cs.CL [Total: 30]
cs.CV [Total: 135]
cs.RO [Total: 11]
cs.DC [Total: 1]
cs.LG [Total: 5]
cs.HC [Total: 4]
cs.CY [Total: 1]
cs.GR [Total: 1]
cs.AI [Total: 10]
cs.IR [Total: 1]
cs.SD [Total: 2]
eess.IV [Total: 4]
cs.CR [Total: 1]

cs.CL [Back]

[1] Inference-Time Mitigation of Adversarial Political Bias in Large Language Models cs.CL | cs.AIPDF

Tejaswi V. Panchagnula, Bruce Coburn, Bryce J. Dietrich, Robert X. Browning, Edward J. Delp

TL;DR: 本文针对大型语言模型在信息检索和摘要任务中可能被对抗性提示注入政治偏见的问题，提出了一种推理时缓解策略。通过结合思维链提示和直接偏好优化，特别是递归自校正方法，显著提升了模型生成摘要的政治中立性。

Details

Motivation: 当前基于人类反馈的强化学习等对齐技术虽能确保LLMs遵循安全指令，但易受对抗性提示注入攻击，导致生成带有政治偏见的不安全内容，而现有方法未专门针对政治偏见进行缓解。

Result: 在立法视频摘要任务上，提出的递归自校正方法将政治中立性李克特量表评分从基线2.14提升至4.56（所有模型平均），有效实现了推理时政治偏见的缓解。

Insight: 创新点在于将思维链提示与直接偏好优化结合用于推理时偏见缓解，特别是递归自校正框架；客观来看，该方法为LLMs对抗特定领域（如政治）的对抗性攻击提供了轻量级、无需重新训练的解决方案。

Abstract: As Large Language Models (LLMs) become the mainstay for information retrieval and summarization tasks, ensuring that they are always non-partisan and invulnerable to political bias is a critical step towards safer and more trustworthy Artificial Intelligence (AI). Current model alignment paradigms, such as reinforcement learning from human feedback (RLHF), make LLMs follow overarching safety instructions. However, this instruction tuning can be exploited via adversarial prompt injection and be used to generate unsafe content. In particular, political bias has not been specifically targeted by modern alignment techniques as harmful and biased content. To address this vulnerability of LLMs, we propose mitigation strategies using Chain of Thought (CoT) prompting and Direct Preference Optimization (DPO). Using a public dataset of legislative videos, we generate summaries using LLMs, inject bias via adversarial prompting and evaluate their performance on a four axis scale designed for political summarization. In this paper, we present different methods to shield LLMs against the injection of political bias. Our results demonstrate that the proposed Recursive Self-Correction approach raises model performance from a Political Neutrality Likert scale baseline of 2.14 to 4.56, averaged across all models, demonstrating effective inference-time mitigation of political bias in LLM-generated summaries.

Hailong Yang, Jianqi Wang, Guanjin Wang, Zhaohong Deng

TL;DR: 本文提出了一种多模态生成模糊系统（MMGFS），这是一个受传统模糊系统启发的、由模糊推理引导的多模态问答框架。它旨在通过多模态协同反思机制缓解模态偏差，并利用模糊规则和多跳推理机制支持跨领域知识融合与分层推理，以解决多模态问答中的模态偏差、知识不确定性和推理深度不足等问题。

Details

Motivation: 现有多模态问答方法（包括传统深度学习模型和大模型/提示框架）面临模态偏差、跨领域知识引入的不确定性以及浅层语义匹配导致的推理深度和可解释性有限等关键挑战。

Result: 在开放域数据集（如MultimodalQA和WebQA）和特定领域基准（如BioMol-VQA和EHRxQA）上的综合评估表明，MMGFS在多个数据集上持续优于现有方法，在答案准确性、一致性和泛化性方面实现了优越性能。

Insight: 主要创新点在于将模糊系统框架引入多模态生成任务，通过多模态协同反思机制缓解模态偏差，并利用模糊规则和多跳推理机制增强不确定性建模和深层语义理解，从而提升了模型的推理深度和可解释性。

Abstract: In Multimodal Question Answering (MQA), models are required to jointly encode and integrate heterogeneous information from multiple modalities, including text, images, and speech, to perform complex semantic reasoning and decision making. Despite recent advances, existing approaches, including traditional deep learning models and Large Models (LMs) or prompt-based frameworks, continue to face several critical challenges. First, modality bias arises from discrepancies in feature distributions across different modalities, which limits effective cross modal collaborative understanding. Second, many questions require knowledge drawn from multiple domains, introducing significant uncertainty. Third, current methods often rely on shallow semantic matching, resulting in limited reasoning depth an reduced interpretability. To address these issues, inspired by the traditional fuzzy system (FS) framework, we propose a fuzzy-inference-guided multimodal generative architecture termed the Multi-Modal Generative Fuzzy System (MMGFS). The main contributions of MMGFS are two folds. First, it alleviates modality bias through a multimodal collaborative rumination mechanism. Second, it introduces fuzzy rules and a multi-hop inference mechanism to support cross-domain knowledge fusion and hierarchical reasoning, thereby strengthening uncertainty modelling and deepening semantic understanding. We conduct comprehensive evaluations on open-domain question answering datasets, including MultimodalQA and WebQA, as well as domain-specific benchmarks, including BioMol-VQA and EHRxQA. Experimental results demonstrate that MMGFS consistently outperforms existing methods across multiple datasets. It effectively mitigates modality bias and question uncertainty while achieving superior performance in answer accuracy, consistency, and generalization.

[3] Characterizing Rhetorical Misalignment in Decision-Making with Language Models cs.CL | cs.AIPDF

Zirui Cheng, Joey Chan, Simo Du, Chenhao Tan, Yue Guo

TL;DR: 本文提出了一种决策理论框架来研究语言模型在决策过程中的修辞错位现象，即LLM在特定决策情境下使用不恰当的修辞表达方式，从而导致人类做出次优决策。通过临床决策实验发现，LLM平均导致2.81%的有害决策反转（从正确答案改为错误答案），并开发了基于LLM模拟决策者的可扩展评估方法。

Details

Motivation: 随着LLM越来越多地融入高风险人机决策，需要理解其输出是否会放大认知偏差、如何影响人类决策，并评估是否会导致有害后果。

Result: 在美国医师执照考试数据集上的实验表明，不同LLM平均导致2.81%的有害决策反转率；基于LLM模拟决策者的计算评估方法实现了对修辞错位的可扩展测量。

Insight: 揭示了即使事实对齐的模型也可能通过修辞呈现造成危害的安全隐患；提出了将修辞错位量化为决策理论框架的创新方法，并发现LLM语言可能诱发锚定效应、权威偏见和损失厌恶等多种认知偏差。

Abstract: Human decision-making is often shaped by a range of well-documented cognitive biases. As large language models (LLMs) become increasingly integrated into high-stakes human-AI decision-making, it is important to understand whether their outputs can amplify potential biases, how this influences human decisions, and crucially, whether it can lead to harmful consequences. In this work, we develop a decision-theoretic framework to study rhetorical misalignment, a failure mode where an LLM uses rhetorically inappropriate forms of presentation for a given decision context, thereby inducing suboptimal human decisions. We empirically investigate this phenomenon through a human-subject experiment in realistic clinical decision-making using a dataset curated from the United States Medical Licensing Examination. By measuring how LLM-generated information affects decisions, we observe that LLMs induce an average 2.81% rate of harmful decision flips across different models, where clinician participants change from a correct to an incorrect answer. Rationales reported by participants provide evidence that these revisions are closely related to the language used by LLMs that may induce different types of cognitive biases, including anchoring, authority bias, and loss aversion. To enable scalable evaluation, we instantiate our theoretical framework using decision-makers simulated by LLMs to computationally measure rhetorical misalignment. Our findings reveal a safety concern previously unrecognized in high-stakes domains: a model can be factually aligned yet still induce harm through its rhetorical presentation.

[4] Beyond Tokens: A Survey on Decoding Methods for Large Language and Vision-Language Models cs.CLPDF

Haoran Wang, Xiongxiao Xu, Philip S. Yu, Kai Shu

TL;DR: 这篇论文是一篇关于大型语言模型和大型视觉语言模型解码方法的综述。它系统性地回顾了通过指导词元级选择、执行序列级生成或并行生成词元来控制模型生成并加速过程的推理时方法，旨在强调解码方法的效率和有效性，并提供其应用的实用视角。

Details

Motivation: 尽管LLMs和LVLMs展示了强大的生成能力，但确保其输出符合用户意图仍然具有挑战性。大多数现有方法在训练阶段解决此问题，而解码方法等推理时方法提供了更高效和可扩展的解决方案。

Result: 本文是一篇综述，未报告具体的定量实验结果或基准测试排名，而是对现有方法进行了系统性回顾，并指出了持续的挑战和未来的研究方向。

Insight: 本文的创新点在于识别并系统梳理了LLMs和LVLMs解码方法的三个新兴范式，为研究者提供了该领域的结构化概览和资源集合。从客观角度看，其价值在于将分散的解码技术整合到一个统一的框架中进行审视，有助于推动该方向的研究。

Abstract: Large language models (LLMs) and large vision-language models (LVLMs) have demonstrated impressive generative capabilities, yet ensuring their outputs align with user intent is still challenging. While most existing approaches address this issue at the training stage, inference-time approaches like decoding methods offer a more efficient and scalable solution. Decoding methods control model generation by guiding token-level selection, performing sequence-level generation, or generating tokens in parallel to accelerate the process. In this survey, we identify three emerging paradigms from recent works on decoding methods for LLMs and LVLMs, provide a systematic review of these methods, highlight ongoing challenges, and discuss potential future research directions. Our goal is to underscore the efficiency and effectiveness of decoding methods and offer a practical view of their applications. Paper lists and more resources on decoding methods for LLMs and LVLMs can be found at https://github.com/wang2226/Awesome-LLM-Decoding.

[5] RamseyGadgets: A Graph Construction Dataset for LLMs cs.CL | cs.AIPDF

Zohair Raza Hassan, Deepak Pandita

TL;DR: 该论文提出了一个名为RamseyGadgets的新型图构造数据集，包含70个未被充分探索的、需要寻找具有特定性质（如包含固定颜色边）的Ramsey-good图的问题。该数据集旨在评估大型语言模型（LLMs）在图构造任务中的推理能力，而非对训练数据的记忆。作者评估了五个开源LLMs在该数据集上的表现，发现它们在困难问题上的准确率仅为37.70%，并展示了如何利用该数据集分析提示信息对模型性能的影响。

Details

Motivation: 动机是测试LLMs是否能够利用其推理能力来构造具有指定性质的图，而非仅仅回忆训练数据。由于许多经典的图构造问题（如寻找极值Ramsey-good图）在文献中已被广泛研究，难以区分LLM的输出是推理结果还是记忆，因此需要一个新的、未被充分探索的数据集来进行有效评估。

Result: 在提出的RamseyGadgets数据集上评估了五个开源LLMs，结果显示它们在困难问题上的总体准确率仅为37.70%，其中Gemma-4-31B模型在五个模型中表现最佳。数据集中的问题解决方案规模适中（最多10个顶点），可通过SAT求解器自动验证。

Insight: 论文的创新点在于构建了一个新颖、可扩展的图构造数据集，该数据集专注于未被充分探索的Ramsey-good图变体问题，从而能够更纯粹地评估LLMs的推理能力而非记忆。该数据集的设计允许通过改变要避免的单色子图来轻松扩展，并为分析不同提示如何帮助LLMs解决此类问题提供了平台。

Abstract: Constructing special graphs is an important task within graph theory and computer science. Many popular graph constructions are the result of a comprehensive exploration of relevant graphs and human ingenuity. Given the rise of generative AI usage in mathematics, it is natural to test whether LLMs are able to construct graphs with specified properties using their reasoning capabilities. Unfortunately, many natural graph construction problems, such as finding extremal Ramsey-good graphs (i.e., avoiding specific monochromatic subgraphs), have been explored extensively in the literature, making it difficult to ascertain whether a construction is the product of an LLM’s reasoning capabilities or its recollection from training data. In this work, we introduce \textbf{RamseyGadgets}, a novel dataset of 70 underexplored graph construction problems that require finding Ramsey-good graphs with special properties (e.g., containing an edge with a fixed color). These problems have reasonably sized solutions (at most 10 vertices) that can be verified by SAT solvers, making them suitable for automatic evaluation. Our dataset is easily expandable, as one can simply change the monochromatic subgraphs being avoided to obtain a new set of problems. We evaluate the performance of five open-source LLMs on our dataset and report the results. Our findings show that LLMs achieve only 37.70% accuracy on the hard-tier problems in our dataset, with Gemma-4-31B achieving the highest performance out of the five. We also showcase how our dataset allows us to ascertain what kind of hints help LLMs perform better at this task.

[6] Interpretable Cross-Lingual Alignment in Small Language Models: Probing Cultural and Pragmatic Reasoning in Japanese-English Bilingual LLMs cs.CL | cs.LGPDF

Florian Braun

TL;DR: 本文针对日语-英语双语小语言模型（TinySwallow-1.5B）中跨语言对齐的可解释性问题，提出了J-PragEval-v0基准来评估日语语用现象（如敬语、内外群体指代等），并采用线性探测和教师强制对数概率分析模型内部表示。研究发现不同语用特征在模型中的编码方式各异，并提出了无需参数的推理时方法‘语用表示引导’来编辑残差流激活。

Details

Motivation: 解决大语言模型在远离英语类型学的语言（如日语）上表现不佳且难以理解的问题，特别是模型在日语语用现象（如敬语、语境敏感礼貌等）上的失败，现有评估方法（如JGLUE）无法将这些现象与表面流畅性分离。

Result: 在TinySwallow-1.5B模型上，敬语记录在残差流中线性可解码（层15平衡准确率0.96），隐式主语和内外群体指代虽线性不可解码但生成时对比明显（翻转率0.77-0.79），间接拒绝则因混淆礼貌与生成长度而表现不佳（翻转率0.43）。对比激活添加基线在存在线性信号时接近逻辑回归准确率。

Insight: 创新点包括设计J-PragEval-v0基准以隔离日语语用现象进行细粒度评估，以及提出语用表示引导方法通过编辑残差流激活来引导模型推理；客观分析认为研究揭示了小模型中语用知识的分层编码机制，为跨语言对齐的可解释性提供了新视角。

Abstract: Large language models work well on English and behave in poorly understood ways on languages typologically far from it. Japanese is a clean example, where evaluation still leans on translation quality and JGLUE-style benchmarks, which roll lexical, syntactic and pragmatic competence into a single score. The phenomena on which general-purpose models fail Japanese users are pragmatic: honorifics, in-group and out-group reference, context-sensitive politeness, zero anaphora. I introduce J-PragEval-v0, a minimal-pair benchmark isolating four such phenomena from surface fluency, and combine it with linear probes and teacher-forced log-probability evaluation to ask where inside TinySwallow-1.5B (28 layers, hidden size 1536) the corresponding contrasts live. The four features split three ways. Honorific register sits cleanly in the residual stream: 0.96 balanced accuracy at layer 15, and the model flips its preferred continuation with the scenario on 93 percent of items. Implicit subject and in-group reference are not linearly decodable at the final prompt token (0.48 and 0.38), yet flip rates are 0.77 and 0.79, so the contrast is worked out during generation rather than stored at the prompt. Indirect refusal is the negative case: 0.95 probe accuracy collapsing to a 0.43 flip rate under length-normalised teacher forcing, because the current minimal pairs conflate politeness with continuation length. I also specify Pragmatic Representation Steering, a parameter-free inference-time method that edits residual-stream activations along the class-mean-difference directions probing identifies. Feasibility is argued indirectly rather than demonstrated: the contrastive activation addition baseline, the same geometry the method would inject, recovers probe accuracy within one to two points of logistic regression wherever a linear signal exists. Scaling to Llama-3.1-Swallow-8B is the next step.

[7] Handoff-H1: An Orchestrated Vision-Agent System for Material Quantity Takeoff from Construction Blueprints cs.CL | cs.AI | cs.CVPDF

Bruno Chicelli, Henrique Alves, Rodrigo Anselmo, Joshua Weinberg, Felipe Lemos

TL;DR: 论文提出了Handoff-H1系统，这是一个用于从建筑蓝图中自动进行材料数量估算的视觉-智能体协同系统。该系统通过三层架构实现：专用计算机视觉模型提取图元、配备图像操作和内部视觉任务工具的智能体、以及基于建筑知识库构建的持久化分层项目基础。在Construction Blueprint Takeoff Benchmark上评估，该系统在端到端处理原始PDF时，综合得分达到81.6%，超越了前沿智能体模型和独立专业估算师。

Details

Motivation: 解决从一套建筑蓝图到完整材料数量估算的自动化问题，该任务需要跨图纸的视觉感知、维度与多跳推理，以及对图纸未明确说明的建筑惯例的理解。

Result: 在Construction Blueprint Takeoff Benchmark（包含10套真实住宅蓝图及专家验证的估算结果）上评估。Handoff-H1从原始PDF端到端处理，综合得分为81.6%（覆盖率为86.1%，数量精度P@.25为78.8%），比最强前沿智能体高出约20个百分点，并在覆盖率和精度上超越了独立专业估算师（77.6%）。

Insight: 创新点在于构建了一个三层协同系统，将专用CV模型、工具使用智能体与基于领域知识的结构化项目基础相结合，实现了对复杂、隐含领域知识任务的端到端自动化处理。其评估框架和系统架构为类似视觉-语言-推理任务提供了可借鉴的范式。

Abstract: Converting a set of architectural blueprints into a complete material quantity takeoff requires visual perception across drawing sheets, dimensional and multi-hop reasoning, and grounding in construction conventions that the drawings never state. We present Handoff-H1, a takeoff system built from three layers: purpose-built computer-vision models that extract primitives; tool-using agents equipped with image operations and in-house visual-task tools, including CV-model-backed counting, detection and plan decomposition; and a persistent, hierarchically structured project foundation, grounded in a curated construction knowledge base. We evaluate on the Construction Blueprint Takeoff Benchmark: 10 real residential blueprint sets paired with consensus-validated expert takeoffs - 2,009 verified line items, restricted for scoring to the 1,348 primary-tier materials that drive an estimate - scored per trade by an LLM judge on material coverage and quantity Precision@25% (P@.25) and combined into a weighted composite. Under identical scoring from the raw PDF, seven frontier and open-weight models span composites of 35-61, and independent professional estimators - scored against the same reconciled gold standard - post 77.6% (65.5% coverage, 87.9% P@.25). Handoff-H1, working end-to-end from the raw PDF, reaches 81.6% (86.1% coverage, 78.8% P@.25): roughly 20 points above the strongest frontier agent, and above the independent estimators by pairing near-human quantity precision with coverage they do not reach. The evaluation harness is public for the open harbor framework; the blueprint sets and ground truth are available upon request for research use.

[8] TRACE-BN: Transferring Bangla-English Tutoring Behavior to a Sub-1B Offline Language Model cs.CLPDF

Khan Raiyan Ibne Reza, Sanjana Aktar Maria, Mohammad Tushar Abdullah, Asfee Bhuiyan Leen, Sumaiya Tabassum Nimi

TL;DR: 本文提出了TRACE-BN，一个为孟加拉语英语学习者（CEFR A1-A2级别）构建的结构化辅导轨迹数据集，并成功将这种多组件的辅导行为迁移到一个参数量小于10亿（sub-1B）的离线语言模型Qwen3-0.6B上。

Details

Motivation: 孟加拉语英语辅导不仅需要正确的翻译，还需要语法差异解释、学习者错误预测和针对性练习，现有模型难以在资源受限的离线环境下满足这些需求。

Result: 在保留测试集上，模型输出的模式有效性从85.4%提升至95.8%；与教师模型参考相比，chrF++从15.28提升到34.77，BLEU从4.52提升到21.03。人工评估也显示在翻译、语法解释、错误诊断和练习对齐方面均有改善。

Insight: 主要创新点在于利用课程大纲引导生成结构化的辅导轨迹数据集，并通过LoRA与4位量化技术，成功将复杂的多组件教学行为迁移到资源受限的小型离线模型上，证明了结构化监督的有效性。

Abstract: Bangla-English tutoring requires more than producing a correct translation: learners also need explanations of grammar differences, awareness of their likely errors, and targeted practice. We present TRACE-BN, a curriculum-guided dataset of structured tutoring traces for Bangla-speaking learners of English at the CEFR A1-A2 level. Each trace combines word-level glosses, literal and natural translations, Bangla grammar explanations, a plausible learner error, and a targeted practice question with its answer. The traces are generated by Gemini 3.5 Flash Lite as the teacher model from NCTB Classes 9-10 English curriculum units, then filtered for structural validity, script integrity, and semantic duplication. We transfer the resulting structured tutoring behavior to Qwen3-0.6B using LoRA with 4-bit quantization for resource-constrained offline deployment. On held-out inputs, schema validity increases from 85.4% to 95.8%, while, against teacher-model references, chrF++ improves from 15.28 to 34.77 and BLEU from 4.52 to 21.03. Field-level evaluation by two independent judges shows improvements across translation, grammar explanation, learner-error diagnosis, and practice alignment, while a human audit supports the quality of the supervision data. The results show that curriculum-guided structured supervision can transfer multi-component tutoring behavior to a sub-1B model under these resource constraints. The dataset, model checkpoints, and code are publicly available at https://huggingface.co/datasets/RaiyanKhaan/Trace-BN

[9] Why Vision Fails as a Universal Bridge: Rectifying Modality Asynchrony in Multilingual MLLMs cs.CLPDF

Yihang Du, Juhao Liang, Zhengzhao Lai, Siyu Li, Yan Hu

TL;DR: 本文研究了多模态大语言模型在非英语视觉推理任务中性能下降的问题，发现其根源在于模态异步现象，即语言翻译在模型早期层已完成，而视觉语义化尚未成熟。为此，作者提出了ANCHOR训练框架，通过主动视觉锚定技术加速早期视觉语义的出现，从而恢复视觉信号在翻译过程中的因果影响，并在多个基准测试中取得了优于基线的性能。

Details

Motivation: 尽管纯文本骨干模型展现出强大的多语言能力，但多模态大语言模型在非英语视觉推理任务中性能显著下降，其背后的多模态机制尚未被探索。

Result: 在XMMMU、MaXM和CVQA基准测试上的广泛实验表明，ANCHOR框架在微调和零样本语言设置下均持续优于标准基线，实现了稳健的跨语言视觉推理。

Insight: 论文揭示了多模态大语言模型中存在的’幽灵锚点’现象，即模态异步导致早期视觉信号功能不可见，并创新性地提出了通过主动视觉锚定来加速视觉语义化，从而对齐模态，这一机制干预方法为解决跨语言多模态对齐问题提供了新思路。

Abstract: Multimodal large language models (MLLMs) exhibit substantial performance degradation in non-English visual reasoning, despite the strong multilingual competence of their text-only backbones. While mechanistic evidence from text-only models suggests that non-English inputs are routed through an English-centric latent space, the multimodal implications of this phenomenon remain unexplored. Through rigorous mechanistic analysis, we identify the \textbf{Ghost Anchor} phenomenon: a temporal modality asynchrony where linguistic translation to the English semantic manifold completes in early layers, while visual semanticization remains immature. Consequently, visual signals are physically present yet functionally invisible during the early alignment window. To rectify this, we propose \textbf{ANCHOR}, a training framework employing Proactive Visual Anchoring (PVA) to accelerate early visual semantic emergence, ensuring visual representations proactively guide linguistic translation. Mechanistic interventions confirm that ANCHOR successfully restores the causal influence of visual signals during early translation. Furthermore, extensive experiments on XMMMU, MaXM, and CVQA demonstrate that ANCHOR consistently outperforms standard baselines, achieving robust visual reasoning across both fine-tuned and zero-shot languages.

Catherine Bao, Vivek Srikumar

TL;DR: 该论文研究了大型语言模型是否与人类共享时间错觉。通过构建包含6,684个叙事对的基准测试，涵盖五种时间错觉，发现人类仅在两种错觉中表现出预期偏好，而14个LLM在四种错觉中选择了文献预测的场景，与人类行为存在差异。分析表明，模型的这种对齐可能源于对已发表心理学研究的检索，而非类人的时间偏差。

Details

Motivation: 动机是探究仅通过书面叙事是否能引发人类时间错觉，并检验LLM是否表现出与人类相似的主观时间感知偏差，以深入理解模型内部的时间表征机制。

Result: 在涵盖五种时间错觉的基准测试上，人类参与者仅在两种文本中操纵直接可见的错觉中表现出预期偏好；而评估的14个LLM在四种错觉中选择了文献预测的正确场景，与人类行为不一致。推理轨迹显示约70%的模型响应明确引用了心理学研究。

Insight: 创新点在于构建了首个基于叙事文本的时间错觉基准，揭示了LLM的时间判断可能严重依赖对已知科学事实的检索而非内部模拟，这挑战了模型具有类人主观时间感知的假设，为评估模型的世界模型真实性提供了新视角。

Abstract: Human perception of time is subjective. Well-documented temporal illusions show that the brain relies on context and relational cues for judging duration instead of tracking elapsed time directly. Prior studies established these effects with visual and auditory stimuli. Existing LLM evaluations of temporal perception focus on estimating event durations or multi-step temporal reasoning. In this work, we investigate whether written narratives alone can evoke human temporal illusions, using a new benchmark of 6,684 narrative pairs spanning five illusions. We find that human readers (60 participants) prefer expected scenarios in only two of the five illusions, those where the manipulation is directly visible in text rather than requiring readers to internally simulate duration. We evaluate 14 LLMs on the same benchmark. Surprisingly, we find that models pick the literature-predicted scenario across four of the five illusions, diverging from human behavior. Reasoning traces show that ~70% of responses explicitly evoke psychology research, suggesting that this alignment is consistent with retrieval of published findings rather than human-like temporal biases.

[11] Do Language Models Consistently Encode the Current Year? cs.CL | cs.LGPDF

Suze van Adrichem, Aditi Bhaskar, Diyi Yang, Christopher Potts, Jing Huang

TL;DR: 这篇论文研究了语言模型如何编码当前年份的概念，提出了两种概念上不同的探测任务：关联任务（通过动词时态推断当前年份）和陈述任务（直接查询当前年份）。研究发现，指令微调后的语言模型在这两个任务上都能准确估计到训练数据截止年份的一年内，而基础模型的关联任务预测可作为预训练数据截止年份的强代理指标。然而，两种任务在内部机制上存在分歧，导致更新当前年份时面临挑战，提示、监督微调或权重编辑等方法均无法同时成功调整两种年份表示。

Details

Motivation: 论文的动机是探究语言模型是否一致地编码当前时间的概念，这对于时间推理至关重要，但当前模型如何表示当前时间尚不明确。

Result: 在实验中，指令微调模型在两个任务上估计的当前年份与训练数据截止年份的平均误差在一年内；基础模型的关联任务预测作为预训练数据截止年份的代理，在13个模型上平均误差仅为10个月。但更新尝试显示，提示方法能成功更新陈述年份（94.6%成功率），却几乎无法改变关联年份（1.7%成功率）；监督微调和权重编辑也未能同时有效调整两种年份表示。

Insight: 论文的创新点在于揭示了语言模型中当前年份编码的不一致性：关联概念（深植于预训练学到的语言结构）和陈述概念（后训练习得）使用不同的因果机制，导致更新困难。这为模型时间表示的可控性和可编辑性提供了重要见解。

Abstract: A consistent concept of the current time is important for temporal reasoning, yet how language models represent the current time is not well understood. We contribute two tasks that probe the current year in conceptually distinct ways: an associative task, which infers the current year from verb tense, and a declarative task, which directly queries for the current year. Both tasks estimate current years within one year of the post-training data cutoff of instruction-tuned language models. For base models, predictions on the associative task serve as a strong proxy for the pre-training data cutoff, with an average error of only 10 months across 13 models. However, their internal mechanisms diverge: the associative task uses mechanisms similar to factual recall, while the declarative task lacks consistent causal pathways. This divergence poses a challenge for updating the current year in language models. None of prompting, SFT, or weight editing succeed in shifting the associative and declarative years simultaneously. Prompting updates the declarative year (94.6% success across 351 target years) but leaves the associative year nearly unchanged (1.7% success). Year-shifted SFT also fails to shift the associative year, matching the target year in only one of eight models. Weight editing, while effective for both tasks individually, does not generalize across both. Overall, our results show that the current year is not consistently encoded in language models: The associative notion, deeply ingrained in linguistic structures learned in pre-training, uses different causal mechanisms and resists the same modifications that easily shift the declarative notion learned in post-training.

[12] Why Summaries Turn Neutral: Policy Attribution for Sentiment Drift in Reinforcement Learning from Human Feedback cs.CLPDF

Mikhail Krasitskii, Alexander Gelbukh, Olga Kolesnikova, Grigori Sidorov

TL;DR: 本文研究了基于人类反馈的强化学习（RLHF）在文本摘要任务中导致的情感漂移问题，即模型生成过于中性、缺乏情感细微差别的摘要。论文提出了策略归因框架，通过梯度和logit分解追踪漂移原因，发现其源于奖励模型信号和KL惩罚。作者还提出了一种情感感知正则化方法，有效减少了情感漂移。

Details

Motivation: RLHF虽然能提升大语言模型摘要的流畅性和安全性，但会导致情感漂移，即摘要变得过度中性，丢失了原文的情感细微差别。本文旨在诊断RLHF为何会充当情感中和器，并理解其内在机制。

Result: 在Reddit TL;DR和CNN/DailyMail数据集上，RLHF生成的摘要获得了更高的奖励分数，但其情感方差降低了30-40%。跨八种语言的实验表明情感漂移具有语言无关性，且形态更丰富的语言情感抑制更严重。提出的情感感知正则化方法将漂移减少了18-22%，且未损害摘要质量。

Insight: 论文的创新点在于提出了策略归因框架来诊断RLHF中的情感漂移，揭示了其源于奖励模型不确定下的“低风险”策略。提出的情感感知正则化方法为解决RLHF的副作用提供了一种有效且可推广的技术路径。

Abstract: Reinforcement learning with human feedback (RLHF) aligns LLMs with human preferences, improving summarization fluency and safety, but causes sentiment drift: overly neutral summaries stripped of emotional nuance. We diagnose why RL acts as a sentiment neutralizer and present Policy Attribution, a framework using gradient and logit decomposition to trace drift to reward model (RM) signals and KL (Kullback-Leibler) penalty. Sentiment drift reflects a strategic bias toward “low-risk” tokens maximizing expected rewards under preference uncertainty (Stiennon et al., 2020; Gao, Schulman, and Hilton, 2023). On Reddit TL;DR and CNN/DailyMail, RLHF summaries get higher rewards but show 30-40% lower sentiment variance. Cross-lingual analysis across eight languages shows language-independent drift, with morphologically richer languages more suppressed (Krasitskii et al., 2026). We propose and validate a sentiment-aware regularization technique reducing drift by 18-22% without harming summary quality. The code and toolkit will be public.

[13] TaoLive Digital Avatar Agent Technical Report: Training Agents to Evolve with Their Harness cs.CLPDF

TaoLive AIGC LLM Team, Yuhan Sun, Wenhao Lin, Yongdong Luo, Yibo Hu

TL;DR: 本文提出了一种名为Harness-Aware Training（HAT）的方法，用于训练可演化的数字人直播代理。该方法通过将系统工具、提示词等运行时环境（Harness）的状态作为训练分布的一部分，解决了传统微调模型难以适应动态业务策略更新、而零样本大模型又延迟过高的问题。

Details

Motivation: 解决电商直播中AI数字人主播面临的挑战：需要实时、准确、低延迟地回答问题并适应快速变化的业务策略、合规要求和风格，而传统方法在模型适应性与推理速度之间存在矛盾。

Result: 在包含4500多个案例的四个评估集上，采用HAT训练的35B紧凑模型在真实直播问答（Live-Stream QA）上得分为94.8，优于基础模型（80.3）和评估中最强的通用LLM（93.0）；在Harness变体问答上得分为94.6，并在IFEval上保持83.5分，而固定Harness的SFT会使IFEval下降7.7分。在单块NVIDIA H20 GPU上，系统P50延迟为3.407秒，P95延迟为8.114秒。

Insight: 核心创新在于提出了Harness-Aware Training（HAT）框架，通过Harness状态增强（HSA）将运行时环境的变化纳入训练过程，使紧凑模型既能快速推理，又能灵活适应系统配置的更新，而不牺牲通用指令遵循能力。这为构建可演化的实时AI代理提供了一种有效的训练范式。

Abstract: AI-powered digital-avatar streamers in live e-commerce must answer product questions, engage viewers, and execute changing business strategies in real time. This requires low latency, factual and effective replies, and rapid adaptation to updated campaign, compliance, and style requirements. We develop an evolvable Harness that decouples Skills, Hooks, system prompts, and tools from model weights, allowing runtime behavior to change without retraining. However, Harness evolution creates a moving execution environment: compact models fine-tuned on one configuration may memorize names, schemas, and prompt templates rather than follow the Harness currently provided, while stronger zero-shot models are too slow for real-time use. We address this tension with Harness-Aware Training (HAT), which makes Harness states part of the training distribution. HAT applies task-preserving Harness-State Augmentation (HSA) to Skills, tool schemas, prompt structures, and interaction constraints, and comprises three stages: HSA-based supervised fine-tuning, general on-policy distillation to recover general capabilities, and HSA-based agentic reinforcement learning in a production-informed live-room simulator. Across four evaluation sets with more than 4,500 cases, our compact 35B model scores 94.8 on real-world Live-Stream QA, versus 80.3 for the base model and 93.0 for the strongest evaluated general LLM, while scoring 94.6 on Harness-Variant QA and retaining 83.5 on IFEval. By contrast, fixed-Harness SFT reduces IFEval by 7.7 points. In a controlled complete-agent replay on one NVIDIA H20 GPU with MTP enabled, the system achieves 3.407 s P50 and 8.114 s P95 latency. These results show that HAT produces a latency-feasible compact agent that remains effective under evaluated Harness changes without sacrificing general instruction following.

[14] MicroVerse: An Instrument for Measuring Self-Authored Identity Drift in Long-Horizon Multi-Agent Language-Model Simulations cs.CLPDF

Sky Ng, Brihi Joshi, Ishan Gupta, Shirley Huang, Zonglin Di

TL;DR: 该论文提出了MicroVerse，一种用于测量长周期多智能体语言模型模拟中身份漂移的行为科学工具。该工具通过一个资源稀缺的网格环境，迫使智能体在生存压力下行动，并利用三层记忆架构和定期反思机制，来量化智能体与其初始核心身份（’灵魂文件’）的偏离程度。

Details

Motivation: 当前缺乏有效的工具来测量在长周期、多智能体的语言模型模拟中，被赋予特定人设的智能体在持续压力下是否能保持其身份一致性（即身份漂移问题）。

Result: 通过一个受控的种子运行（n=25）和反思阈值扫描实验，论文发现两个主要结果：1）反自我欺骗是身份修改中最大的语义类别（占新增边界的24%）；2）系统具有阈值鲁棒性，即更低的反思阈值会加速并增加修订频率，但不改变漂移的方向。所有结果均为初步的效应验证，而非统计显著性结论。

Insight: 论文的创新点在于设计了一个专门用于测量身份漂移的标准化工具（MicroVerse），其核心包括：不可变的’灵魂文件’、资源稀缺的生存压力环境、三层记忆与反思机制、解耦测量与行为的快照系统，以及使用基于释义感知和价值锚定的多寄存器差异（而非原始余弦相似度）来离线评分身份漂移。

Abstract: Long-horizon, multi-agent language model (LM) simulations are widely proposed for studying social behavior, yet instruments to measure whether persona-conditioned agents maintain identity fidelity under sustained pressure are lacking. We present MicroVerse, a behavioral-science instrument that measures identity drift in generative agents. Agents carry an immutable “soul file” (core values, moral boundaries, personality, goals) and inhabit a resource-scarce 50 x 50 environment where water is a non-respawning survival constraint. Scarcity is operationalized via a per-tick existence-cost gradient. The eight-verb action space maps directly to moral boundaries (trade, talk, attack, scavenge). Using a three-layer memory architecture, agents periodically revise a mutable current identity against their immutable original soul via importance-triggered reflection. To mitigate survivor bias, MicroVerse decouples measurement from behavior using uniform longitudinal engine snapshots every N ticks alongside a forced-end snapshot of all living and dead agents. Identity drift is scored offline using a paraphrase-aware, value-anchored, multi-register diff rather than raw cosine similarity. We evaluate the instrument via a controlled seed run (n = 25) and a reflection-threshold sweep (thresholds {40, 80, 150}) to determine if drift dynamics are gate artifacts or threshold-robust properties. We report two primary findings: (1) Anti-self-deception emerges unprompted as the single largest semantic category of identity modification (27 of 111 added boundaries, 24%). (2) The system is threshold-robust; lower gates accelerate and increase revision frequency but preserve drift direction. All empirical results are strictly preliminary existence proofs and effect shapes (one model, one seed per arm, n = 25) rather than statistical significance claims.

[15] SEER: Long-Context Reasoning via Selective Visual-Text Compression cs.CL | cs.CVPDF

Jiawei Xu, Zhilin Zhai, Jinrui Fang, Ruohan Xu, Mingfei Lu

TL;DR: 本文提出了SEER框架，通过选择性视觉-文本压缩来解决大语言模型处理长上下文时计算开销大的问题。SEER通过视觉扫描选择与查询相关的图像，仅在需要时检索文本内容，结合了视觉压缩的效率和基于文本推理的精度。在长上下文基准测试中，SEER在保持平均提示令牌节省的同时提高了提取精度。

Details

Motivation: 长上下文推理因注意力机制的二次复杂度而计算昂贵，现有视觉-文本压缩方法采用统一压缩，可能牺牲需要详细提取时的精度。

Result: 在LongBench基准测试中，SEER实现了51.11%的平均准确率，比视觉-文本基线Glyph-9B高出2.33个百分点，比Qwen3-8B高出3.49个百分点。

Insight: 创新点在于学习选择性压缩，通过监督微调工具交互轨迹来适应性地调用工具进行选择和检索，实现了效率与精度的平衡。

Abstract: Long-context reasoning remains computationally expensive for large language models due to the quadratic complexity of attention over text tokens. Visual-text compression offers a promising alternative by rendering text into images and processing them with vision-language models, often reducing token usage. However, existing approaches apply uniform compression regardless of query relevance, potentially sacrificing precision where detailed extraction is required. We present SEER, a framework that learns to select query-relevant images through visual scanning and retrieve textual content only where needed, combining the efficiency of visual compression with the precision of text-based reasoning. Through supervised fine-tuning on tool-interaction trajectories, SEER learns adaptive tool invocation for selection and retrieval. Experiments on long-context benchmarks show that SEER improves extraction precision through selective text retrieval while retaining average prompt-token savings relative to full-text baselines. On LongBench, SEER achieves 51.11% average accuracy, outperforming the visual-text baseline Glyph-9B by 2.33 points and Qwen3-8B by 3.49 points. Code can be accessed at https://github.com/jiaweixu98/SEER

[16] LLMs Get Smarter from Targeted Synthetic Multilingual Data cs.CL | cs.AIPDF

Ishika Agarwal, Arkajyoti Charaborty, Tanner Sorensen, Neha Gupta, Andreas Stolcke

TL;DR: 本文提出了一种名为HOTFIXR的数据生成框架，旨在解决大语言模型中的语言特定能力差异问题。该框架通过探测模型的多语言弱点并生成针对性合成数据来提升模型的多语言推理性能。

Details

Motivation: 针对大语言模型在不同语言提示下表现不一致的语言特定能力问题，现有方法要么牺牲语言表达能力，要么降低整体性能。本文旨在通过数据中心的视角，生成能够针对性弥补模型弱点的多语言合成数据。

Result: 在三个分布内任务、三个分布外任务和四种分布外语言的评估中，HOTFIXR平均将分布内性能提升6.2%，将微调引发的分布外任务灾难性遗忘减少3.7%，并将分布外语言性能提升7.1%。

Insight: 创新点在于提出了一个通过模型自我探测弱点并生成针对性合成训练数据的数据中心框架，而非依赖语言平衡数据或英语路由，从而在提升多语言性能的同时避免了现有方法的局限性。

Abstract: Language-specific competency (LSC) is the phenomenon of a language model performing better or worse depending on the language of the prompt. In other words, a language model outputs different (and potentially incorrect) responses to the same semantic query when prompted in different languages. Prior work attributes this to an internal misalignment of semantic representation across languages. Currently, there are two main approaches to address LSC in the literature: (1) routing all queries through English, improving performance, but limiting language expressivity to English; or (2) training on language-balanced data, equalizing model performance across languages, but reducing overall performance. In this work, we take a data centric perspective and introduce HOTFIXR: Hardness Optimized Training data For Improving X-Lingual Reasoning. It is a data generation framework that uses models to probe and learn a student model’s multilingual weaknesses, and generates data to mitigate them. HOTFIXR can generate multilingual synthetic training data that can improve multilingual performance. We evaluate on three in-distribution tasks, three out-of-distribution tasks, and four out-of-distribution languages. On average, HOTFIXR (1) improves in-distribution performance by 6.2%, (2) reduces catastrophic forgetting (induced by fine-tuning) on OOD tasks by 3.7%, and (3) on OOD languages by 7.1%. Overall, as many real-world applications requires multilingual LLMs, our work contributes to the efforts of making LLMs multilingually proficient. We will release code upon acceptance.

[17] From Sequence to Structure: Relational Uncertainty Propagation for LLM Agents cs.CL | cs.AIPDF

Zhengzhao Ma. Boxi Cao, Yaojie Lu, Hongyu Lin, Xianpei Han, Le Sun

TL;DR: 本文提出RUPA（面向智能体的关系不确定性传播框架），一种轨迹级不确定性量化方法，用于提升LLM智能体在复杂交互环境中的可靠性。该方法将执行历史建模为有向轨迹图，通过传播不确定性来捕捉跨步骤的风险累积，并结合行为特征与目标对齐信息生成整体置信度估计。

Details

Motivation: 现有不确定性量化方法主要依赖局部信号（如词元概率、预测熵），忽略了长程依赖关系中错误在轨迹上的累积，导致无法识别由早期推理或交互步骤引发的智能体失败。

Result: 在τ-2、Terminal-Bench-2和GAIA等代表性智能体基准测试上，使用6个开源LLM的实验表明，RUPA在提供更准确的不确定性估计、实现更早的失败检测以及改进不确定性引导的智能体执行方面，均持续优于现有方法。

Insight: 核心创新在于显式建模执行轨迹中状态与交互间的时序和语义依赖关系，通过图结构传播不确定性以捕捉风险的长程传递，为长视野LLM智能体提供了更可靠的不确定性量化基础。

Abstract: Reliable uncertainty quantification (UQ) is essential for deploying large language model (LLM) agents in complex interactive environments. Existing UQ methods largely rely on local signals, such as token probabilities, predictive entropy, or per-step confidence, and therefore overlook the long-range dependencies through which errors accumulate across an execution trajectory. As a result, they may fail to identify agent failures whose causes originate several reasoning or interaction steps before the final answer. We propose RUPA (Relational Uncertainty Propagation for Agents), a trajectory-level UQ framework for LLM agents. RUPA represents an execution history as a directed trajectory graph in which reasoning states, tool interactions, and environment feedback are nodes connected by temporal and semantic dependency edges. It then propagates uncertainty over this graph to capture how execution risk accumulates and transfers across interaction steps. The propagated signal is combined with trajectory-level behavioral features and goal-alignment information to produce a confidence estimate for the full agent trajectory. We evaluate RUPA on representative agent benchmarks, including $τ$-2, Terminal-Bench-2, and GAIA, using 6 open-source LLMs spanning multiple model families. Experimental results show that RUPA consistently outperforms existing UQ methods by providing more accurate uncertainty estimates, enabling earlier failure detection, and improving uncertainty-guided agent execution across diverse agent tasks. These results demonstrate that explicitly modeling relational dependency is crucial to reliable UQ for long-horizon LLM agents, providing a practical foundation for trustworthy agent execution.

[18] $R^3$-Bench: LLMs Struggle with Resource-Rational Reasoning under Shared Budgets cs.CLPDF

Peisong Wang, Zhiwei Ma, Bowen Liu, Feixue Liu, Aochuan Chen

TL;DR: 该论文提出了R^3-Bench基准测试，用于评估大型语言模型在共享计算预算下的资源理性推理能力。研究发现，模型在独立任务上表现出的能力与在共享预算下的实际表现之间存在显著差距，即使简单的固定调度策略也常常优于模型自身的动态决策。

Details

Motivation: 现有基准测试通常为每个任务分配独立预算，而共享预算下的研究未能将模型在任务套件中的表现与其在单个问题上已证明的能力进行校准对比。论文旨在填补这一空白，评估LLMs在有限计算资源下的多任务决策效率。

Result: 在涵盖数学、竞赛编程和抽象推理的72个主要测试单元中，基于单问题响应曲线构建的经验Oracle均值在所有单元中都匹配或超过了模型竞赛均值，且在71个单元中严格更高。在中等无工具压力下，六分之四的模型中，均等分配重放策略也优于竞赛表现。

Insight: 创新点在于引入了校准单问题能力的离线经验Oracle作为性能上限，并设计了共享预算评估框架。核心发现是LLMs在资源受限的多任务环境中策略更新有限，其动态调度能力未能充分利用已证明的单问题解决能力，这揭示了当前模型在资源理性推理方面的系统性缺陷。

Abstract: In cognitive science, resource rationality asks how an agent should allocate limited computation to maximize expected value. Most reasoning and agent benchmarks use independent per-task budgets; existing shared-budget studies do not calibrate suite performance against the same model’s demonstrated single-problem competence. We introduce $R^3$-Bench, which evaluates six-problem suites under shared budgets across mathematics, competitive programming, and abstract reasoning in tool-free and agentic settings. Matched single-problem response curves define an offline empirical oracle over observed successes. Across 72 main-table cells for six models, the oracle mean matches or exceeds the contest mean in all cells and is strictly higher in 71. Under moderate tool-free pressure, equal-allocation replay also exceeds contest performance for four of six models. Trajectory diagnostics reveal limited strategy updating and pressure-dependent failure patterns. In a three-model diagnostic under strong agentic pressure, at least one fixed scheduler exceeds the contest mean in six of nine cells, but no policy dominates across domains. These results expose a persistent gap between demonstrated competence and shared-budget realization.

[19] Skill2Query: Exploiting Skill Structure to Generate Pseudo-Queries for Agent Skill Retrieval cs.CL | cs.IRPDF

Lihui Ding, Zihan Guo, Bingwei Lu, Chenyu Zhou, Yuanjian Zhou

TL;DR: Skill2Query是一个用于缓解智能体技能检索监督瓶颈的框架，通过将技能文档解析为技能知识图谱，并采用风格模仿、查询模板生成和参数填充的三阶段过程来生成伪查询，以增强检索信号。

Details

Motivation: 现有文档级伪查询生成方法通常隐含技能内部能力、参数和使用示例之间的丰富关系，导致生成的查询可能主题相关但缺乏能力基础和参数一致性，因此需要显式利用技能文档的内部结构来产生更有效的检索信号。

Result: 在TheoremQA、LogicBench、ToolQA和CHAMP四个基准测试上，使用近3万个跨领域技能生成了70万个类别多样的伪查询，Skill2Query在稀疏、密集和技能路由检索中平均Recall@1提升了6.70个百分点，其生成的训练数据在评估的生成基线中取得了最佳的Recall@1和nDCG@1，且改进的技能检索转化为更高的智能体任务成功率。

Insight: 创新点在于将技能文档结构化解析为知识图谱，并设计了三阶段的伪查询生成流程，这不仅可用于离线索引增强、在线查询扩展和检索器训练，还通过显式建模技能内部关系提升了生成查询的质量和检索效果。

Abstract: Pseudo-query generation can alleviate the supervision bottleneck for agent skill retrieval, but existing document-level approaches typically leave the rich internal relations among capabilities, parameters, and usage examples implicit. As a result, generated queries may be topically relevant to a skill while lacking capability grounding and parameter consistency, raising the question of whether explicitly exploiting a skill document’s internal structure can produce more effective retrieval signals. We therefore propose Skill2Query, a framework that first parses a skill document into a Skill Knowledge Graph and then generates pseudo-queries through a three-stage process including style mimicking, query template generation, and parameter filling. The generated queries can be used for offline index augmentation, online query expansion, and retriever training. Four benchmarks (TheoremQA, LogicBench, ToolQA, and CHAMP) are used to evaluate Skill2Query with large-scale skill candidate pools across multiple downstream applications, including skill retrieval, retriever training, and end-to-end agent execution. Using nearly 30K skills across diverse domains, we generate 700K category-diverse pseudo-queries. Skill2Query consistently improves sparse, dense, and skill-routing retrieval, with an average Recall@1 gain of 6.70 percentage points across retrieval settings. Skill2Query-generated training data also achieves the best Recall@1 and nDCG@1 among the evaluated generation baselines. Further evaluations with multiple LLM backends demonstrate that improved skill retrieval translates into higher agent task success rates. Code and resources are available at https://github.com/MatZaharia/Skill2Query.

[20] STAIR: Semantic-Temporal Automaton for Interpretable Reasoning in Temporal Question Answering cs.CL | cs.AIPDF

Xinlong Dai, Jinchuan Zhang, Lei Gao, Xinzhe Hu, Yuefeng He

TL;DR: STAIR是一种用于时序问答任务的神经符号系统，它将语义解释与精确时序推理分离：一个免答案的LLM适配器将复杂问题映射为规范化的时序意图，而一个具有有限控制和守卫转换的确定性时序自动机则在规范化证据上执行相应策略。该方法遵循规则优先设计，减少自由形式推理，使时序决策可验证且可解释。

Details

Motivation: 现有基于提示的神经符号系统仍依赖LLM同时进行语义解释和精确时序推理，导致关于区间、时间锚点和有序状态的离散决策易受概率性错误影响且难以验证。

Result: 在TimeQA-Easy、TimeQA-Hard、TempReason-L2和TempReason-L3数据集上的时序问答任务中，STAIR在使用匹配模型设置下持续优于强基线，使用Qwen2.5-7B和GPT-4o-mini模型时分别实现了平均F1分数16.57%和3.10%的提升。

Insight: 核心创新在于将语义解释（由LLM适配器处理）与确定性时序推理（由时序自动机执行）解耦的规则优先架构。守卫执行确保精确的时间点包含和前后关系选择，而语义适应则处理非精确区间和时间锚定查询，从而在保持可解释性的同时提升对边界敏感和顺序敏感查询的处理能力。

Abstract: By leveraging large-scale pretraining, LLMs can interpret diverse temporal expressions and question formulations without task-specific training. However, existing prompt-based neuro-symbolic systems continue to rely on LLMs for both semantic interpretation and exact temporal inference. Consequently, discrete decisions regarding intervals, time anchors, and ordered states remain vulnerable to probabilistic errors and difficult to verify. We present STAIR, a \textbf{S}emantic-\textbf{T}emporal \textbf{A}utomaton for \textbf{I}nterpretable \textbf{R}easoning. STAIR separates semantic interpretation from precise temporal inference: an answer-free LLM adapter maps complex question formulations to normalized temporal intents, while a deterministic temporal automaton with finite control and guarded transitions executes the corresponding policies over canonicalized evidence. Following a rule-first design, STAIR resolves standard questions without invoking an LLM and applies semantic adaptation only when the rule path fails to produce an executable intent. This approach reduces free-form reasoning, making temporal decisions verifiable and interpretable. Specifically, guarded execution supports precise point-time containment and before/after selection, while semantic adaptation handles non-exact intervals and time-anchored queries. Across the TimeQA-Easy, TimeQA-Hard, TempReason-L2, and TempReason-L3 datasets, STAIR consistently outperforms strong baselines in the TQA task using matched model settings, achieving average F1 improvements of 16.57% and 3.10% when utilizing the Qwen2.5-7B and GPT-4o-mini models, respectively. Furthermore, ablations and diagnostic analyses demonstrate that STAIR excels at handling both boundary-sensitive and order-sensitive queries, while its guarded execution and semantic adaptation ensure precise point-time reasoning and inexact intervals, respectively.

[21] Step-Level On-Policy Distillation: Interpolating Between On-Policy Distillation and Supervised Fine-Tuning cs.CL | cs.AIPDF

Changhui Sun, Lanbo Liu, Hang Lei, Tong Ling, Jiahang Xie

TL;DR: 本文提出了一种名为Step-Level On-Policy Distillation (SOPD)的新蒸馏方法，旨在融合监督微调（SFT）的长视野修正能力与在线策略蒸馏（OPD）的策略优势，为模型生成的完整轨迹提供步骤级别的监督。

Details

Motivation: 标准令牌级在线策略蒸馏（OPD）只能对错误的学生模型轨迹提供零碎的修正，无法展开完整且正确的修复路径，因此需要一种能提供更长视野指导的方法。

Result: 在推理和智能体任务（如ALFWorld）上，SOPD显著优于传统的SFT和OPD方法，例如在ALFWorld上将平均成功率比Vanilla OPD提升了13.4个百分点。

Insight: SOPD的创新点在于通过调整步骤长度，可以在SFT和OPD之间平滑插值，既利用了教师模型基于学生轨迹的条件响应以更好地对齐状态，又提供了比OPD更完整的长序列指导，为蒸馏方法研究提供了新视角。

Abstract: On-policy distillation (OPD) aligns a student model with a teacher’s logit distribution on student-generated trajectories. This approach has achieved strong empirical gains and can often surpass conventional off-policy distillation with substantially less data. However, standard token-level OPD can provide only fragmented corrections along an erroneous student trajectory and cannot unfold a complete and correct repair path. Motivated by this limitation, we propose \emph{Step-Level On-Policy Distillation} (SOPD), which combines the long-horizon correction of supervised fine-tuning (SFT) with the on-policy advantage of OPD to provide step-level supervision over complete student-generated trajectories. We show that, at different limits of step length, SOPD reduces to SFT or approximates OPD. Compared with SFT, the teacher responses in SOPD are conditioned on student trajectories and therefore align more closely with student-visited states; compared with OPD, SOPD provides longer-horizon corrections rather than fragmented token-level guidance. Across both reasoning and agent tasks, SOPD substantially outperforms conventional SFT and OPD. For example, on ALFWorld, SOPD improves the average success rate by 13.4 points over Vanilla OPD. We hope this work offers a new perspective for future research on distillation methods.

[22] Mint-Agent: Introducing Finance-Native Agentic Foundation Models cs.CL | cs.LGPDF

Mint-Agent Team, B. Zhang, Yaze Geng, Lei Tang, Yaoyang Yi

TL;DR: 本文提出了Mint-Agent，一个面向金融领域的智能体基础模型家族，旨在构建兼具可靠性和执行力的金融智能体。其核心基于数据、交互框架和算法三大支柱，通过专门的数据引擎、稳定的交互框架以及结合SFT、OPD和RLVR的训练方法，开发了金融推理和智能体执行专家模型，并最终融合为紧凑的通用金融智能体。

Details

Motivation: 解决现有金融智能体不仅需要回忆领域知识，还必须具备可靠地执行基于证据的精确操作以及执行长期、可审计的研究任务的双重挑战。

Result: 在专业金融基准测试中表现出色：Mint-Ag在RFC-Bench上达到98.33%，超越GPT-5.6-Sol和Claude-Opus-4.8；Mint-Cu在FinSearchComp T2上达到69.86%，超越Agents-A1-35B和Nex-N2-mini；Mint-Ag在FinanceAgentBench v1.1和v2上分别达到76.00%和60.49%，确立了其在可靠性和可执行性方面的SOTA水平。

Insight: 创新性地将领域专业知识、长期任务执行和可审计证据轨迹共同工程化为一个统一的基础模型框架，并通过模型合并与多教师策略蒸馏技术将专家能力集成到紧凑模型中，为构建可信赖的金融智能体提供了系统化路径。

Abstract: Financial agents must do more than recall domain knowledge: they must be both reliable, executing precise operations over grounded evidence, and executive, sustaining long-horizon research whose conclusions remain auditable. We present Mint-Agent, a family of finance-native agentic models designed around these two scales of financial intelligence. Mint-Agent is built upon three pillars: data, harness, and algorithm. Our data engine constructs clean, specialized tasks for atomic financial capabilities and long-horizon agentic execution from real-world financial sources. MintHarness enables stable interaction with open-ended environments and maintains auditable evidence trails across extended research trajectories. Our training recipe combines SFT, critical-step OPD, and RLVR to develop separate financial reasoning and agentic execution experts, which are then unified through model merging and multi-teacher on-policy distillation into compact, general-purpose financial agents. This pipeline yields two flagship models, Mint-Cu (9B) and Mint-Ag (27B). Across professional financial benchmarks, our models demonstrate two defining strengths: (1) Reliability: Mint-Ag achieves 98.33% on RFC-Bench, surpassing GPT-5.6-Sol and Claude-Opus-4.8 by 3.66 and 3.00 points; and (2) Executability: Mint-Cu reaches 69.86% on FinSearchComp T2, outperforming Agents-A1-35B and Nex-N2-mini by 22.83 and 12.78 points, while Mint-Ag achieves 76.00% and 60.49% on FinanceAgentBench v1.1 and v2, respectively. These results establish a path toward trustworthy financial intelligence in which domain expertise, long-horizon execution, and auditable evidence are jointly engineered as a unified foundation for frontier agentic models.

[23] D2-ScaleAgent: Dual-Dimensional Scaling for Long Document Understanding cs.CLPDF

Hao Zhang, Longrong Yang, Lunhao Duan, Ziyang Wang, Qing-Guo Chen

TL;DR: 本文提出了D2-ScaleAgent，一个用于视觉丰富长文档理解的多模态检索增强生成（RAG）智能体框架。其核心创新在于引入了基于查询内在难度的双维度（检索与推理）动态扩展范式，通过一个由验证器智能体驱动的动态路由循环和一个持续更新的证据库，实现对外（检索扩展）和对内（推理扩展）的灵活计算资源分配，以解决现有方法工作流固定、计算无法动态扩展导致证据不足的问题。

Details

Motivation: 现有面向视觉丰富长文档理解的多模态RAG方法正向多智能体系统演进，但它们通常依赖固定工作流，缺乏在测试时动态扩展计算的能力，这常常导致检索到的证据不充分。

Result: 在MMLongBench-Doc、LongDocURL等长且视觉丰富的文档基准测试上进行的广泛实验表明，D2-ScaleAgent是有效的。

Insight: 主要创新点是提出了一个双维度（检索与推理）动态扩展范式，通过一个以证据库为动态工作记忆、由验证器智能体驱动的路由循环，实现了根据查询难度自适应地向外扩展检索（属性分解与并行检索）和向内扩展推理（动态选择不同粒度和数量的子智能体），最终在证据链上实现逻辑闭合。这为多模态RAG系统提供了更灵活、更自适应的计算资源分配机制。

Abstract: Multi-modal retrieval-augmented generation (RAG) is a key technique for visually rich long document understanding. Existing multi-modal RAG methods are progressively advancing toward multi-agent systems: they first retrieve relevant pages based on a query, and then iteratively understand information within those pages. However, these methods typically rely on fixed workflows and lack the ability to dynamically scale computation at test time, often leading to insufficient evidence. To address this, we propose D2-ScaleAgent, an agentic framework that introduces a dual-dimensional scaling paradigm for retrieval and reasoning. The core of D2-ScaleAgent is a Verifier agent-driven dynamic routing loop based on the intrinsic difficulty of the query, centered around a continuously updated evidence bank that serves as the agent’s dynamic working memory: when retrieval needs to be expanded, the agent routes outward (retrieval scaling), decomposing the query into attributes and performing parallel page retrieval, followed by adaptive pruning to ensure comprehensive evidence coverage. When fine-grained reasoning is required, the agent routes inward (reasoning scaling), dynamically selecting sub-agents with varying granularity and count to extract evidence from pages. Finally, D2-ScaleAgent achieves logical closure over the evidence chain. Extensive experiments demonstrate that D2-ScaleAgent is effective on long and visually rich document benchmarks like MMLongBench-Doc, LongDocURL, etc.

[24] BabelSteering: Multilingual Safety Alignment via English Steering Vectors cs.CLPDF

Emma V. Stein, Dominik Meier, Terry Ruas, Jan Philip Wahle, Bela Gipp

TL;DR: 本文提出BabelSteering方法，通过从英语安全监督中提取的拒绝方向向量，作为轻量级推理时干预手段，以提升多语言大模型的安全性。该方法在八种语言上评估，能有效增加对有害请求的拒绝率，同时保持任务效用基本不变，但会略微增加对伪有害提示的过度拒绝。

Details

Motivation: 当前大语言模型的安全研究和对齐工作主要集中在英语，导致其他语言用户面临较弱的安全保障，因此研究如何将英语学到的安全信号泛化到多语言场景。

Result: 在Gemma 7B等模型上，BabelSteering使多语言有害提示拒绝率平均提升11个百分点（如孟加拉语提升17个百分点），在Global MMLU基准上未损失效用，但伪有害拒绝率平均增加13个百分点。

Insight: 创新点在于利用英语安全监督衍生的激活向量进行跨语言安全干预，提供了一种低成本、实用的多语言安全对齐机制；同时构建了多语言翻译与评估流程以支持后续研究。

Abstract: Large language models (LLMs) are deployed globally in high-stakes settings, yet most safety research and alignment efforts remain concentrated on English. Thus, users interacting with LLMs in other languages may encounter weaker safeguards despite relying on the same systems for similarly sensitive tasks. In this work, we investigate whether safety signals learned from a high-resource language, like English, can improve multilingual safety. We propose BabelSteering, an activation steering method that acts as a lightweight inference- time intervention, using refusal directions derived from English safety supervision to generalize across languages. Our evaluation includes eight languages and jointly measures refusal of harmful requests, over-refusal, and general task utility. The results show that BabelSteering increases the refusal of harmful requests across languages, with only a marginal to no reduction in task utility but with some increase in refusal of pseudo-harmful prompts. For example, for Gemma 7B, we see an average increase in the refusal of harmful prompts across languages of 11 percentage points (pp), with individual languages like Bengali seeing an increase of 17 pp, with no loss of utility on Global MMLU, while pseudo-harmful refusals increase by 13 pp on average. We also introduce a multilingual translation-and-evaluation pipeline to facilitate future work on cross-lingual safety interventions. Overall, our findings suggest that activation steering may provide a practical, low- cost mechanism for extending English-derived safety signals to other languages. Warning: this paper contains examples with unsafe content

[25] Ask, Condition or Abstain: Reinforcement Learning for Missing-Premise Reasoning cs.CLPDF

Yongqi Tong, Zhenyu Zhang, Zimi Liu, Kewei Fu, Mingli Song

TL;DR: 本文提出了ACA-RL（Ask-Condition-Abstain Reinforcement Learning）框架，用于训练推理模型处理前提缺失的查询。该框架通过数据增强和强化学习，指导模型在面对不完整问题时，选择询问缺失前提、基于未知量给出条件性答案或放弃回答。作者还引入了MPB（Missing-Premise Benchmark）基准来评估模型在此类任务上的表现。

Details

Motivation: 现实中的许多查询往往缺少得出唯一答案所需的前提，而现有的仅基于答案的强化学习方法主要针对完全明确的问题进行训练，无法有效处理这种不确定性。

Result: 在Qwen3和Llama模型上，ACA-RL在MPB基准上取得了持续改进，同时在定义良好的推理任务上保持了有竞争力的性能。

Insight: 创新点在于提出了一个结构化奖励框架，将模型响应行为分为五类（如询问、条件化、放弃等），并通过推理图引导的数据增强生成训练实例。这为NLP评估引入了新方向：不仅评估模型回答明确问题的能力，更评估其识别任务不确定性并妥善处理的能力。

Abstract: Answer-only reinforcement learning (RL) trains reasoning models to solve fully specified problems, but many realistic queries omit a premise needed for a unique answer. In this setting, the useful response is not always refusal: the model should ask for the missing premise, condition its answer on the unknown quantity, or abstain when no informative conditional response is available. We present \emph{Ask-Condition-Abstain Reinforcement Learning} (ACA-RL), a data-augmented RL framework for this setting. Its reasoning-graph-guided pipeline converts well-posed problems into missing-premise training instances with localized gap annotations; ACA-RL then trains on these instances with a structured reward over five observable response behaviors. We also introduce the \emph{Missing-Premise Benchmark} (MPB), a 274-instance human-verified benchmark spanning mathematical, logical, and real-world word problems. Across Qwen3 and Llama models, ACA-RL consistently improves on MPB while preserving competitive performance on well-posed reasoning tasks. Together with the released code, MPB, and training data, this work supports a new mission for NLP evaluation: measuring whether models can recognize when a task is underdetermined and handle uncertainty, not only whether they can answer fully specified questions.

[26] When Do Explanations Help In-Context Learning? A Comparative Study of Natural Language Explanation Types and Faithfulness cs.CL | cs.AIPDF

Mahdi Dhaini, Adam Dejl, Juraj Vladika, Volkan Özer, Barbara Plank

TL;DR: 本文系统比较了在上下文学习中，不同来源（人工撰写、模型自生成、外部LLM生成）和不同选择策略（随机筛选与基于忠实度筛选）的自然语言解释（NLEs）对下游任务性能的影响。研究发现，在分类任务中，添加NLEs通常能提升少样本提示的准确率，其中外部LLM生成的解释表现优异，甚至可与人工解释媲美；而在数学推理任务中，效果则更依赖于模型和解释来源。

Details

Motivation: 当前在上下文学习中，自然语言解释被广泛用作输入以影响模型行为，但不同来源和筛选方式的NLEs对下游性能的影响尚不明确，因此需要系统评估以指导实际应用。

Result: 在六个基准测试和四个指令调优模型上的评估表明：在分类任务中，添加NLEs通常能提高准确率；外部LLM生成的NLEs表现强劲，与人工解释相当；自生成NLEs对选择策略更敏感；基于忠实度的筛选整体提升有限，且效果因任务和模型而异。

Insight: 创新点在于首次系统比较了不同NLE来源和选择策略在上下文学习中的效用，并揭示了外部LLM生成解释的竞争力以及忠实度筛选的局限性，为实际提示工程中解释的选择和报告提供了实证依据。

Abstract: Natural language explanations (NLEs) are increasingly used as inputs, for example, as few-shot rationales that influence model behavior in in-context learning (ICL). However, it remains unclear how different types of NLEs compare in their effects on downstream model performance in explanation-augmented prompting. Therefore, we provide a comparative evaluation across six benchmarks and four instruction-tuned models, studying how NLE source (human-written when available, self-generated explanations, generated by an external LLM) and NLE selection (random vs faithfulness-based filtering) affect downstream utility of NLEs when used in ICL settings. Our extensive evaluation shows that, on classification-style benchmarks, adding NLEs to few-shot prompts often improves accuracy over few-shot prompting without explanations; among NLE sources, externally generated LLM-NLEs often provide strong downstream utility and remain competitive with human rationales where both are available, whereas self-NLEs are more sensitive to the selection strategy. On math reasoning, the effects are more model- and source-dependent. We further show that faithfulness-based selection of self-NLEs yields small average gains overall, but can improve or reduce performance depending on the metric, task, and model. Different faithfulness metrics can disagree substantially, affecting which self-NLE examples are selected and their downstream predictive utility. Robustness tests with randomly swapped and out-of-distribution rationales indicate partial robustness, suggesting that semantic alignment contributes to performance gains. Overall, our results provide insights for selecting and reporting explanations that influence model behavior in practical prompting pipelines.

[27] Every Coin Has Two Sides: On the Dual Nature of Generalization in On-Policy Distillation of Large Language Models cs.CLPDF

Zhaoyi Li, Deyang Kong, Yuan Wei, Evan Yang, Ranran Shen

TL;DR: 本文研究了策略蒸馏（OPD）在大型语言模型中的泛化行为，通过控制变量实验探讨了从域内分布偏移到跨域迁移及多教师设置下的泛化特性。研究发现，OPD主要传递教师的推理行为而非具体答案，且训练难度影响不大；泛化效果强烈依赖于师生模型的起源关系，同源对能实现跨语言、推理范围甚至跨域的迁移，而异源对则主要拟合训练分布。

Details

Motivation: 策略蒸馏通过监督学生自身策略采样的轨迹来传递教师能力，但其泛化行为尚不明确，现有研究多在单一领域或接近训练数据的基准上评估，缺乏系统性分析。

Result: 实验表明，同源师生对在跨语言、推理范围和其他领域上能使学生接近教师水平，而异源对泛化有限；多教师设置中，结合不同教师会导致能力间的混合依赖跷跷板效应，影响泛化。

Insight: 创新点在于揭示了OPD传递的是推理行为而非答案，并强调了起源关系对泛化的关键影响；客观分析认为，这为诊断多教师OPD提供了新视角，有助于理解泛化的双刃剑特性，即广泛迁移能力可能带来能力混合的不稳定性。

Abstract: On-policy distillation (OPD) transfers teacher capabilities by supervising trajectories sampled from the student’s own policy, yet its generalization behavior remains poorly understood, as most studies evaluate OPD on a single domain and on benchmarks close to the training data. We present a controlled study that varies one generalization factor at a time, from in-domain distribution shifts to cross-domain transfer and the multi-teacher setting. We find that OPD transfers a teacher’s reasoning behavior rather than its answers to particular problems: training difficulty barely matters, and even problems the teacher never solves are useful. Transfer depends strongly on the origin relationship between teacher and student: same-origin pairs bring the student close to the teacher across languages, reasoning horizons, and even other domains, whereas cross-origin pairs mostly fit the trained distribution. This broad reach is a double-edged sword: since routing prompts to domain experts cannot confine each teacher’s influence, combining them yields a mixture-dependent seesaw among their capabilities. These results clarify when OPD generalizes and offer a useful perspective for diagnosing multi-teacher OPD.

[28] ClawGym II: Exploring Black-Box RL on Agent Harness cs.CL | cs.AI | cs.LGPDF

Huatong Song, Fei Bai, Ming Yang, Renyuan Li, Jia Deng

TL;DR: 本文提出了ClawGym II，一个用于通过复杂代理工具链进行黑盒强化学习的统一框架。该框架通过沙盒隔离执行、代理调用捕获与树结构重组、以及混合工具链训练等技术，实现了对通用代理在长视野任务中的稳定、可扩展优化。

Details

Motivation: 代理工具链通过协调代理与环境的交互，显著提升了长视野任务的性能，但通过复杂工具链进行强化学习仍未被充分探索，因为将其扩展到长视野代理任务存在根本性挑战。

Result: 使用Qwen3-30A3B模型，黑盒RL在ClawGym-Bench基准上通过OpenClaw和Claude Code工具链分别将Pass@1提升了9.98和14.81个百分点，并在200-400个优化步骤中保持稳定；在JobBench和OfficeQA等更具挑战性的任务上也取得了一致的性能提升。

Insight: 创新点包括：1) 沙盒化执行基础设施实现大规模并发；2) 通过服务代理捕获模型调用并重组为前缀树以优化轨迹；3) 混合工具链训练使单一模型能被异构工具链联合优化；4) 整个框架实现了对黑盒工具链的有效、稳定、可扩展优化，并支持跨异构执行系统的统一训练。

Abstract: Agent harnesses have substantially improved performance on long-horizon tasks by coordinating agent interactions with the environment. However, reinforcement learning through complex harnesses remains largely unexplored, as scaling such training to long-horizon agent tasks introduces fundamental challenges. In this work, we present a unified black-box RL framework for stable and scalable optimization of general agents through complex harnesses. Concretely, we first build a sandbox-based execution infrastructure that isolates task environments and harnesses within temporary sandboxes for large-scale concurrent rollouts. We then decouple policy optimization from opaque harness execution and place a serving proxy at the model boundary to capture model calls. To reconstruct multi-turn trajectories and improve training efficiency, we organize the captured calls into prefix trees and further adapt both critic-based PPO and critic-free GRPO to optimize over the recovered tree structure. Meanwhile, we maintain training-inference consistency throughout the optimization process. Finally, we introduce mix-harness training, allowing a single model to be jointly optimized by heterogeneous harnesses. With Qwen3-30A3B, black-box RL improves Pass@1 on ClawGym-Bench by 9.98 and 14.81 points through OpenClaw and Claude Code, respectively, while remaining stable over 200-400 optimization steps. Moreover, the framework yields consistent gains on more challenging tasks such as JobBench and OfficeQA. Overall, our framework enables effective, stable, and scalable optimization of general agents through black-box harnesses, supporting unified training across heterogeneous execution systems.

[29] PCA-guided Activation Scaling for Monotonic Bidirectional Control over LLM Sycophancy cs.CLPDF

Zheng Chen, Zhaoxin Feng, Yip Tin Po, Jianfei Ma, Emmanuele Chersoni

TL;DR: 本文提出了一种名为PCA引导的激活缩放（PAS）的激活引导框架，用于对大型语言模型（LLM）中的谄媚行为（即倾向于同意用户观点而忽视事实准确性）进行单调双向控制。该方法通过主成分分析（PCA）识别出与谄媚-诚实行为相关的子空间，并对该子空间和正交残差应用不同的缩放指数，从而实现对谄媚行为增强或减弱的可预测、渐进的调控。

Details

Motivation: 大型语言模型存在谄媚行为，即倾向于迎合用户观点而忽视事实，这可能强化错误观念，但完全消除该行为又可能过度纠正合理观点。因此，需要一种既能减少也能增强谄媚行为，且具有可预测、渐进效果的控制方法。现有方法无法确保在跨模型和数据集时，引导强度与行为结果之间存在单调、双向的关系。

Result: 在三个LLM和三个数据集上的实验表明，PAS方法实现了强单调性（斯皮尔曼相关系数ρ = +0.92），并且每个方向（增强或减弱谄媚）平均能带来15.4%的行为改变，而基线方法的平均改变仅为8.7%。消融研究证实了分解、非对称指数和层选择对于维持单调控制都是必要的。

Insight: 论文的创新点在于利用PCA分解残差流激活，将行为引导问题解耦到特定的语义子空间（谄媚-诚实子空间）和正交残差，并通过对这两个部分应用不同的缩放指数来实现单调、双向的精细控制。这为模型行为调控提供了一种可解释、可预测且效果显著的子空间干预方法。

Abstract: Large language models (LLMs) exhibit sycophancy, a tendency to agree with user beliefs regardless of factual accuracy. This can reinforce misconceptions, but eliminating it entirely risks over-correction against valid opinions. Effective control must therefore both reduce and increase sycophancy with predictable and gradual effect. Yet, existing methods fail to ensure a bidirectional and monotonic relationship between steering strength and behavioral outcome across models and datasets. We introduce PCA-guided Activation Scaling (PAS), an activation steering framework that decomposes residual stream activations into a PCA-identified sycophancy-honesty subspace and an orthogonal residual, then applies distinct scaling exponents to achieve monotonic, bidirectional control. Across three LLMs and three datasets, PAS achieves strong monotonicity (Spearman $ρ$ = +0.92) and an average shift of 15.4% per direction, compared with 8.7% for the baselines. Ablation studies confirm that the decomposition, asymmetric exponents, and layer selection are each essential for maintaining monotonic control. The data and code are available at https://github.com/Bellafc/PCS.

[30] Model Hypnosis: Strong control of AI via additive subliminal effects cs.CL | cs.AIPDF

Enric Boix-Adsera, Benedict Tessler

TL;DR: 本文揭示了AI模型普遍存在一种称为’模型催眠’的现象，即通过组合提示中看似无关的弱线索，可以系统性地强烈控制模型行为。该现象跨越不同模型家族和规模，包括前沿推理模型，且催眠提示可在模型间迁移。由于模型被文本细节（如改写和拼写错误）控制，这对AI安全性和可解释性提出了新挑战。

Details

Motivation: 研究动机是探索AI模型对提示中细微文本变化的敏感性，揭示模型行为可能被看似无关的线索组合所强烈控制的潜在风险。

Result: 实验表明模型催眠现象广泛存在于不同模型家族和规模中，包括前沿推理模型，且催眠提示具有跨模型迁移性，但未提及具体基准测试或定量结果。

Insight: 创新点在于首次系统性地定义了’模型催眠’现象，揭示了AI模型对文本细微变化的脆弱性，为AI安全性和可解释性研究提供了新的重要方向。

Abstract: We demonstrate that AI models are broadly susceptible to a phenomenon we call model hypnosis, in which individually weak and seemingly irrelevant cues in the prompt can be systematically combined to strongly control model behavior. Model hypnosis occurs across model families and scales, including in frontier reasoning models, and hypnotic prompts can transfer between models. Because the model is controlled by inconspicuous textual choices, such as paraphrases and typos, model hypnosis presents new challenges and avenues for AI safety, and is a major hurdle for AI interpretability.

cs.CV [Back]

[31] Deep Analog: Open-Set Film Emulation with Reference-Conditioned 3D LUTs cs.CV | cs.GR | cs.LG | cs.MMPDF

Yitong Mu

TL;DR: 本文提出了Deep Analog，一个用于开集胶片模拟的流程，旨在从单个参考帧中复制任何模拟胶片的视觉效果。核心是预测一个参考图像条件化的3D查找表（LUT），并围绕其构建了包含色调匹配和物理光学渲染器的完整处理流程。

Details

Motivation: 解决开集胶片模拟问题，即从单个参考示例中匹配任意胶片外观。现有基于固定LUT库的实时图像增强方法存在专家混合模型的门控崩溃和闭集限制问题。

Result: 在350个自监督生成的图像对上，颜色处理阶段达到22.05 dB PSNR / 0.925 SSIM，完整流程达到21.72 dB / 0.923。颜色处理路径在1080p分辨率下运行时间为5.2 ms（192 FPS），并可导出标准.cube LUT文件。

Insight: 提出了参考图像条件化的3D LUT预测网络（StyleLUTNet），无需配对数据或重新训练即可泛化到未见过的胶片类型，解决了闭集限制。揭示了条件LUT训练中的残差尺度崩溃问题，并提出辅助正则化必须服从于重建任务的一般原则。

Abstract: Film emulation reproduces the look of an analog film stock on a new digital photograph. We target its open-set form – matching any reference film frame from a single example – with a 3D lookup table (LUT) predicted from that reference. Real-time image enhancement predicts per-image weights over a fixed bank of 3D LUTs and blends them. We show this is a gated mixture of experts and inherits its failure: trained end-to-end against reconstruction, the gate collapses onto a single expert, so a bank of K LUTs delivers the capacity of one. An entropy term, the enhancement-setting analogue of mixture-of-experts load balancing, restores utilization and recovers about 1 dB PSNR. The deeper constraint survives: a fixed LUT basis is closed-set, freezing the achievable looks at training time. We therefore discard the basis and predict a single 3D LUT as a residual from a reference image (StyleLUTNet), trained by self-supervision on procedurally generated color transforms. The conditional design removes the gate and generalizes open-set to unseen film stocks without paired data or retraining. Around this color backbone we build Deep Analog, a film-emulation pipeline that adds histogram-based tone matching and a physics-informed optical renderer – multi-scale grain and per-channel halation driven by parameters an inverse network regresses from the reference. On 350 self-supervised pairs the color stage reaches 22.05 dB PSNR / 0.925 SSIM and the full pipeline 21.72 dB / 0.923; the color path runs in 5.2 ms at 1080p (192 FPS) and exports a portable .cube LUT for standard editing tools. A second degeneracy in conditional LUT training – residual-scale collapse – shares the root cause and yields a general principle: auxiliary regularization must stay subordinate to reconstruction.

[32] Xemo-Talker: Unlock Emotions Explicitly for Audio-Driven Talking Portrait Synthesis cs.CV | cs.SDPDF

Chaolong Yang, Yinuo Guo, Kai Yao, Yuyao Yan, Jie Sun

TL;DR: 本文提出了Xemo-Talker，一种用于音频驱动说话头像合成的模型，旨在解决现有方法在情感控制上依赖隐式调节、导致控制不直接且不充分的问题。该方法的核心发现是，情感线索主要分布在运动空间的非主成分中，因此通过在这些非主成分上施加判别性监督，可以更好地平衡情感控制与唇部同步。模型首先学习一个中性的语音到运动映射以保证稳定的发音和唇部同步，然后引入一个轻量级的情感分支，该分支受到非主子空间监督的引导，并辅以Tri-Loss（类间分离、类内紧凑性和非主成分对比学习）来增强情感控制。

Details

Motivation: 现有音频驱动说话头像合成系统依赖隐式情感调节，导致对情感的控制间接且不充分。同时，在整个运动空间使用显式情感相关损失进行训练存在困难，因为需要权衡准确的唇部同步和细粒度的情感控制。

Result: 在给定音频输入、参考图像和情感标签的条件下，Xemo-Talker在情感分类准确率上达到了最先进水平（SOTA），同时保持了有竞争力的唇部同步性能和高推理效率，其性能接近在真实视频上测量的结果。

Insight: 论文的创新点在于揭示了情感线索主要编码在运动空间的非主成分中（主成分主要编码高能量的发音和姿态变化），并基于此提出了在非主子空间进行集中监督的策略。从客观角度看，将运动解耦为中性映射和轻量级情感分支，并设计Tri-Loss进行联合优化，是实现高效、显式情感控制的一个有效且新颖的架构设计思路。

Abstract: Precise emotion control in audio-driven talking heads remains a challenge due to the reliance on implicit emotion regulation in existing systems, which often leads to indirect and insufficient control. Additionally, training with explicit emotion-related losses across the entire motion space poses significant difficulties due to the inherent trade-off between accurate lip synchronization and fine-grained emotion control. In this paper, we reveal a key finding: although emotional cues are distributed throughout the motion space, concentrating discriminative supervision on less-principal components achieves a better emotion-lip synchronization balance, as principal components mainly encode high-energy articulation and pose variations. Building on this insight, we propose Xemo-Talker, which first learns a neutral speech-to-motion mapping for stable articulation and lip synchronization, and then introduces a lightweight emotion branch guided by less-principal subspace supervision. To enhance emotion control, we design a Tri-Loss consisting of inter-class separation, intra-class compactness, and less-principal contrastive learning. Given an audio input, a reference image, and an emotion label, Xemo-Talker achieves state-of-the-art emotion classification accuracy while maintaining competitive lip synchronization and high inference efficiency, with performance approaching that measured on real videos.The source code is publicly available at https://github.com/chaolongy/Xemo-Talker.

[33] On Cross-Validation for Hyperparameter Optimization of Deep Learning Image Classifiers cs.CV | cs.LGPDF

Ljubomir Buturovic

TL;DR: 本文系统比较了三种超参数优化（HPO）协议——固定保留集、重排保留集和5折交叉验证——在深度学习图像分类器中的性能估计误差。研究发现，在小样本医学影像数据集上，交叉验证能显著降低绝对性能估计误差（AEE），尤其适用于样本量有限的情况；而在大规模自然图像数据集上，三种协议的AEE差异可忽略。

Details

Motivation: 超参数优化对深度学习图像分类器性能至关重要，但缺乏关于如何选择验证策略（特别是小样本场景如医学影像）的实证指导。本文旨在比较不同HPO协议在性能估计可靠性上的差异。

Result: 在RSNA肺炎X光片和二值化HAM10000皮肤病变数据集上，交叉验证在所有样本量下均优于两种保留集协议，且AEE降低幅度在小样本时最大；在Tiny ImageNet（200类）上，三种协议的AEE均接近零。测试AUROC在不同协议间总体相似。

Insight: 研究证实了交叉验证在小样本医学图像分类任务中能通过额外计算成本换取更可靠的测试性能预估，为资源允许的场景提供了明确的HPO协议选择依据；同时揭示了样本规模对验证策略有效性的关键影响。

Abstract: Hyperparameter optimization (HPO) can materially affect the performance of deep learning (DL) image classifiers, but there is little empirical guidance on how to derive the validation signal that drives it, especially for the small sample sizes common in fields such as medical imaging. We compared three HPO protocols in terms of {\em absolute performance-estimation error} (AEE; the absolute difference between the winning configuration’s validation AUROC and its test AUROC): fixed holdout (F), reshuffled holdout (R), and 5-fold cross-validation (C). The search space, sampler, training procedure, architecture, and test set were held identical across protocols. We evaluated the protocols on three public datasets spanning two regimes: binary medical imaging (RSNA pneumonia radiographs and binarized HAM10000 skin lesions) and 200-class natural imaging (Tiny ImageNet), across a range of development set sizes $n$ and two backbones (ResNet-18 on all datasets, Vision Transformer (ViT-S/16) on RSNA). On the medical datasets, every point estimate favored cross-validation over both holdout protocols, with reductions in AEE largest at small sample sizes and diminishing as $n$ increased. This pattern remained robust under conservative family-wise adjustment. On Tiny ImageNet, AEE was negligible under all three protocols. Test AUROC was generally similar among protocols. Fixed holdout had lower mean AEE than reshuffled holdout in 11 of 12 medical conditions, although this secondary finding was less uniformly supported. For small-sample medical image classification, we recommend cross-validation-based HPO when computational resources permit because it trades additional computation for a more reliable development-time estimate of subsequent test performance.

[34] Equilibrium Forcing: Adaptive Video Generation Without Noise Conditioning cs.CV | cs.AI | cs.LGPDF

Hansen Jin Lillemark, Alex Rojas, Zachary Novack, Runqian Wang, Yilun Du

TL;DR: 本文提出了一种名为Equilibrium Forcing (EqF) 的新框架，用于无需噪声级别条件化的视频去噪生成模型。该框架将去噪场的学习与采样过程解耦，允许推理时算法通过适应样本反馈进行闭环操作，从而在具有挑战性的自回归视频生成基准上提升了视频质量和一致性。

Details

Motivation: 标准的基于扩散和流匹配的自回归视频生成算法依赖于固定的训练目标和静态采样计划，限制了推理过程对数据的适应性。本文旨在解决这一问题，通过移除噪声级别条件化，使模型能够更灵活地适应数据。

Result: 在具有挑战性的自回归视频生成基准测试中，EqF 通过其数据依赖的推理特性，超越了标准的噪声级别条件化去噪视频方法的性能，提升了视频质量和一致性。

Insight: 主要创新点在于提出了无需噪声级别条件化的模块化训练和推理设计，将去噪场学习与采样解耦，实现了推理时的闭环自适应。这为生成模型提供了更大的灵活性，使其能够根据样本反馈动态调整，从而可能提高生成质量。从客观角度看，这种解耦和自适应机制是推动视频生成技术发展的一个值得借鉴的方向。

Abstract: Standard autoregressive video generation algorithms based on Diffusion and Flow Matching rely on rigid training objectives and static sampling schedules, limiting inference procedures from adapting to the data. We introduce Equilibrium Forcing (EqF), a simplified framework for video denoising generative models without noise level conditioning. EqF pioneers modular training- and inference-time designs for noise-unconditional generation that decouple learning the denoising field from sampling. This flexibility allows for inference-time algorithms that operate in a closed loop by adapting to feedback from the sample, improving video quality and consistency on challenging autoregressive video generation benchmarks. Extensive analysis elucidates exactly how removing the noise level conditioning enables EqF’s data-dependent inference properties to surpass the performance of standard noise level-conditional denoising video methods.

[35] VideoGAIA: A Benchmark for General AI Assistants on Agentic Video Understanding cs.CV | cs.CLPDF

Fan Zhang, Guangming Yao, Jinyang Wu, Hao Wu, Zheng Lian

TL;DR: 本文介绍了VideoGAIA，一个用于评估通用人工智能助手在代理式视频理解能力上的新基准。它超越了传统的单轮视频问答，将视频理解构建为一个多轮、工具增强的交互过程，要求模型迭代感知视频、调用外部工具、收集补充信息并整合多轮多模态证据。该基准包含271个人机协同设计的任务，覆盖多样复杂的现实场景，并由三位人类专家独立验证，旨在推动从传统视频理解向代理式视频理解的转变。

Details

Motivation: 现有领先模型在传统视频理解基准（如Video-MME）上准确率已接近90%，表明单轮视频理解任务趋于饱和，不足以评估先进多模态大语言模型的智能水平。因此，需要一个新的、更具挑战性的基准来评估下一代模型的代理式视频理解能力。

Result: 在VideoGAIA基准上，所有被评估的多模态大语言模型（包括前沿模型如GPT-5.5和Kimi-K3）的准确率均低于60%，凸显了该基准的高质量和及时性，能有效区分当前模型的性能极限。

Insight: 论文的核心创新在于将视频理解从单轮问答范式转变为多轮、工具增强的代理式交互过程，这更贴近现实世界中复杂、迭代的推理需求。其构建的基准通过人机协同设计任务、专家独立验证确保了任务的多样性和难度，为评估和推动下一代AI助手的关键能力（如工具使用、多轮信息整合）提供了重要方向。

Abstract: Video understanding is a fundamental task for evaluating the capabilities of multimodal large language models (MLLMs). However, existing leading models have already achieved approximately 90% accuracy on the Video-MME leaderboard, suggesting that conventional single-turn video understanding tasks are becoming increasingly saturated and insufficient for assessing the intelligence of advanced MLLMs. Towards this end, we introduce VideoGAIA, an agentic video understanding benchmark for general artificial intelligence (AI) assistants. Moving beyond one-shot video question answering, VideoGAIA formulates video understanding as a multi-turn, tool-augmented interaction process, where models must iteratively perceive videos, invoke external tools, gather complementary information, and integrate multimodal evidence across turns. VideoGAIA contains 271 model-human co-designed tasks covering diverse and complex real-world scenarios. Each video-question-answer instance is independently verified by three human experts to ensure both correctness and appropriate difficulty. All evaluated MLLMs, including frontier models such as GPT-5.5 and Kimi-K3, achieve less than 60% accuracy on VideoGAIA, highlighting its value as a high-quality and timely benchmark for evaluating next-generation MLLMs. We hope that VideoGAIA will facilitate the transition from conventional video understanding toward agentic video understanding.

[36] DeCo-MIL: Debiased Counterfactual Reasoning for Long-Tailed Whole Slide Image Analysis cs.CV | cs.AIPDF

Xiaoxiao Li, Xitong Ling, Jiawen Li, Weiming Chen, Zhenyang Cai

TL;DR: 本文提出DeCo-MIL方法，用于解决长尾分布下全切片图像（WSI）弱监督分析中的嵌套双重长尾问题。该方法通过频率去偏的反事实推理，联合缓解了切片间的类别长尾和切片内具有判别性证据的实例长尾，从而提升尾部类别的识别能力。

Details

Motivation: 动机在于解决基于多示例学习（MIL）的WSI分析在长尾分布下面临的嵌套双重长尾问题：切片间类别分布不均（头部类别样本多）与切片内判别性证据稀缺且被冗余信息掩盖，这两个问题相互耦合，导致模型偏向头部类别，损害尾部类别识别。

Result: 在三个长尾WSI基准测试上的大量实验表明，DeCo-MIL在尾部类别识别和整体分类上都达到了最先进的（SOTA）性能。

Insight: 创新点在于提出了一种频率去偏的反事实推理框架，通过聚类组织形态锚点、进行反事实干预并估计其贡献来指导冗余掩蔽，同时结合锚点分层伪包构建、尾部感知过采样和一致性正则化，以联合缓解嵌套的双重长尾问题，有效保留稀缺的判别性实例并增加尾部类别的监督信号。

Abstract: Multiple instance learning (MIL) is widely used for weakly supervised whole slide image (WSI) analysis. However, under long-tailed distributions, MIL-based WSI analysis faces a nested dual long-tail: an inter-slide class long tail and an intra-slide long tail of instance-level discriminative evidence. The two long tails are coupled: tail classes have few training slides, while their limited diagnostic evidence is concentrated in a few patches and obscured by abundant within-bag redundancy. This coupling biases models toward head classes and degrades rare-class recognition. To address this, we propose DeCo-MIL for long-tailed WSI analysis, which jointly alleviates the nested dual long-tail through frequency-debiased counterfactual reasoning. For the inner long tail, DeCo-MIL clusters patches into tissue-morphology anchors, replaces each anchor with its matched normal prototype to perform a counterfactual intervention, and estimates its counterfactual contribution to the ground-truth class using class-frequency-corrected predictions. These contributions guide redundancy masking to preserve scarce discriminative instances. For the outer long tail, DeCo-MIL constructs anchor-stratified pseudo-bags from redundancy-reduced bags and combines tail-aware oversampling with consistency regularization, increasing effective supervision for tail classes while preserving tissue-morphology composition. Extensive experiments on three long-tailed WSI benchmarks demonstrate that DeCo-MIL achieves state-of-the-art performance in both tail-class recognition and overall classification.

[37] AeroGround: A Comprehensive Benchmark for Aerial-Ground Collaborative Reasoning cs.CVPDF

Shenghong Yi, Lin Zhang, Muzian Li, Jiakang Yuan, Haoyu Zhang

TL;DR: 本文提出了AeroGround基准测试，用于评估视觉语言模型在无人机与地面协作场景中的理解和推理能力。该基准基于包含约29,000组多模态观测数据的模拟数据集，提供了2,250个高质量问答实例，涵盖跨视图对应、空间理解和推理任务。实验表明，现有模型与人类性能存在显著差距，最佳模型平均准确率仅为54.4%，而人类达到93.3%。

Details

Motivation: 现有无人机基准主要关注空中视角场景，但视觉语言模型在实际应用（如救援和基础设施检查）中空中-地面协作场景下的理解和推理能力尚未得到充分探索。

Result: 在16个预训练视觉语言模型及两个领域适应变体上的实验显示，最佳模型平均准确率为54.4%，远低于人类的93.3%，揭示了当前模型在AeroGround基准上的性能局限。

Insight: 创新点在于构建了首个专注于空中-地面协作推理的综合基准，通过模拟数据集和多样化任务系统评估模型能力，为开发更强大的空中-地面协作具身智能系统奠定了基础。

Abstract: Vision-language models (VLMs) have been widely employed in understanding and reasoning tasks for unmanned aerial vehicles (UAVs). Existing UAV benchmarks primarily focus on aerial-view scenarios. However, whether current VLMs can perform well on understanding and reasoning tasks in aerial-ground collaborative scenarios which are practical in real-world applications like rescue and infrastructure inspection remains underexplored. To address this gap, we introduce AeroGround, a comprehensive benchmark for evaluating VLMs in aerial-ground collaborative reasoning. AeroGround is built upon a simulated aerial-ground dataset containing approximately 29,000 multimodal observation groups from diverse open environments, and provides 2,250 high-quality question-answering instances covering cross-view correspondence, spatial understanding, and reasoning. Experiments on 16 pretrained VLMs, together with two domain-adapted variants, reveal a substantial gap between current models and human performance: the best model achieves an average accuracy of 54.4%, whereas humans reach 93.3%. By systematically revealing the strengths and limitations of existing models in aerial-ground collaborative reasoning, AeroGround provides a foundation for developing more capable aerial-ground collaborative embodied intelligence systems.

[38] Braided Vision Transformer for Stroke Detection in Multi-view Retinal Fundus Imaging cs.CVPDF

Aysen Degerli, Mika Hilvo

TL;DR: 本文提出了一种名为编织视觉变换器（BViT）的模型，用于通过多视角视网膜眼底成像检测中风和短暂性脑缺血发作。该模型首次将视觉变换器应用于中风评估的视网膜成像领域，能够从双眼的黄斑中心和视神经头中心视图中提取代表性特征，并捕获视图间的相互关系。

Details

Motivation: 中风是全球主要的死亡和致残原因，需要准确且即时的评估。视网膜眼底成像作为一种非侵入性、成本效益高且便携的筛查方式，能够反映脑血管和神经系统的风险因素，为中风评估提供了有前景的替代方案。

Result: 在作者收集的Stroke-Data数据集上的实验表明，BViT模型在中风检测任务上取得了0.75的AUC分数，其性能优于常规的视觉变换器模型。

Insight: 论文的主要创新点在于首次将视觉变换器架构引入视网膜眼底成像的中风评估任务，并提出了BViT模型。该模型的创新之处在于其能够同时处理多视角图像并捕获不同视图（双眼的黄斑和视神经头视图）之间的相互关系，从而更全面地理解与脑血管事件相关的视网膜生物标志物。

Abstract: Stroke remains a leading cause of mortality and morbidity worldwide, emphasizing the importance of its accurate and immediate assessment. Retinal fundus imaging has emerged as a promising modality for stroke assessment, as the retina reflects cerebrovascular and neurological risk factors. Contrary to conventional neuroimaging techniques, retinal fundus imaging offers a non-invasive, cost-effective, and portable alternative for rapid screening. This paper explores the feasibility of retinal fundus imaging for stroke and transient ischemic attack (TIA) detection using macula-centric and optic nerve head-centric views captured from both eyes. Our study introduces, to the best of our knowledge, the first vision transformer model for retinal fundus imaging in stroke assessment, offering a novel approach for capturing retinal patterns. Thereby, we propose the Braided Vision Transformer (BViT) model, which extracts representative features from the given multi-view images while simultaneously capturing inter-view relationships across both eyes, enabling a more informative understanding of retinal biomarkers associated with cerebrovascular events. Experiments conducted on our collected Stroke-Data dataset demonstrate that BViT achieves an AUC score of 0.75 for stroke detection, outperforming regular vision transformers.

[39] A Vision Transformer for ECG-Based Detection of Left Ventricular Systolic Dysfunction Across Multiple Clinical Sites cs.CV | cs.LGPDF

Burcu Ozek, Aruna Mohan, David Vorchheimer, Daniel Weiss, Eyal Kedar

TL;DR: 该论文提出了一种基于视觉Transformer（ViT）的模型，用于从常规12导联心电图中检测左心室射血分数降低（LVEF ≤ 40%）。模型在来自美国多个临床站点的10,142名患者数据上进行训练，并在独立的外部队列（4,092名患者）中验证，实现了高准确率（AUROC 0.88）和阴性预测值（97.8%）。研究强调了心电图作为可扩展的初步筛查工具，以识别需要进一步超声心动图检查的患者。

Details

Motivation: 左心室射血分数降低通常无症状且诊断延迟，而心电图虽常规记录却未被充分利用，因为LVEF降低没有单一诊断波形。论文旨在开发一种基于深度学习的自动化方法，从常规心电图中高效检测LVEF降低。

Result: 在外部验证队列（真实世界LVEF降低患病率8.72%）中，模型实现了AUROC为0.88（95% CI 0.86-0.89），敏感性81.2%，特异性81.0%，阴性预测值97.8%。敏感性在不同性别、种族、民族和合并症亚组中保持较高，而特异性在老年患者及房颤或心肌病患者中较低。

Insight: 创新点包括：1）使用视觉Transformer从零开始训练，并采用集成策略处理心电图图像；2）在心跳级别进行分析，提高了模型的可解释性，注意力图显示模型一致关注QRS波群而非P波；3）模型在多个独立临床站点的大规模真实世界数据上验证，展示了跨人群的泛化能力，为医疗AI的部署提供了参考。

Abstract: Reduced left ventricular ejection fraction (LVEF) is frequently asymptomatic and often detected only after advanced heart failure develops. Electrocardiograms are recorded routinely yet underused for this condition, because reduced LVEF has no single diagnostic waveform. We trained an ensemble of vision transformers from scratch to detect reduced LVEF ($\leq$40%) from 12-lead ECGs, analyzing each heartbeat individually, using 10,142 patients across seven sites in three US health systems. In a held-out external cohort of 4,092 patients from three geographically independent US clinical sites at a real-world reduced-LVEF prevalence of 8.72%, the model achieved an AUROC of 0.88 (95% CI 0.86-0.89), sensitivity 81.2%, specificity 81.0%, and negative predictive value 97.8%. Sensitivity remained high across sex, race, ethnicity, and comorbidity subgroups, while specificity was lower in older patients and those with atrial fibrillation or cardiomyopathy. Beat-level attention maps provided interpretability into the model’s predictions, showing consistent focus on the QRS complex rather than the P wave. These findings support the potential of routine ECGs as a scalable first-pass triage step to identify patients who should undergo echocardiography for reduced ejection fraction across diverse patient populations.

[40] Privacy-Preserving Dataset Curation for Kuala Lumpur Urban Traffic: Grounded Vision-Language Detection with Spatial Vehicle-Context Filtering cs.CV | cs.AI | cs.LGPDF

Mohammed Abdul Al Arafat Tanzin, Rudzidatul Akmam Dziyauddin

TL;DR: 本文针对马来西亚吉隆坡复杂热带城市环境下的交通数据集匿名化难题，提出了一种结合Grounded Vision-Language Detection与空间车辆上下文过滤的自动化隐私保护框架。该框架通过集成Grounding DINO视觉语言模型和新型空间车辆ROI约束引擎，有效抑制环境误报，自动模糊人脸、头部和车牌，在1266帧的初步评估中达到约95%的成功率。

Details

Motivation: 解决在吉隆坡高密度摩托车、深色亚克力车牌、动态相机倾斜和极端热带眩光等复杂条件下，传统匿名化方法（如Haar级联和YOLOv8）失效，导致个人可识别信息（PII）匿名化面临严重挑战的问题。

Result: 在吉隆坡道路数据集（1266帧）上的初步评估显示，该框架实现了约95%的成功率，剩余失败案例仅限于小尺寸、严重遮挡、倾斜或模糊的目标。

Insight: 主要创新点在于将零样本开放集视觉语言变换器（Grounding DINO）与新型空间车辆ROI约束引擎相结合，通过要求车牌质心位于已验证的车辆边界内，来抑制环境误报。同时，结合时间持续性机制和自动质量控制审计器，在保护下游视觉任务场景上下文的同时，最小化隐私相关的漏报。

Abstract: The rapid advancement of intelligent transportation systems and autonomous driving relies heavily on multi-modal urban traffic datasets. However, curating high-fidelity video imagery in complex tropical urban environments—specifically Kuala Lumpur, Malaysia—presents severe challenges for Personally Identifiable Information (PII) anonymization due to high motorcycle density, dark acrylic license plates, dynamic camera tilt, and extreme tropical glare. We propose an automated anonymization framework tailored for the Kuala Lumpur Road Dataset, captured via a mobile cycling platform at 2 FPS. We document how legacy Haar cascades and YOLOv8 fail under these conditions—generating false positives on background elements while missing rotated or occluded targets. Our architecture resolves this by integrating Grounding DINO—a zero-shot open-set vision-language transformer—with a novel Spatial Vehicle Region of Interest (ROI) Containment Engine. By requiring license plate centroids to reside within validated vehicle boundaries, the pipeline suppresses environmental false positives while automatically obfuscating faces, heads, and license plates. An initial evaluation on 1,266 frames demonstrates a $\sim$95% success rate, with remaining failures restricted to small, heavily occluded, oblique, or ambiguous targets. Coupled with temporal persistence mechanisms and an automated quality-control auditor, the framework minimizes privacy-related false negatives while preserving scene context for downstream vision tasks. While formal legal compliance depends on broader governance procedures, this publicly available pipeline and demonstration notebook provide an auditable preprocessing stage for privacy-aware dataset curation.

[41] Spatial Attention Noise Masking for Causally Sufficient Interpretability cs.CVPDF

Benjamin Formby, Kuang-Ching Wang, D Hudson Smith

TL;DR: 本文提出了一种新颖的因果可解释性方法，通过动态掩码输入图像来为计算机视觉模型提供因果解释。该方法包含一个UNet风格的掩码生成器和一个ResNet18编码器分类器，通过约束掩码稀疏平滑且掩码图像嵌入与原图一致，生成的特征归因图具有强因果性。

Details

Motivation: 现有可解释性方法多为被动应用于已训练模型，产生的是相关性而非因果性解释；而现有因果方法多为事后分析，因果主张较弱，且缺乏明确将责任分配给输入特征的主动解释方法。

Result: 定量评估表明，该方法生成的掩码具有忠实性，在五个分类任务上即使大幅掩码图像信息仍能保持接近基线的分类性能，并对背景替换和自然对抗样本等分布偏移具有鲁棒性。定性比较显示其与最先进的（SOTA）特征归因方法具有竞争力的可解释性。

Insight: 创新点在于提出了一种主动的、因果充分的解释框架，通过联合训练掩码生成器和分类器，并施加稀疏性、平滑性和嵌入一致性约束，使生成的归因图不仅是相关的，而且是因果的。这为高风险领域提供了更可靠的可解释性工具。

Abstract: We present a novel causal approach to interpretability for computer vision models that dynamically masks the input image prior to classification. The interpretability of deep learning predictions is critical in high-stakes fields such as medical imaging, security, and autonomous driving. Most interpretability methods are applied passively to already trained models, which typically result in correlational rather than causal explanations. Existing causal interpretability methods are limited to post hoc analysis, weakening the causal claims. Additionally, existing active methods generally lack explanations that explicitly assign responsibility to input features. This work proposes a spatial attention noise masking framework that provides causal explanations about the features sufficient for the prediction. The proposed framework consists of: 1) a UNet-style mask generator, and 2) a Resnet18 encoder and linear classifier that classifies both masked and unmasked versions of an input image. The generated masks are regularized to be sparse and spatially smooth, while masked image embeddings are constrained to remain consistent with embeddings from the corresponding unmasked images. The resulting masks can be interpreted as feature attribution maps that are competitive with related interpretability methods while additionally providing strong causal explanations of model predictions. Quantitative evaluations demonstrate mask faithfulness, near-baseline classification performance across five classification tasks despite substantial masking of image information, and robustness to distribution shifts such as background swapping and natural adversarial examples. Qualitative comparisons further demonstrate mask behavior and competitive interpretability relative to state-of-the-art feature attribution methods.

[42] Low Cost Two-Stage Fabric Defect Detection at the Edge cs.CVPDF

Rasel Hossen, Diptajoy Mistry, Mosaddek Hossain Kamal

TL;DR: 本文提出了一种用于低收入经济体服装行业织物缺陷检测的低成本两阶段级联系统，部署在NVIDIA Jetson Nano边缘设备上。第一阶段使用紧凑的卷积自编码器进行异常检测以筛选可疑帧，第二阶段仅在可疑帧上调用完整的YOLOv5n检测器。该系统在保持高召回率的同时，减少了计算负担并提升了处理速度。

Details

Motivation: 解决低收入经济体中小型纺织厂因商业视觉系统价格高昂而依赖人工织物检测的问题，通过级联设计在缺陷稀疏的生产环境中实现低成本、高效率的自动化缺陷检测。

Result: 在包含249张图像（20张缺陷，229张正常）的独立测试集上，第一阶段在优先召回率的阈值下成功标记所有20张缺陷图像（95% CI 0.83-1.00），假阳性率为49.3%，相比普通自编码器减少19.3%假阳性（p=0.011）。并行流水线达到13.45 FPS，比仅使用YOLO的串行循环（9.86 FPS）快1.36倍。

Insight: 创新点包括：1) 采用紧凑自编码器结合解码器注意力门、边缘加权重建损失和来自冻结YOLOv5n的特征级蒸馏作为高效异常检测器；2) 关键发现是速度提升主要源于JPEG解码与推理的重叠优化，而非级联本身，这警示了在评估级联加速时需控制数据路径；3) 系统定位为AI辅助分诊而非全自主验收，强调了在假阳性受限场景下的实用校准策略。

Abstract: Fabric inspection in the garment industries of low-income economies remains largely manual, and commercial vision systems are priced beyond most small and medium mills. Because defects are sparse under controlled production, a natural response is a cascade: screen every frame with a cheap anomaly detector and invoke a full detector only on suspicious frames. We build such a cascade for four knit-fabric defect classes and deploy it end-to-end on an NVIDIA Jetson Nano with TensorRT FP16. Stage 1 is a compact convolutional autoencoder with decoder attention gates, an edge-weighted reconstruction loss, and feature-level distillation from a frozen YOLOv5n teacher; Stage 2 is YOLOv5n, invoked only on flagged frames. On a 249-image benchmark disjoint from detector training (20 defective, 229 non-defective), Stage 1 at a recall-prioritised threshold flags all 20 defective images (95% CI 0.83-1.00) at a false-positive rate of 49.3% (113/229), reducing false positives by 19.3% relative to a plain autoencoder (p=0.011). The parallel pipeline reaches 13.45 FPS against 9.86 FPS for a sequential YOLO-only loop. Our central finding comes from decomposing that 1.36x: 91% of it is attributable to overlapping JPEG decode with inference rather than to the cascade, which contributes only a 5.1% inference reduction at the measured forwarding rate p = 0.534. We further show that forwarding here is false-positive-limited rather than prevalence-limited - 85% of forwarded frames are false alarms - and quantify the 29-45% inference reduction attainable under tighter calibration. We report this as a caution for cascade speedups measured without controlling the data path, and position the system as AI-assisted triage rather than autonomous acceptance.

[43] From Dense Prediction to Visual Editing: Structured Supervision for Unified Image and Video Creation cs.CVPDF

Zhefan Rao, Bin Zou, Haoxuan Che, Xuanhua He, Chong Hou Choi

TL;DR: 该论文提出了一种通过引入密集预测任务（如深度和表面法线预测）作为结构化视觉监督，来提升统一图像和视频生成模型性能的方法。该方法使用单一的多模态扩散Transformer（MMDiT）主干网络，结合互上下文注意力（MCA）、配对视频数据构建和渐进式训练课程，旨在增强模型在遵循指令的同时保持视觉上下文（如身份、几何和时序结构）的能力。

Details

Motivation: 现有的统一图像和视频生成模型主要依赖语义层面的条件输入和生成任务训练，缺乏对局部结构和时序一致性的显式监督，导致在精确、时序一致的编辑任务上表现不足。

Result: 在统一系统的比较中，该方法获得了最高的综合得分（4.15）。具体地，添加密集监督后，在OpenVE基准上的Overall分数从3.98提升至4.06，Local Add分数从3.92提升至4.18。

Insight: 核心创新点在于将感知导向的密集预测任务（如深度、法线估计）作为图像形式的去噪目标，并将其作为结构化监督整合到生成框架中，从而将语义理解与空间对齐的视觉注入解耦。这有效提升了模型在下游生成任务，特别是编辑的局部性和保持性方面的能力，但作者强调其目的并非成为独立的顶级密集预测器。

Abstract: Unified image and video creation requires a model to follow diverse instructions while preserving identity, geometry, and temporal structure from visual context. However, semantic-only conditioning and creation-only training do not explicitly supervise the local structure needed for precise, temporally consistent editing. We therefore formulate depth and surface-normal prediction as image-form denoising targets, using these dense tasks as structured visual supervision within the same creation interface. Our framework decouples semantic interpretation from spatially aligned visual injection while sharing one multimodal diffusion transformer (MMDiT) backbone across all tasks. Mutual Context Attention (MCA), a paired-video data-construction procedure, and a progressive training curriculum then connect the learned structural cues to temporally localized editing and reference-conditioned creation. A single checkpoint obtains the highest overall score in the reported comparison of unified systems (4.15); adding dense supervision improves OpenVE Overall from 3.98 to 4.06 and Local Add from 3.92 to 4.18. These results support a deliberately bounded conclusion: perception-oriented dense supervision transfers useful structural knowledge to downstream creation, especially editing locality and preservation; we do not claim superiority as a standalone dense predictor.

[44] PolyComp: A Polycube-based Benchmark for Compositional 3D Spatial Reasoning in Multimodal Models cs.CV | cs.AIPDF

Siddharth Patel

TL;DR: 本文提出了PolyComp基准测试，这是一个用于评估多模态模型在组合式三维空间推理能力的程序化生成和验证的基准。该基准包含120个问题，涵盖四种几何家族，每个问题有三种呈现格式（单图或多图）。测试要求模型从四个选项中识别出哪一对多立方体组件可以组合成目标立体形状。

Details

Motivation: 动机是创建一个能够有效评估多模态模型视觉识别和组合式空间推理能力的基准，当前模型在这类需要复杂三维空间理解的任务上表现尚不明确。

Result: 在三种呈现格式下（每个模型共360个问题），GPT-5.6 Sol（最大努力）准确率为50.0%（问题簇CI 43.3-56.7%），平均成本为每个问题0.951美元；Claude Fable 5（最大努力）为39.4%（33.1-46.1%），成本0.701美元；Gemini 3.1 Pro Preview（高思考级别）为27.5%（22.8-32.5%），接近25%的随机猜测基线，成本0.350美元。不同几何家族间的准确率差异大于不同呈现格式间的差异。

Insight: 创新点在于提出了一个程序化生成和验证的、专注于组合式三维空间推理的基准测试PolyComp，并提供了详细的问题开发与评估协议、成本与令牌核算方法。该基准能有效揭示当前先进多模态模型在复杂空间推理任务上的能力局限与成本效益差异。

Abstract: We introduce PolyComp, a procedurally generated and verified benchmark that stresses visual recognition and compositional spatial reasoning. In each problem, a model must identify which of four options shows a pair of polycube components that can be combined to form a target solid. The benchmark contains 120 problems across four geometry families, and each problem has three different presentation formats using either a single image or multiple images. The random guessing baseline is 25%. Across the three presentations (360 presented problems per model), GPT-5.6 Sol with max effort attains 50.0% accuracy (95% problem-cluster CI 43.3-56.7%) at a mean cost of $0.951 per presented problem, Claude Fable 5 with max effort attains 39.4% (33.1-46.1%) at $0.701, and Gemini 3.1 Pro Preview with thinking level high attains 27.5% (22.8-32.5%), near the 25% random guessing baseline, at $0.350. The observed accuracy spread across geometry families is larger than across presentation formats. We present a problem development and evaluation protocol, cost and token accounting, and release the 120 problems.

[45] Beyond Boundary Noise: Aggregated Aleatoric Uncertainty Fails to Capture Presence Ambiguity in 3D Lung Nodule Segmentation cs.CV | cs.LGPDF

Simon Baur, Arne Schernich, Ekin Böke, Wojciech Samek, Jackie Ma

TL;DR: 本文系统评估了基于熵的认知不确定性在3D肺结节分割中捕捉存在性模糊（即病理是否存在的病例级分歧）的能力，发现其与边界噪声和微小标注差异相关，但无法有效区分存在性模糊。相比之下，在冻结分割特征上训练的轻量级监督模糊头显著优于所有基于熵聚合的基线方法，并达到或超过显式建模模糊的先进方法。

Details

Motivation: 深度学习在医学图像分割中的安全临床部署需要可靠的不确定性估计，但基于熵的认知不确定性是否能够捕捉临床上有意义的存在性模糊（即病理是否存在的分歧）尚不明确，本文旨在系统评估这一问题。

Result: 在LIDC-IDRI和外部验证队列LNDb上的实验表明，基于熵聚合的不确定性方法在存在性模糊任务上表现不佳；而提出的轻量级监督模糊头在多种架构、评估指标和队列上均显著优于基线，并与Probabilistic U-Net、Annotator-Confusion 3D-UNet等方法相当或更优。

Insight: 研究发现，存在性模糊信息已编码在逐像素训练网络的冻结编码器特征中，但被分割输出及其熵聚合所丢弃；这揭示了认知不确定性的理论承诺与实际行为之间的根本错配，提示在安全关键应用中不应依赖基于熵的不确定性作为临床模糊的代理。

Abstract: Uncertainty estimation is critical for the safe clinical deployment of deep learning in medical image segmentation, with aleatoric uncertainty theoretically designed to capture irreducible data ambiguity. However, whether entropy-based measures reflect clinically meaningful ambiguity, i.e. case-level disagreement about whether a pathology is present at all, remains poorly understood. Contrary to most prior work, which focused on pixel-wise boundary disagreement, we systematically evaluate how well aleatoric uncertainty captures presence ambiguity. Our evaluation spans 3D lung nodule segmentation across four architectures with Monte Carlo dropout and deep ensembles, on LIDC-IDRI and an external validation cohort (LNDb). We find that entropy-based uncertainty maps align with boundary noise and minor drawing variation but carry insufficient discriminative signal for presence ambiguity. In contrast, a lightweight supervised ambiguity head trained on frozen segmentation features substantially outperforms all entropy-aggregation-based baselines across architectures, metrics, and both cohorts, and matches or exceeds methods that explicitly model ambiguity under disagreement supervision (Probabilistic U-Net, Annotator-Confusion 3D-UNet). A qualitative feature-space analysis shows that presence ambiguity is already encoded in the frozen encoder features of pixel-wise-trained networks, only to be discarded by the segmentation output and its entropy aggregation. Our findings expose a fundamental mismatch between the theoretical promise of aleatoric uncertainty and its practical behavior, and suggest that practitioners should not rely on entropy-based uncertainty as a proxy for clinical ambiguity in safety-critical applications.

[46] NARRATE: A Multimodal Real-World Australian Driving Dataset for Human-Centred Explanations in Automated Driving cs.CV | cs.AI | cs.CL | cs.ROPDF

Ashkan Yousefi Zadeh, Zishuo Zhu, Xiaomeng Li, Andry Rakotonirainy, Sebastien Glaser

TL;DR: 本文介绍了NARRATE，一个多模态的真实世界澳大利亚驾驶数据集，包含来自35名经验丰富的驾驶员和驾驶教练在公共道路上收集的2,050个标注事件。每个事件都基于同步的视觉、定位、运动和LiDAR数据流，并与车内和/或驾驶后的自由文本解释配对。该数据集提供了动作标签、涵盖六个高级别和32个细粒度类别的场景上下文标签，以及对驾驶员解释的感知、理解和预测层面的情境意识标注。四个基准任务表明，这种结构可以从驾驶员语言中学习，但细粒度上下文识别和解释生成仍具挑战性。

Details

Motivation: 现有语言标注的驾驶数据集多为观察者编写、事后记录、基于模拟或从传感器输入生成，而非从执行动作的驾驶员处获取，这限制了自动驾驶系统向乘客提供可理解、可监控和可信的解释能力。

Result: 在四个基准任务（情境意识、场景上下文、驾驶员动作分类和解释生成）上，该数据集的结构被证明可从驾驶员语言中学习，但细粒度上下文识别和解释生成任务仍面临挑战。

Insight: 创新点在于通过真实驾驶场景中驾驶员主动提供的多模态同步数据与自由文本解释，构建了一个以人为中心的驾驶数据集，支持情境意识标注，为开发更符合人类认知的自动驾驶解释模型提供了新路径。

Abstract: Automated vehicles must explain their decisions in ways that passengers can understand, monitor, and trust. Existing language-annotated driving datasets are mostly observer-written, post-hoc, simulation-based, or generated from sensor inputs, rather than elicited from the driver performing the action. We introduce NARRATE, a multimodal real-world Australian driving dataset comprising 2,050 annotated events from 35 experienced drivers and driving instructors on public roads. Each event is grounded in synchronised visual, localisation, motion, and LiDAR streams and paired with in-vehicle and/or post-drive free-text explanations. NARRATE provides action labels, scenario-context labels spanning six high-level and 32 fine-grained categories, and span-level Situational Awareness (SA) annotations over driver explanations for Perception, Comprehension and Projection. Four benchmark tasks (SA, scenario-context, driver-action classification, and explanation generation) show that this structure is learnable from driver language, while fine-grained context recognition and explanation generation remain challenging. NARRATE paves a path towards more human-centred and domain-aware explanation models for automated driving.

[47] AMPLIFAI: A Multiphase CT Dataset for Benchmarking Clinical Reasoning in LI-RADS Assessment of Liver Lesions cs.CV | cs.LGPDF

Pranav Kulkarni, Nikhil Shah, Amritansh Suryavanshi, Jana Delfino, James Tonascia

TL;DR: 论文介绍了AMPLIFAI数据集，这是首个公开的多期相腹部CT扫描数据集，专门用于基于LI-RADS标准的肝细胞癌（HCC）临床推理评估。该数据集包含LI-RADS类别标注，并对动脉期高强化、廓清和强化包膜三个关键特征进行了分割，旨在推动人工智能在HCC自动化诊断中的发展。

Details

Motivation: 肝细胞癌是全球癌症相关死亡的第三大原因，早期检测可显著提高生存率。虽然LI-RADS标准为基于影像的自动化诊断提供了框架，但缺乏高质量标注的大型公开数据集限制了相关AI模型的开发。

Result: 论文未在摘要中提及具体的模型性能或基准测试结果，而是专注于介绍数据集本身及其构建过程。

Insight: 创新点在于创建了首个公开的、符合LI-RADS标准的多期相CT数据集，并提供了详细的标注（包括类别和关键特征分割），遵循《Datasheets for Datasets》格式以确保透明度和可重复性，为AI在肝病变临床推理中的基准测试奠定了基础。

Abstract: Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related mortality worldwide, with early detection improving survival from <20% to >70%. The standardized LI-RADS criteria establish a biopsy-free, fully imaging-based framework that can serve as a foundation for automating HCC diagnosis with artificial intelligence (AI). However, the lack of large, publicly available datasets with high-quality labels has limited the development of AI models for LI-RADS characterization. We introduce the \textbf{AMPLIFAI} dataset, the first public dataset of multiphase abdominal CT scans annotated with LI-RADS categories and segmented for three major LI-RADS features: arterial phase hyperenhancement, washout, and enhancing capsule. Following the \emph{Datasheets for Datasets} format, this paper details the dataset’s composition, curation process, and annotation pipeline to facilitate transparent, reproducible research.

[48] Qwen-Video-Edit: Instruction-Based Video Editing by Repurposing an Image Editing Model cs.CVPDF

Yunpeng Bai, Yossi Gandelsman, Michaël Gharbi, Qixing Huang

TL;DR: 本文提出Qwen-Video-Edit，一种基于指令的视频编辑方法。其核心创新在于直接利用强大的图像编辑模型（Qwen-Image-Edit）来编辑视频，而无需昂贵的视频预训练主干网络。方法将视频VAE的潜在帧排列为一个大虚拟图像的图块，通过轻量级投影桥接图像与视频潜在空间，并在Ditto-1M数据集上微调，最终通过Wan~2.2进行可选的时间增强。

Details

Motivation: 动机是探索一条不同于主流（基于视频预训练扩散模型）的、更经济的视频编辑路径。旨在解决基于指令的视频编辑通常需要昂贵视频扩散模型适配的问题，转而尝试复用成熟的图像编辑先验知识。

Result: 方法在公开的Ditto-1M编辑三元组数据集上进行微调。结果表明，即使未经训练（零样本），基础图像编辑器也能对视频潜在表示进行可识别的编辑，微调后主要弥补了保真度差距，证明了图像编辑先验向视频域迁移的有效性。

Insight: 宣称的创新点在于通过将视频帧潜在表示平铺为虚拟图像并重用图像位置编码，以极小的适配成本（轻量级投影层）将图像编辑模型成功应用于视频编辑。客观来看，其核心洞察是：尽管视频潜在空间训练投入巨大，但逐帧的视频潜在表示与图像域足够接近，成熟的图像编辑先验可以以最小代价迁移过来，这为视频生成/编辑模型设计提供了新思路。

Abstract: Instruction-based video editing is commonly built on video-pretrained generative backbones: a video diffusion transformer is adapted, at considerable cost, to condition on a source video and an editing instruction. In this report we explore a different route and show that a strong instruction-based image editing model can edit videos by operating directly on video-VAE latents. Starting from Qwen-Image-Edit, we arrange the latent frames of a Wan2.1 video VAE as tiles of one large virtual image, reuse the editor’s image positional encoding for every tile, and bridge the two latent spaces with a pair of lightweight input/output projections warm-started from the editor’s own patchify and unpatchify layers, so that at initialization a (static) video is embedded exactly as an image the model already understands. The whole system is then fine-tuned on the public Ditto-1M editing triplets, and a few denoising steps of Wan2.2 serve as an optional temporal enhancer. We motivate the design with a chain of zero-training observations: the stock image editor already edits a video presented as a contact sheet; it is indifferent to whether the sheet’s tokens come from one joint encode or from per-frame encodes stitched in latent space; and it even edits genuine video latents zero-shot to a clearly recognizable degree, leaving fine-tuning only a fidelity gap to close. Our results suggest that, despite the large investment in training video latent spaces, per-frame video latents remain close enough to the image domain that mature image editing priors transfer with minimal adaptation. Project Page: https://yunpeng1998.github.io/Qwen-Video-Edit-Page; Code: https://github.com/yunpeng1998/Qwen-Video-Edit; Model: https://huggingface.co/yunpeng1998/Qwen-Video-Edit.

[49] Where the Cost Falls: A Deployment-Aware Adoption Order for Stability Enhancements to Cycle-Consistent Adversarial Networks cs.CVPDF

Rowan Hussein, Mohamed Ouf

TL;DR: 本文针对CycleGAN在无配对图像翻译中遇到的训练不稳定、纹理漂移和局部伪影等问题，提出了四种改进方法，并基于计算成本与部署影响提出了一个采用顺序建议。

Details

Motivation: CycleGAN在应用中面临对抗训练振荡或崩溃、循环一致性仅保留粗略布局而细纹理漂移、以及单一判别器忽略局部伪影等障碍，需要系统性地评估改进方法的实际部署成本。

Result: 在horse-to-zebra翻译任务上，结合所有四种改进的模型报告了Fréchet Inception Distance和Kernel Inception Distance；但论文指出仍需进一步协议来全面评估各变体的性能。

Insight: 创新点在于根据改进方法对训练与部署阶段的不同成本影响（如自注意力机制导致内存平方增长）提出资源受限团队的采用优先级，而非仅比较输出质量。

Abstract: Teams that adopt cycle-consistent adversarial networks for unpaired image-to-image translation meet the same obstacles: adversarial training oscillates or collapses, cycle consistency preserves coarse layout while finer texture drifts, and a single discriminator judging global realism misses local artifacts. Four enhancements address these failures, and they are usually compared on output quality alone. We show that they also divide sharply by where their cost falls, and that this division, which follows from the architecture and not from any particular run, yields an adoption order for teams under a compute or latency budget. A Wasserstein objective with gradient penalty, a VGG19 perceptual loss on the cycle reconstruction, and multi-scale discriminators change training only, so a team can adopt or drop them without altering what ships. Self-attention alone persists into the deployed generator, with memory growing as the square of the feature-map size, which makes it the one component a resource-constrained team should defer. We integrate all four onto a lightly tuned baseline for horse-to-zebra translation, introduced one at a time on a fixed control and then combined, and for each we give the failure mode it targets and how it integrates. We document the collapse and reconstruction-artifact modes the baseline produced, report what visual inspection of saved samples showed for each variant, and report Fréchet Inception Distance and Kernel Inception Distance for the combined model. We specify the protocol still needed, covering the individual variants, perceptual similarity, and downstream segmentation, to rank these enhancements on measured evidence.

[50] Zero-Shot Adaptation of Medical Vision Foundation Models for High-Frequency Micro-Ultrasound Prostate Segmentation cs.CV | cs.LGPDF

Ayusha Abbas, Saram Abbas, Kabita Adhikari

TL;DR: 本文提出了一种针对高频微超声前列腺分割的零样本适应方法，利用预训练的医学视觉基础模型MedSAM进行定位，并结合CLAHE增强、二值膨胀和傅里叶平滑等技术来优化分割边界，无需训练数据即可实现高精度分割。

Details

Motivation: 解决高频微超声（29 MHz）图像中因密集声学斑点导致前列腺边界模糊、临床医生分割一致性差的问题，并克服有监督方法成本高、泛化性差的局限性，实现无需训练数据的零样本分割。

Result: 在Micro-Ultrasound Prostate Segmentation数据集的20名患者测试集上，该方法将Dice系数从0.749提升至0.865，HD95误差从217.2像素降低至120.1像素，分割一致性比非专家标注者高38-52%，且仅需近似边界框作为提示即可部署。

Insight: 创新点在于首次将零样本适应流程应用于高频微超声前列腺分割，结合预训练基础模型与后处理技术，显著提升分割精度和一致性；客观来看，该方法通过简单的边界框提示实现了高效部署，避免了数据收集和重新训练的成本。

Abstract: Prostate cancer claims a life every 80 seconds. Early detection is needed to prevent disease progression, and both PSA density calculation and biopsy decisions rely on knowing the exact boundary of the gland. Conventional ultrasound at 6-12 MHz blurs this boundary, missing one in three high-risk cancers. Micro-ultrasound (29 MHz) improves resolution threefold but introduces dense acoustic speckle that obscures the outer wall; given the same image, two clinicians draw outlines differing by over 10% in area. Supervised methods are costly and generalise poorly across scanners. Can a foundation model segment the prostate with no training data? We present the first zero-shot pipeline for this modality: MedSAM, pre-trained on over 1.5 million medical images, localises the prostate; we then apply CLAHE to sharpen the outer wall, binary dilation to recover missed pixels, and Fourier smoothing (4 modes, s=1.05) to refine the boundary. MedSAM requires a spatial prompt, so we evaluate bounding-box and point-click strategies across 75 patients of the Micro-Ultrasound Prostate Segmentation dataset (2,621 slices). On the 20-patient held-out test set, the pipeline reduces mean boundary-distance error by 45% (Dice 0.749+/-0.043 to 0.865+/-0.029; HD95 217.2+/-36.9 to 120.1+/-26.1 px), reaching Dice 0.859 across the cohort. Its mean overlap shows no significant difference from the three non-expert rater groups (p>0.19), while segmenting 38-52% more consistently (lower inter-patient standard deviation). Point-click prompts fail regardless of placement (best Dice=0.350), because speckle gives no stable local contrast. Only an approximate bounding box is required, so any clinic can deploy it without data collection, annotation, or retraining.

[51] Artificial Intelligence as a Tool for Combating Child Labour: A Real-Time Edge Vision Pipeline for Child Detection and Age Estimation cs.CV | cs.AIPDF

Mark Nowak

TL;DR: 本文提出了一种用于打击童工的实时边缘视觉管道，结合了多任务人员与面部检测、级联年龄估计、跟踪与重识别技术，旨在为童工监测与补救系统提供持续、基于存在的证据渠道。该系统在嵌入式硬件上实现了超过两倍的实时性能，并在津巴布韦农场进行了为期17天的无人值守实地试点，显著提升了检测效率并减少了误报。

Details

Motivation: 全球仍有约1.38亿儿童处于童工状态，现有基于定期家访和访谈的监测系统存在系统性漏检问题。本研究旨在探索通过计算机视觉技术为童工监测系统提供连续、基于视觉证据的可行性，以弥补传统方法的不足。

Result: 在检测方面，人员mAP@0.5从基线0.390提升至0.683；儿童专用年龄估计模型在仅含儿童的验证集上达到1.944年平均绝对误差，而广泛使用的开源栈误差高达18-23年。FP8 TensorRT编译带来1.77倍加速，使管道在嵌入式硬件上达到超过两倍的实时性能。在代理视频中，系统发现634个独特儿童候选者，而前代系统仅发现285个。实地试点显示，软件调优使检测产出提升36倍，身份合并将过度报告从9.1倍降至1.8-3.9倍，且无误合并。

Insight: 创新点在于构建了一个集成了检测、年龄估计、跟踪与重识别的端到端实时边缘视觉管道，并专门针对儿童（0-12岁）优化了年龄估计模型。客观来看，该研究在嵌入式部署优化（如FP8量化）和实地系统调优（如身份合并策略）方面提供了实用的工程见解，同时强调了数据保护和人在回路保障的必要性。

Abstract: An estimated 138 million children remain in child labour worldwide, and the monitoring systems used by affected sectors, built on periodic household visits and interviews, systematically under-detect them. We present a real-time computer-vision pipeline, built and operated solely as a research prototype, that studies the feasibility of giving Child Labour Monitoring and Remediation Systems (CLMRS) a continuous, presence-based evidence channel. The pipeline combines a multi-task person and face detector (YOLO26x backbone in the CerberusDet framework), cascaded age estimation pairing MiVOLO v2 with a child-specialist model for ages 0-12, ByteTrack tracking, ArcFace and DINOv2 re-identification, and track-level fusion producing reviewable per-person records. The detector raises person mAP@0.5 from 0.390 to 0.683 over the previous-generation baseline; the child specialist reaches 1.944 years MAE on children-only validation, where widely used open-source stacks err by 18-23 years. FP8 TensorRT compilation yields a 1.77x speedup at +0.002 years MAE, bringing the pipeline above twice real-time on embedded hardware. On 26.8 hours of proxy video the system finds 634 unique child candidates versus 285 for its predecessor. We further report a seventeen-day unattended field pilot on a farm in Zimbabwe (38.7 million frames, six cameras) evaluated against a daily attendance register: software tuning improved detection yield 36-fold, and identity consolidation under a simultaneity veto cut over-reporting from 9.1x to 1.8-3.9x with zero proven-false merges. We document training and quantisation failures alongside successes, and the data-protection and human-in-the-loop safeguards such a system requires.

[52] OvDSGG: End-to-End Open-Vocabulary Dynamic Scene Graph Generation cs.CVPDF

John Helsby, Yi Yang, Bodo Rosenhahn, Michael Ying Yang

TL;DR: 本文提出了OvDSGG，首个端到端的开放词汇动态场景图生成框架，旨在解决现有方法局限于固定训练词汇且难以处理罕见概念长尾分布的问题。该框架通过空间主干、时间主干、三元组特征提取模块和视觉-语言对齐模块，实现了无需昂贵知识蒸馏的开放词汇识别，并在新构建的基准上显著优于现有基线。

Details

Motivation: 现有动态场景图生成方法为闭集，仅能识别固定训练词汇中的对象和谓词，且难以处理罕见概念的长尾分布，限制了实际应用；而现有开放词汇模型通常依赖预训练大语言模型，导致多阶段训练和推理成本高昂。

Result: 在基于Action Genome构建的严格开放词汇DSGG基准上，OvDSGG在所有指标上显著优于开放词汇基线，零样本Recall@K分数比次优基线高10.0-20.4个百分点；在闭集DSGG任务上，其性能与最先进模型保持竞争力。

Insight: 创新点包括端到端的开放词汇DSGG框架设计、连接空间与时间主干的三元组特征提取模块，以及通过在联合视觉-语言特征空间中学习自适应决策边界来保持开放词汇识别的视觉-语言对齐模块，避免了现有方法中昂贵的知识蒸馏过程。

Abstract: Dynamic scene graphs (DSGs) capture spatio-temporal interactions across videos as $\langle$subject, predicate, object$\rangle$ triplets, and underpin downstream tasks such as video captioning, video question answering, and action analysis. However, end-to-end dynamic scene graph generation (DSGG) methods are closed-set: they recognize only objects and predicates from a fixed training vocabulary and struggle with the long-tailed distribution of rare concepts, severely limiting their real-world applicability. Existing open-vocabulary models typically inherit pretrained large language models, resulting in multi-stage training and inference with substantial cost. We introduce OvDSGG, the first end-to-end framework for open-vocabulary DSGG. OvDSGG builds on top of an open-vocabulary Spatial Backbone and a Temporal Backbone; we further propose a Triplet Feature Extraction Module that bridges them, and a Visual-Language Alignment Module that preserves open-vocabulary recognition by learning an adaptive decision boundary in the joint visual-language feature space, without expensive knowledge distillation in existing methods. We further introduce a rigorous open-vocabulary DSGG benchmark adapted from Action Genome, with disjoint Base/Novel splits for both objects and predicates. OvDSGG significantly outperforms open-vocabulary baselines across all metrics, with zero-shot Recall@$K$ scores 10.0–20.4 percentage point higher than the next-best baseline, while on closed-set DSGG remaining competitive with state-of-the-art models. Code and benchmark are publicly available at https://github.com/jhelsby/OvDSGG/.

[53] Zero-MELO: Test-Time Evidence Calibration with Multimodal LLMs for Zero-Shot Micro-Gesture Recognition cs.CVPDF

Chengyan Wang, Hanliang Xie, Yueyi Yang, Haoyu Chen

TL;DR: 本文提出了一种名为Zero-MELO的测试时证据校准框架，旨在提升多模态大语言模型在零样本微手势识别任务中的性能。该框架通过树搜索机制获取细粒度视觉证据，并使用校准模块缓解评分偏差，最终通过多线索融合进行预测。

Details

Motivation: 多模态大语言模型在通用视频理解上表现出色，但在细粒度、以动作为中心的任务（如微手势识别）上能力有限。作者认为其性能受限源于单次前向推理模式，并希望通过精心设计的测试时引导来增强其运动敏感的表征能力。

Result: 在iMiGUE和MA-52基准测试上，该方法分别取得了26.84%和22.10%的平均类别准确率，显著超越了Qwen2.5-VL基线的16.15%和10.20%，实现了性能的大幅提升。

Insight: 创新点在于诊断了MLLMs在零样本MGR任务中的两大瓶颈（局部证据不足和评分偏差），并提出了一个集成了渐进式树搜索证据收集、测试时偏差校准和多线索融合的端到端测试时优化框架，无需训练即可有效提升模型对细微动作的判别能力。

Abstract: While Multimodal Large Language Models (MLLMs) excel in general video understanding, their capability in fine-grained and motion-centric tasks remains limited. This limitation is particularly critical in micro-gesture recognition (MGR), where micro-gestures (MGs) - subtle, short-duration, and spatially localized human movements - serve as key discriminative signals for implicit affective analysis, yet are easily neglected following common prompting practices. Although MGR has been intensively studied by many discriminative approaches, the use of MLLMs for MGR is underexplored, with notably poor performance. We hypothesize that the motion-sensitive representation ability of MLLMs is constrained by their inherent single-pass forward inference, which can be substantially enhanced through carefully designed test-time guidance. Motivated by this, building on our prior findings regarding temporal insensitivity in Video LLMs, we diagnose zero-shot MGR errors in the Negative Log-Likelihood (NLL) space. We observe that MLLMs suffer from two bottlenecks: 1) insufficient localized evidence and 2) severe score biases driven by language and motion-agnostic appearances. Thus, we propose a novel test-time evidence calibration framework that improves both reasoning details and prediction reliability. Specifically, we introduce a tree search mechanism to progressively acquire localized, fine-grained visual evidence, coupled with a test-time calibration module to mitigate score biases. The multi-cue fusion module then integrates evidence from multiple cues without relying on a single cue for final prediction. Our framework achieves mean-class accuracies of 26.84% on iMiGUE and 22.10% on MA-52, significantly outperforming the Qwen2.5-VL baseline, which produces 16.15% and 10.20%, respectively. The code will be available at https://zero-melo.github.io/Zero-MELO.

[54] SpIn-ViT: Designing a Sparsity-Induced Vision Transformer That Is Mechanistically Interpretable cs.CV | cs.AIPDF

Philip H. Lee, Parth Padalkar

TL;DR: 本文提出SpIn-ViT框架，将预训练的视觉Transformer（ViT）与改进的稀疏自编码器（SAE）进行端到端联合训练，使稀疏的patch级表示与图像分类目标直接对齐。该方法在保持竞争力的预测性能的同时，学习到了语义连贯且可定位有意义图像区域的神经元激活。

Details

Motivation: 现有后处理稀疏自编码器（SAEs）在ViT优化完成后对其冻结表示进行训练，其潜在特征与下游分类目标未直接对齐，限制了机制可解释性的效果。

Result: 在九个图像分类基准测试上，SpIn-ViT相比之前最先进的后处理SAE方法，平均分类准确率提高了8.84%，基于AI的可解释性分数提高了近四倍，人工评估分数提高了一倍以上。其提取的神经符号模型平均准确率也提高了5.97%，且所需规则集规模缩小了58.8%。

Insight: 核心创新在于将SAE与ViT进行端到端联合训练，使稀疏特征学习直接服务于分类目标，从而在机制可解释性与模型性能之间取得了更好的平衡。该方法还展示了如何利用SAE神经元构建更紧凑、更准确的神经符号模型。

Abstract: Mechanistic interpretability has recently expanded to Vision Transformers (ViTs), with Sparse Autoencoders (SAEs) increasingly used as post-hoc tools to decompose internal representations into sparse and more interpretable features. However, because post-hoc SAEs are trained on frozen representations after the ViT has already been optimized, their latent features are not directly aligned with the downstream classification objective. We introduce SpIn-ViT, a framework that jointly trains a pretrained ViT and a modified SAE end-to-end, directly aligning sparse patch-level representations with image classification. SpIn-ViT learns semantically coherent neuron activations that localize meaningful image regions while maintaining competitive predictive performance. We evaluate SpIn-ViT across nine image-classification benchmarks using classification accuracy, quantitative interpretability metrics, AI-based and Human evaluations. Compared with the previous state-of-the-art post-hoc SAE method, SpIn-ViT achieves 8.84% higher average classification accuracy, an AI-based interpretability score nearly four times as high, and a human-evaluation score more than twice as high. We further extract interpretable rule-sets using the SAE neurons to create neurosymbolic models which achieve 5.97% higher average classification accuracy while requiring a 58.8% smaller rule-set than the neurosymbolic models created from the SOTA post-hoc SAE method.

[55] PaSTel: Anchoring Histology in Spatial Transcriptomics via Multi-Scale Hierarchical Bio-Prior Contrastive Pretraining cs.CV | cs.AIPDF

Azim Dehghani Amirabad, Junchao Zhu, Pushpak Pati, Walid Abdelmoula, Tommaso Mansi

TL;DR: PaSTel是一个用于空间转录组学的多尺度分层生物先验对比预训练框架，通过整合斑点级、功能级和区域级的生物先验，将组织学图像与基因表达对齐，以解决现有方法中基因选择受管家基因主导和独立斑点-图像对齐忽略空间依赖性的问题。

Details

Motivation: 现有方法在空间转录组学中面临两个关键限制：空间信息基因选择常被普遍存在的管家基因主导，导致表示区分性弱；独立的斑点-图像对齐无法捕捉对组织至关重要的空间依赖性。

Result: 在多个下游任务中，PaSTel持续优于现有的视觉和视觉-组学编码器，表明整合多尺度生物先验能产生更具信息性和可迁移性的表示。

Insight: 创新点包括使用TF-IDF重加权识别空间信息基因、利用KEGG通路作为全局生物语义锚点，以及通过空间聚类聚合邻近斑点建模中尺度组织结构，从而通过分层生物先验增强表示学习。

Abstract: Spatial transcriptomics (ST) links tissue morphology with molecular programs, motivating multimodal pretraining methods that align histology images with gene expression. However, existing approaches suffer from two key limitations: spatially informative gene selection is often dominated by ubiquitous housekeeping genes, leading to weakly discriminative representations, and independent spot-patch alignment fails to capture spatial dependencies that are critical for tissue organization. To address these challenges, we introduce PaSTel, a hierarchical multimodal pretraining framework that integrates biological priors at three levels. At the spot level, TF-IDF reweighting is used to identify spatially informative genes; at the functional level, curated KEGG pathways serve as anchors for encoding global biological semantics; and at the regional level, spatial clustering aggregates neighboring spots to model meso-scale tissue structure. Across multiple downstream tasks, PaSTel consistently outperforms existing vision and vision-omics encoders, demonstrating that incorporating multiscale biological priors yields more informative and transferable representations for spatial transcriptomics.

[56] Beam-Wise Statistical Background Subtraction for Static Roadside LiDAR: A Cross-Sensor Benchmark Study cs.CV | cs.ROPDF

Alexander Baumann, Marcel Vosshans, Thao Dang

TL;DR: 本文针对静态路边LiDAR的背景减除问题，提出了基于光束级统计建模的方法，并引入了新的多LiDAR数据集HighwayScene及扩展的CoopScenes数据集进行可重复评估。研究表明，结合轻量级光束级模型与空间一致性滤波能显著提升精度，同时保持高召回率和实时性。

Details

Motivation: 解决静态路边LiDAR感知中背景减除缺乏系统性跨传感器评估和可重复研究的问题，以高效分离动态交通参与者。

Result: 在多个场景和异构传感技术中，光束级统计建模表现出鲁棒性和可迁移性；结合空间滤波后，在保持高召回和实时性的同时显著提升了精度。

Insight: 创新点在于将背景估计公式化为光束级时间建模问题，并探索互补的统计策略以捕捉主导和多模态背景结构，同时结合角域和3D域的空间滤波；客观分析认为其提出的跨传感器基准和公开数据集推动了该领域的可重复研究。

Abstract: Background subtraction is a key preprocessing step for infrastructure-based LiDAR perception, enabling efficient isolation of dynamic traffic participants without semantic annotations. However, systematic cross-sensor evaluations and reproducible studies for static roadside LiDAR are missing. This paper presents a comparative benchmark of beam-wise statistical background subtraction for statically mounted LiDAR sensors. We formulate background estimation as a per-beam temporal modeling problem and investigate complementary statistical strategies that capture dominant as well as multi-modal background structures, combined with spatial filtering in the angular and 3D domain. To enable reproducible evaluation, we introduce HighwayScene, a new multi-LiDAR dataset recorded in a static roadside setup, and extend the public CoopScenes dataset with static/dynamic point-wise annotations. Across multiple scenes and heterogeneous sensing technologies, we demonstrate that beam-wise statistical modeling provides a robust and transferable solution. Combining lightweight per-beam models with spatial consistency filtering substantially improves precision while maintaining high recall and real-time capability. All datasets, annotations, and implementations are publicly released.

[57] Risk-Adaptive Edge–Cloud Visual Reasoning for Communication-Efficient Autonomous Driving cs.CVPDF

Meng Ma, Shuyang Li, Naigang Wang, Ruimin Ke

TL;DR: 本文提出了一种风险自适应的边缘-云端视觉推理架构，用于通信高效的自动驾驶。该架构通过车载交通评估来决定何时请求云端推理，结合车载视觉语言模型和轻量级检测器进行本地保守响应和选择性云端访问，在保持任务成功率的同时显著减少了云端请求和紧急制动激活。

Details

Motivation: 云端视觉语言模型虽然具有更强的上下文推理能力，但频繁上传视觉数据会增加通信开销和决策延迟，影响自动驾驶的实时性和安全性。本文旨在通过风险自适应机制，在保证决策质量的前提下，减少对云端模型的依赖和通信负担。

Result: 在CARLA仿真实验中，该方法在任务成功率上与定期云端访问方法相当，但将云端请求减少了54.1%，并记录了更少的自动紧急制动激活次数。在三种模拟网络配置下，该方法持续减少了云端流量，尽管变道耗时比定期访问更长。

Insight: 创新点在于将车载交通风险评估作为触发选择性云端推理的实用机制，实现了本地轻量检测与云端深度推理的协同。其风险自适应决策框架可借鉴用于其他需要平衡计算、通信与实时性能的边缘智能系统。

Abstract: Cloud-hosted vision-language models (VLMs) offer greater contextual reasoning capabilities than smaller onboard models, but frequent visual uploads increase communication overhead and add network and inference latency to tactical decisions. We present a risk-adaptive edge-cloud architecture in which onboard traffic assessment determines when cloud reasoning is requested. An onboard VLM and a lightweight detector capture temporal traffic conditions and path-relative hazards for conservative local response and selective cloud access. The cloud model provides tactical advice, while validation, vehicle control, and automatic emergency braking remain local. In CARLA experiments, our method matched the task success rate of periodic cloud access while reducing cloud requests by 54.1% and recording fewer automatic emergency braking (AEB) activations. In a delayed-roadwork ablation, semantic events triggered requests before the next scheduled audit. Across three emulated network profiles, the method continued to reduce cloud traffic, although lane changes took longer than with periodic access. Onboard traffic assessment therefore served as a practical trigger for selective VLM inference in these experiments.

[58] Registration-Free Hyperspectral Reconstruction from RGB via a Permutation-Invariant Gram-Matrix Principle cs.CVPDF

Jiangsan Zhao, Masayuki Hirafuji, Seishi Ninomiya, Jakob Geipel, Wei Guo

TL;DR: 该论文提出了一种无需配准和已知相机响应函数（CRF）的高光谱图像重建方法。核心是利用置换不变的格拉姆矩阵监督原理，通过匹配解混后的丰度图格拉姆矩阵来学习从高分辨率RGB图像到高光谱图像的映射，从而摆脱了对空间对齐和预定义CRF的依赖。

Details

Motivation: 传统的高光谱图像重建方法通常依赖于精确的空间配准和已知的相机响应函数，这在多传感器实际应用中难以满足。本文旨在消除这两个强假设，实现更鲁棒的重建。

Result: 在室内、自然场景和遥感基准测试中，该方法达到了与需要这些假设的方法相当的精度，并且在假设被违反时仍保持鲁棒性。在完全随机置换像素的极端条件下，现有SOTA方法失效，而本方法在逆重排后重建结果不变。

Insight: 主要创新点在于提出了置换不变的格拉姆矩阵监督原理，将重建问题转化为对材料组成（丰度）统计特性的匹配，而非像素级对应。这允许网络在无配准、无已知CRF、无成对监督的情况下进行训练。一个关键洞察是重建精度对用于匹配格拉姆矩阵的具体差异度量不敏感，表明性能主要源于该原理本身而非损失函数调优。

Abstract: Reconstructing a spatially and spectrally high-resolution hyperspectral image (HR-HSI) from a low-resolution HSI (LR-HSI) and a high-resolution RGB image (HR-RGB) usually assumes precise registration and a known camera response function (CRF). Both assumptions are difficult to satisfy with different sensors. We remove both through a permutation-invariant supervision principle: the Gram matrix of an unmixed abundance map depends on shared material composition but not on pixel ordering. Matching abundance Gram matrices therefore allows RGB-to-HSI mapping to be learned without spatial correspondence and without a predefined CRF. Under a full random permutation of HR-RGB pixels, a state-of-the-art fusion method collapses, whereas our reconstruction is unchanged after inverse reindexing for evaluation. Building on this principle, a residual spectral super-resolution function maps HR-RGB directly to HR-HSI without registration, known CRF, or paired supervision. Across indoor, natural-scene, and remote-sensing benchmarks, the method achieves accuracy comparable to approaches that require these assumptions while remaining robust when they are violated. Loss ablations further show that reconstruction accuracy is largely insensitive to the specific discrepancy used to match the Gram matrices, indicating that performance arises primarily from the permutation-invariant principle rather than loss tuning.

[59] MetaReason: Precise Interleaved Multimodal Reasoning via Editing Meta Information for Solving Geometry Problems cs.CV | cs.AI | cs.MMPDF

Penghao Yin, Haomin Wang, Qihong Tang, Xiaoye Qu, Hongjie Zhang

TL;DR: 本文提出MetaReason框架，通过编辑几何图形的元信息实现精确的多模态推理，以解决平面几何问题。该框架首先将几何图像解析为结构化元信息，利用预定义工具进行可控编辑以合成高保真视觉状态，然后基于这些增强视图进行推理。为支持该框架，作者构建了包含图像-元信息转换样本和推理轨迹的TutorGeo数据集，并通过监督微调和强化学习训练模型，在真实考试问题衍生的ExamGeo基准上取得了优于开源模型、与专有模型相当的性能。

Details

Motivation: 现有视觉语言模型在解决几何任务时过度依赖纯文本推理，而引入中间视觉状态的方法又常因几何表示不准确和渲染保真度低导致输出不可靠，因此需要一种能实现精确多模态推理的新方法。

Result: 在从真实考试问题构建的ExamGeo基准上，MetaReason显著优于现有开源模型，并与专有模型取得了竞争性性能。

Insight: 创新点在于利用结构化元信息进行可控编辑以合成高保真视觉状态，从而支持精确的辅助线构造和交错多模态推理；从客观角度看，将几何问题转化为可编辑的元信息流程，并结合大规模合成数据集进行训练，是提升几何推理可靠性的有效途径。

Abstract: Although visual reasoning is crucial for solving complex geometry tasks, existing vision-language models rely heavily on text-only reasoning. Some recent methods introduce intermediate visual states to facilitate reasoning, but they are often hindered by inaccurate geometric representations and low rendering fidelity, ultimately leading to unreliable outputs. To address these limitations, we propose MetaReason, a framework for multimodal reasoning in plane geometry that leverages structured meta-information to enable accurate auxiliary-line construction. The framework first parses geometric images into meta-information, performs controllable edits with predefined tools to synthesize high-fidelity visual states, and then conducts reasoning based on these augmented views. To support this framework, we construct TutorGeo, a comprehensive dataset containing 17k image-to-meta conversion samples, 60k text-only reasoning traces, and 60k interleaved multimodal reasoning traces. Using this dataset, we combine supervised fine-tuning and reinforcement learning to develop robust multimodal reasoning capabilities. We also introduce ExamGeo, a benchmark derived from real-world examination problems that enables systematic evaluation across varying difficulty levels. Experimental results demonstrate that MetaReason significantly outperforms existing open-source models and achieves competitive performance against proprietary models.

[60] FZ-VLM: A Two Stage Florence-Zephyr Vision Language Model Framework for Pulmonary Nodule Characterization and Clinical Decision Making cs.CV | cs.AIPDF

Pramit Dutta, Jenita Manokaran, Richa Mittal, Ryan Appleby, Eranga Ukwatta

TL;DR: 本研究提出了FZ-VLM，一个用于肺部CT结节结构化表征和临床决策的两阶段视觉语言模型框架。第一阶段使用微调的Florence-2模型从CT图像中提取放射学属性（如解剖位置、直径、边缘特征和衰减类型），第二阶段使用Zephyr-7B模型基于这些属性生成结节描述、随访建议和纵向分析。

Details

Motivation: 肺癌筛查中，放射科医生手动评估肺结节特征耗时且存在观察者间差异，现有AI方法多专注于孤立任务，缺乏统一的、基于临床的解释框架。

Result: 第一阶段模型在解剖位置、边缘特征、衰减类型分类上分别达到77.18%、67.96%、79.13%的准确率，直径估计的平均绝对误差为2.58毫米，优于评估的GPT-4基线和人类基线。第二阶段输出经放射科专家评估，准确率为93.9%，完整性为98.6%，临床相关性为76.1%，综合得分为89.5%。

Insight: 创新点在于提出了首个用于结构化结节表征和临床决策的两阶段VLM框架，将视觉特征提取与基于属性的语言推理解耦，实现了从图像到结构化属性再到临床文本报告的统一流程，为医学影像AI提供了可解释且临床可用的新范式。

Abstract: Lung cancer remains one of the leading causes of cancer-related mortality worldwide, and Computed Tomography (CT) is a primary imaging tool for screening and followup assessment. After pulmonary nodule detection, radiologists manually assess anatomical location, diameter, margin characteristics, and attenuation type to support risk assessment and clinical decision-making. However, this post-detection workflow is time-consuming and can be affected by inter-observer variability. Existing Artificial Intelligence methods often focus on isolated tasks, limiting their use as a unified, clinically grounded interpretation framework. This study presents FZ-VLM, a two-stage Florence-Zephyr Vision Language Model framework for unified structured pulmonary nodule characterization in lung CT. The framework uses a fine-tuned Florence-2 model to extract radiological attributes from expert-annotated 2D axial CT slices, while a Zephyr-7B model uses these attributes to generate nodule descriptions, follow-up recommendations, and longitudinal analyses. Results showed that the Stage 1 model achieved 77.18% accuracy for anatomical location, 67.96% accuracy for margin characteristics, and 79.13% accuracy for attenuation type, with a Mean Absolute Error of 2.58 mm for diameter estimation, outperforming evaluated GPT-4-based baselines as well as the human baseline. Expert radiologist evaluation of Stage 2 showed 93.9% accuracy, 98.6% completeness score, 76.1% clinical relevance, and an overall score of 89.5%. Safety analysis showed that most outputs were clinically safe, although some follow-up recommendations still required expert review. To the best of our knowledge, this study presents the first two-stage Vision-Language Model framework for structured nodule characterization and clinical decision-making.

[61] Frequency and Edge-Guided Segment Anything Model for Remote Sensing Image Semantic Segmentation cs.CVPDF

Feng Gao, Zizhe Pan, Haoting Wang, Ruzhuang Hua, Jingchao Cao

TL;DR: 本文提出了一种频率和边缘引导的Segment Anything模型（FE-SAM），用于遥感图像语义分割。该方法通过引入频率调制适配器（FMA）自适应地分解和调制频域特征，以增强不同地物类型的信息，并设计了EGRefiner模块来整合多尺度边缘增强信息，以改善对象边界的精细分割。

Details

Motivation: 现有基于SAM的方法在遥感图像语义分割中存在两个主要问题：一是SAM的特征对不同地物类型的多样性适应不足；二是对象边界存在语义模糊，阻碍了精确的轮廓描绘。

Result: 在三个基准数据集上的大量实验表明，FE-SAM的性能优于最先进的方法（SOTA）。

Insight: 创新点在于将频率域分析与边缘引导机制结合到基础视觉模型SAM中，通过FMA进行自适应频率调制以增强特征表示，并通过EGRefiner显式地利用边缘信息来提升边界分割的准确性，为遥感图像分割提供了可扩展且高效的框架。

Abstract: Remote sensing image semantic segmentation (RSISS) has attracted significant attention due to the growing demand for fine-grained land cover information. The Segment Anything Model (SAM), proposed as a foundation vision model, offers strong segmentation performance and generalization capabilities for RSISS tasks. However, existing SAM-based approaches face two limitations: (1) Insufficient adaptation of SAM’s features to the diverse characteristics of land cover types. (2) Semantic ambiguity at object boundaries, which hinders accurate delineation. To address these limitations, we propose Frequency and Edge-guided SAM (FE-SAM), a scalable and efficient framework for RSISS. Specifically, we introduce a Frequency-Modulated Adapter (FMA) that adaptively decomposes and modulates frequency-domain features based on the input data. It selectively enhances informative high- and low-frequency components corresponding to different land cover types. Furthermore, to improve SAM’s ability to capture fine-grained details, we design EGRefiner, which integrates multi-scale edge-enhanced information extracted from the input image. Extensive experiments on three benchmark datasets demonstrate that FE-SAM outperforms state-of-the-art methods. The source codes are available at: https://github.com/oucailab/FE-SAM.

[62] MOSS-VL Technical Report cs.CVPDF

Pengyu Wang, Chenkun Tan, Shaojun Zhou, Qirui Zhou, Yanxin Chen

TL;DR: MOSS-VL是一个开源的视觉语言模型系列，其核心创新是将实时交互能力（边说边看）作为首要设计目标。该模型通过门控交叉注意力、合成的交互语料库以及分阶段的课程学习策略来实现这一能力。离线版本在标准基准测试中表现优异，而实时版本在多个流式基准测试中领先于开源模型，尤其在测试主动行为的子集上优势显著。

Details

Motivation: 为了解决现有视觉语言模型在实时交互场景（如连续感知和生成）中的局限性，将“边说边看”作为模型的一等能力进行设计。

Result: 离线模型MOSS-VL-Instruct在可比较规模下具有竞争力，并在时序推理视频数据集上领先。实时模型MOSS-VL-Realtime在四个流式基准测试中，在三个测试上平均表现最佳（第四个排名第二），在OmniMMI主动预警子集上以66.0分大幅超越最佳基线（37.5分）。模型在视觉上下文增长时，首词生成时间优势从2.8倍扩大到5.1倍。

Insight: 创新点在于将实时交互作为核心能力进行全栈协同设计：采用门控交叉注意力实现生成时持续感知视觉输入；使用合成交互语料监督说话、沉默和修正的时机；通过分阶段课程学习，在强大的离线基础模型上仅用轻量级最终阶段集中训练实时能力。这种设计实现了高效且主动的流式交互。

Abstract: We present MOSS-VL, an open vision-language model family that treats real-time interaction – perceiving while it speaks – as a first-class capability. It is co-designed across the stack: the language decoder attends to vision only through gated cross-attention, so the model can naturally see incoming frames while generating; a synthesized interaction corpus supervises when to speak, when to stay silent, and when to revise; and a staged curriculum concentrates all real-time-specific training in one light final stage over a strong offline foundation. Offline, MOSS-VL-Instruct is competitive at comparable scale and leads temporal-reasoning video sets. Across four streaming benchmarks, MOSS-VL-Realtime posts the best average on three (second on the fourth) among open-source streaming models, sweeping the three subsets that squarely test proactive behavior – 66.0 vs. 37.5 for the best baseline on OmniMMI Proactive Alerting. With 11.3B parameters but visual tokens outside the decoded sequence, MOSS-VL widens its time-to-first-token advantage over same-backbone Qwen3-VL-8B from 2.8x to 5.1x as visual context grows. We release all five checkpoints, the training curriculum, and the real-time inference code at https://github.com/OpenMOSS/MOSS-VL.

[63] MEDR: Query-Independent Frame Selection via Multi-Signal Event Modeling and Dynamic Rescoring cs.CVPDF

Xinlei Pu, Weijie Shi, Wen Yang, Yi Cao, Hao Chen

TL;DR: 本文提出了一种名为MEDR（多信号事件建模与动态重评分）的查询无关帧选择方法，旨在为多模态大语言模型处理长视频时，在有限的视觉令牌和计算预算下，选择更具信息量的关键帧。该方法通过整合视觉、运动和文本信号来建模事件，并动态重评分以构建一个固定的、可跨不同问题重用的帧集合。

Details

Motivation: 现有方法存在局限：均匀采样可能遗漏短暂但信息丰富的内容，而查询依赖的方法虽能检索问题相关帧，但无法在不同问题间共享视觉输入，在多轮视频对话中需重复选择。因此，需要一种查询无关的帧选择方法，在保持固定视觉输入可重用性的同时，提高对信息事件的覆盖。

Result: 在标准基准测试中，MEDR在Video-MME基准上将模型准确率提升了0.63%-0.89%。在LongVideoBench的长视频子集上，使用Qwen3-VL-8B模型时准确率最高提升1.23%。同时，它为视频的每个问题重用完全相同的帧集，整体准确率仍提升0.53%。

Insight: 创新点在于提出了一种无需训练、查询无关的帧选择框架，通过多信号（视觉、运动、文本）事件建模来捕捉互补信息，并引入动态重评分机制，迭代地根据帧级信号强度、额外事件覆盖和时间邻近性来更新候选帧分数，从而构建一个高效、可重用的固定帧集。

Abstract: Frame selection is a fundamental component of multimodal large language models, enabling long videos to be processed under limited visual-token and computational budgets. Uniform sampling preserves temporal coverage but may miss informative content that appears only briefly. To alleviate this limitation, query-dependent methods can retrieve question-relevant frames. However, because the selected frames depend on the current question, the same visual input cannot be directly shared across different questions, and frame selection must be repeated in multi-turn video dialogue. This motivates us to seek a query-independent frame selection method that preserves the reusability of a fixed visual input while improving the coverage of informative events beyond uniform sampling. We propose Multi-Signal Event Modeling and Dynamic Rescoring (MEDR), a training-free and query-independent frame selection method. Multi-Signal Event Modeling organizes complementary visual, motion, and text signals into signal-specific temporal events. Dynamic Rescoring then iteratively reevaluates each candidate relative to the current selected set, updating its score according to frame-level signal strength, additional event coverage, and temporal proximity. The resulting fixed frame set is constructed without observing the query and can be reused across different questions. On the standard benchmark evaluations, MEDR improves model accuracy by 0.63%-0.89% on Video-MME. On the long-video subset of LongVideoBench, it improves accuracy by up to 1.23% with Qwen3-VL-8B. MEDR further improves overall accuracy by 0.53%, while reusing exactly the same frame set for every question about a video.

[64] Do Visual Grounding Decoders Need Feed-Forward Networks? A Controlled Study over Frozen Vision-Language Features cs.CVPDF

Tarun Tomar

TL;DR: 这篇论文研究了在视觉定位解码器中，前馈网络（FFNs）是否在预训练视觉语言模型（VLM）已经编码了图像和语言上下文后仍然提供必要的计算。通过比较仅含注意力模块的解码器（A4）、包含注意力加FFN的解码器（S4）以及参数数量匹配的仅注意力解码器（A8）在冻结VLM特征上的性能，发现仅注意力解码器在多个基准上表现相当或更优，同时显著减少了可训练参数和延迟。

Details

Motivation: 动机是探究在视觉定位任务中，当使用预训练的、特征被冻结的视觉语言模型作为编码器时，解码器中的前馈网络（FFNs）是否仍是必要的组件，旨在简化模型架构并提升效率。

Result: 在RefCOCOg和Ref-Adv-s基准上，A4匹配或略优于S4；在FineCops-Ref的IoU@0.5指标上，A4有微小劣势（0.52个百分点），但A8恢复并反超S4（0.26个百分点）。A4将可训练解码器参数减少了44.4%，缓存解码器延迟降低了10.1%。

Insight: 创新点在于通过控制实验表明，在冻结VLM特征的视觉定位解码器中，前馈网络可能不是必需的，仅使用注意力模块的简化架构可以达到相当甚至更好的性能，同时提升参数和计算效率。这为设计更高效的视觉语言模型解码器提供了新思路。

Abstract: Do feed-forward networks (FFNs) in visual grounding decoders add essential computation once a pretrained vision-language model has already encoded image and language context? We compare a four-block attention-only decoder (A4), a matched four-block attention-plus-FFN decoder (S4), and an eight-block attention-only parameter control (A8) over frozen VLM features. A4 matches or slightly exceeds S4 on RefCOCOg and Ref-Adv-s. FineCops-Ref reveals a small A4 deficit of 0.52 percentage points at IoU@0.5 (95% CI [0.12, 0.95] in favor of S4), but A8 recovers it and finishes 0.26 points above S4. Official FineCops levels do not show a monotonic increase in the gap. A4 reduces trainable decoder parameters by 44.4% and cached-decoder latency by 10.1%, although end-to-end latency remains backbone-dominated. These results concern the trainable grounding decoder, not a complete attention-only VLM.

[65] EgoTac: In-the-wild Tactile Prediction from Egocentric Vision cs.CV | cs.ROPDF

Wenkang Zhang, Chengbo Yuan, Zicheng Zhang, Zhengxue Cheng, Yang Gao

TL;DR: 本文提出了EgoTac模型，旨在从第一人称视角的人类视频中预测丰富的触觉信息。该模型在一个包含超过570万图像-触觉对的数据集上进行训练，能够预测连续的力测量值和二值化的接触信息。实验表明，EgoTac在领域内预测力误差低于0.06N，在领域外接触预测基准上优于现有方法，并能对真实世界视频进行零样本预测。

Details

Motivation: 触觉对于灵巧操作至关重要，但用于机器人学习的大规模人类第一人称数据普遍缺乏触觉信息。由于传感器限制，直接收集大规模触觉数据具有挑战性，而人类视频数据丰富、包含大量接触信息且易于扩展，因此研究能否仅从视觉信息推断触觉信号。

Result: 在领域内预测中，平均力误差低于0.06N。在领域外接触预测基准测试中，EgoTac持续优于最先进的接触估计器。模型还能捕捉真实触觉数据的上升和下降模式，并对无约束的真实世界视频实现零样本预测。缩放分析进一步表明，数据多样性和数量都能稳步提升性能。

Insight: 论文的创新点在于提出了一个通用模型，能够从丰富易得的人类第一人称视频中学习并预测触觉信号，为机器人学习提供了一种可扩展的、获取触觉先验知识的新途径。从客观角度看，其核心在于利用大规模、多样化的视觉-触觉配对数据，学习跨不同交互的细微触觉动态，从而实现了从视觉到触觉的有效迁移。

Abstract: Touch is fundamental to dexterous manipulation, yet most egocentric human data increasingly used for robot learning lacks tactile information. Directly collecting large-scale tactile data is challenging due to sensor limitations, while human video data is abundant, contact-rich, and easily scalable. This motivates a natural question: can tactile signals be inferred purely from vision? To address this, we introduce EgoTac, a generalizable model that predicts rich tactile information directly from egocentric human videos. EgoTac is trained on a unified corpus of over 5.7M image-tactile pairs, covering both continuous force measurements and binary contacts. By learning from this diverse dataset, EgoTac captures nuanced touch dynamics across varied interactions. Experiments demonstrate strong performance: in-domain prediction achieves an average force error below 0.06N. On out-of-domain contact prediction benchmarks, EgoTac consistently outperforms the state-of-the-art contact estimator. It also captures the rise and fall patterns of real tactile data and enables zero-shot predictions on unconstrained real-world videos. Scaling analyses further reveal that both data diversity and volume improve performance steadily. Overall, EgoTac provides a scalable pathway to extract tactile priors from egocentric human videos, enabling broadly applicable tactile-aware robot learning.

[66] SA-GEM: Scale-Adaptive and Geospatial Evidence-Modulated Token Pruning for Efficient Remote Sensing Large Vision-Language Models cs.CVPDF

Kexin Ma, Jing Xiao, Bowen Xing, Liang Liao, Chia-Wen Lin

TL;DR: 本文提出了一种名为SA-GEM的即插即用框架，用于高效遥感大视觉语言模型（RS-LVLMs）的令牌剪枝。该方法通过任务自适应的令牌粒度分配与整体的地理空间证据调制，解决了高分辨率处理导致令牌数量激增而重要视觉证据稀疏的问题。实验表明，SA-GEM在多个基准测试上实现了精度和效率的双重提升。

Details

Motivation: 遥感大视觉语言模型在处理高分辨率图像时，视觉令牌数量随输入分辨率呈二次增长，导致计算效率低下，而关键的视觉证据却稀疏且被稀释。现有令牌剪枝方法多采用与尺度无关的分辨率策略和孤立的重要性线索，限制了任务对齐的粒度适应性和整体证据保留。

Result: 在多个基准测试中，SA-GEM在准确性和效率上均优于现有剪枝方法。具体在XLRS-Bench上，其准确率超过GeoLLaVA-8K模型2.3%，同时实现了2.4倍的总推理加速。

Insight: 创新点在于统一了任务自适应的令牌粒度分配（通过轻量级路由器选择分辨率）与整体的地理空间令牌重要性调制（联合建模任务相关性、空间结构和局部冗余）。核心洞察是：达到足够粒度后，令牌质量比数量更重要，高分辨率并非总是有益。

Abstract: RS-LVLMs have advanced multimodal understanding of Earth observation imagery, yet their performance is fundamentally constrained by high-resolution processing, as visual token counts grow quadratically with linear input resolution while important visual evidence is inherently sparse and increasingly diluted across the expanded sequence. Existing token pruning methods largely rely on scale-agnostic resolution policies and isolated importance cues, limiting task-aligned granularity adaptation and holistic evidence preservation. To address this, we present Scale-Adaptive and Geospatial Evidence-Modulated Token Pruning (SA-GEM), a plug-and-play framework that unifies task-adaptive token granularity allocation with holistic geospatial token importance modulation. Specifically, a lightweight router selects the resolution based on query-dependent token granularity, while a token importance modulator jointly models task relevance, spatial structure, and local redundancy to preserve holistic geospatial evidence. We show that higher resolution is not universally beneficial and, once sufficient granularity is reached, token quality matters more than token quantity. Experiments across various benchmarks demonstrate that SA-GEM achieves consistent gains in both accuracy and efficiency over existing pruning methods. On XLRS-Bench, it surpasses GeoLLaVA-8K by 2.3% in accuracy with a 2.4 times total inference speedup.

Chengbo Huang, Jun-Jie Huang, Long Lan, Tianrui Liu, Xueqiong Li

TL;DR: 本文提出MODAL，一种基于模型驱动稀疏解耦和文本-图像差分滤波的新型多模态物体重识别框架。该框架通过多模态耦合稀疏编码理论，将多模态特征显式分解为单模态特定、双模态共享和三模态共享表示，实现更透明的特征解耦。此外，利用文本语义自适应抑制视觉表示中与任务无关的响应，增强判别性信息，并在不完整模态场景下通过模态感知子空间激活缓解性能下降。

Details

Motivation: 现有方法在多模态物体重识别中缺乏原则性的特征解耦和连贯的多模态集成，导致特征纠缠、跨模态冲突、判别性线索模糊以及在模态缺失条件下的分布偏移问题。

Result: 在四个数据集上的大量实验表明，MODAL实现了最先进的性能，并具有优异的透明度。

Insight: 创新点包括基于多模态耦合稀疏编码的模型驱动深度展开特征解耦模块，以及利用文本语义进行差分滤波以增强视觉表示判别性的方法；从客观角度看，该方法通过理论驱动的解耦和自适应滤波，提升了多模态融合的鲁棒性和可解释性。

Abstract: Multi-modal object re-identification (Re-ID) aims to facilitate cross-camera object retrieval in complex environments by leveraging complementary information from visual (e.g., RGB, NIR, TIR) and textual modalities. However, existing approaches often lack principled feature disentanglement and coherent multi-modal integration, leading to entangled representations that introduce cross-modal conflicts, obscure discriminative cues, and suffer distribution shift under modality-missing conditions. To tackle these challenges, we propose MODAL, a novel multi-modal object re-identification framework, grounded in coupled sparse coding theory and differential suppression principles. A core component of MODAL is a Multi-modal Feature Sparse Decoupling module, developed in a model-driven deep unrolling manner based on multi-modal coupled sparse coding. It explicitly decomposes multi-modal features into uni-modal specific, bi-modal and tri-modal shared representations, thereby achieving more transparent and effective feature disentanglement. Benefiting from the principled feature disentanglement, MODAL naturally mitigates performance degradation in incomplete-modality scenarios via a Modality-Aware Subspace Activation that selectively activates only the consistently shared subspaces. Moreover, we propose a Text-Image Differential Filtering module that leverages coarse-grained textual semantics to adaptively suppress task-irrelevant responses in the decoupled visual representations, thereby enhancing discriminative information. Extensive experiments on four datasets demonstrate that MODAL achieves state-of-the-art performance with superior transparency.

[68] Distribution-free false-alarm calibration and chance-corrected spatial evaluation for industrial anomaly detection cs.CV | cs.LGPDF

Jie Deng

TL;DR: 该论文提出了一种结合无分布上容忍阈值与配对-交叉空间检验的方法，用于工业异常检测中的误报率校准和机会校正的空间评估。该方法旨在更准确地评估检测器在选定阈值下的性能，并校正因重复缺陷位置和掩模几何形状而夸大的重叠度量。

Details

Motivation: 工业视觉检测研究中常用的AUROC和异常图与缺陷掩模重叠度指标，既未指定选定阈值下的误报率，也易受重复缺陷位置和掩模几何形状的影响而夸大性能。因此，需要一种能报告操作点性能并进行机会校正空间证据评估的方法。

Result: 在包含120个点缺陷图像的三个ISP-AD模态和三个固定数据分割上评估了三种检测器。在378个警报中，230个与匹配掩模重叠，但九个检测器-模态单元中有八个的配对与交叉率相似；仅DINOv2-ASM具有正的95%自举下界（提升0.259，区间0.159-0.347）。在独立的Magnetic Tile Defect数据集上，WRN50和ViT-B/16的补丁记忆方法分别获得0.203和0.231的提升，单侧置换检验p值均为10^{-5}。精确样本规划表明，使用150个校准正常样本，仅支持目标误报率≥1.98%的95%置信无分布声明；1%目标需要至少299个正常样本。

Insight: 创新点在于将无分布统计方法（上容忍阈值）与空间检验（配对减交叉）相结合，以校正机会重叠并量化空间证据提升。这提供了一种更稳健的评估框架，可补充传统AUROC和原始掩模重叠指标，强调报告操作点性能和机会校正空间证据的重要性。

Abstract: Studies of industrial visual inspection commonly report the area under the receiver operating characteristic curve (AUROC) and the overlap between anomaly maps and defect masks. Neither measure specifies the false-alarm rate at a selected threshold, while recurrent defect locations and mask geometry can inflate overlap. We combine a distribution-free upper tolerance threshold with a paired-minus-crossed spatial test. This test compares each detector’s score-contributing locations with the matched defect mask and with masks from other images; the difference in rates defines spatial-evidence lift relative to the empirical chance-overlap rate. We evaluate three detectors on 120 point-defect images from three ISP-AD modalities and three fixed data splits. Of 378 alarms, 230 overlap the matched mask. Paired and crossed rates are nevertheless similar in eight of nine detector–modality cells; only DINOv2–ASM has a positive 95% bootstrap lower bound (lift 0.259, 95% interval 0.159–0.347). On the independent Magnetic Tile Defect dataset, the same analysis gives lifts of 0.203 (0.169–0.236) for Wide ResNet-50 (WRN50) patch memory and 0.231 (0.202–0.262) for Vision Transformer B/16 (ViT-B/16) patch memory, with one-sided permutation $p=10^{-5}$ for both. When crossed masks are restricted to the same defect class, the lifts remain 0.185 and 0.210. Exact sample planning shows that, with 150 calibration normals, a 95%-confidence distribution-free claim is supported only for target false-positive rates of 1.98% or higher; a 1% target requires at least 299 normals. The results support reporting operating-point performance and chance-corrected spatial evidence alongside AUROC and raw mask overlap.

[69] Perspective-Invariant Attack with Enhanced Transferability of Adversarial Examples cs.CVPDF

Kaisheng Liang, Yiming Cao, Bin Xiao

TL;DR: 该论文提出了一种名为视角不变攻击（PIA）的新方法，通过引入多自由度顶点采样策略，系统地覆盖从2自由度平移到8自由度投影映射的透视变换层次，以生成几何多样化的输入变体，从而减少对抗性扰动对代理模型的过拟合，提升对抗样本的跨模型可迁移性。

Details

Motivation: 现有增强对抗样本可迁移性的输入变换方法主要依赖自由度有限的局部操作，忽略了由视点变化引起的全局透视变换，这限制了对抗样本的泛化能力。

Result: 在涉及多种DNN架构、先进防御机制以及多模态大语言模型（LLMs）的广泛实验中，PIA及其扩展PIA-Mix在基于迁移的攻击中超越了现有最先进方法。

Insight: 核心创新在于将全局几何透视变换（而非仅局部操作）系统性地引入对抗样本生成过程，以模拟真实世界视点变化，并通过PIA-Mix框架与辅助方法高效结合，这为提升对抗攻击的鲁棒性和通用性提供了新视角。

Abstract: Adversarial examples generated on a surrogate deep neural network (DNN) can often successfully fool other black-box DNN models. This cross-model transferability poses serious security threats to DNNs in practical applications. Input transformation techniques are widely used to enhance adversarial transferability by increasing the diversity of input images. However, existing methods primarily rely on local operations with limited degrees of freedom (DOF), such as block-wise shuffling and resizing, overlooking global perspective transformations that naturally arise from viewpoint changes. In this work, we propose a Perspective-Invariant Attack (PIA), which introduces a multi-DOF vertex sampling strategy that systematically covers the perspective transformation hierarchy from 2-DOF translation to 8-DOF projective mapping. By generating geometrically diverse input variations, PIA effectively reduces overfitting of adversarial perturbations to the surrogate model, thereby improving adversarial transferability. We further propose PIA-Mix, a generic extension that maintains a complementary transformation pool and efficiently combines our perspective transformation with auxiliary methods for improved transferability. Extensive experiments involving various DNN architectures, advanced defense mechanisms, and multimodal large language models (LLMs) demonstrate that PIA and PIA-Mix outperform state-of-the-art transfer-based attacks.

[70] From “What-If” to “What-Is”: Counterfactual Thinking-Inspired Semantic Alignment for Visual Brain Decoding cs.CV | cs.HCPDF

Kaitao Yan, Chi Liu, Congcong Zhu, Huajie Chen, Gengshen Wu

TL;DR: 本文提出ConceptAlign，一种受反事实思维启发的语义对齐框架，用于视觉脑解码任务。该方法通过将解码的视觉token池化并投影到冻结的文本嵌入空间，使其与真实描述对齐，同时与由LLM生成的、仅修改关键对象/属性/关系但保持场景不变的近失替代描述分离，从而学习细粒度语义边界。

Details

Motivation: 现有基于多模态表示和扩散先验的视觉脑解码方法虽然提升了重建的真实感，但生成的图像可能在对象、属性或关系上存在语义错误，而传统重建指标主要评估最终图像，可能掩盖此类语义错误。

Result: 在Natural Scenes Dataset上的实验表明，ConceptAlign在MindEye2骨干网络基础上，提升了重建指标、反事实语义区分能力和表征对齐度。消融实验、独立LLM/人工编写的替代描述以及人工评估均支持该监督方法的有效性和鲁棒性。

Insight: 创新点在于引入反事实思维，通过构造场景保持但关键语义元素变化的近失替代描述来定义细粒度语义边界，从而在解码表征层面进行对齐和约束，而非仅优化最终图像。这提供了一种系统性的语义评估和监督框架。

Abstract: Visual brain decoding reconstructs visual content perceived by a person from neural measurements such as fMRI, providing a computational approach to studying how visual information is represented in the brain. Recent multimodal representations and diffusion priors have improved reconstruction realism. However, visually plausible reconstructions may contain incorrect objects, attributes, or relations because a strong generative prior can complete content not sufficiently specified by the decoded representation. Conventional reconstruction metrics mainly assess the final image and may therefore obscure such semantic errors. We propose ConceptAlign, a counterfactual semantic alignment framework for visual brain decoding. ConceptAlign pools decoded visual tokens and projects them into a frozen text-embedding space, aligning the representation with the ground-truth caption while separating it from scene-preserving near-miss alternatives. Generated offline by an LLM, these alternatives modify one critical object, attribute, or relation while retaining the scene. A margin-based objective learns fine-grained semantic boundaries between the observed stimulus and plausible but incorrect interpretations without requiring LLM calls during inference. We introduce a systematic three-level semantic evaluation framework covering foundational discriminability, counterfactual description discrimination, and representational geometry. Experiments on the Natural Scenes Dataset show that ConceptAlign improves reconstruction measures, counterfactual semantic discrimination, and representational alignment over the MindEye2 backbone. Matched negative-source ablations, independent LLM and human-written alternatives, and human evaluation support the effectiveness and robustness of the supervision, with favorable patterns in fine-grained conflicts, limited-data decoding, and cross-subject structure.

[71] Beyond Natural-Image Foundation Models: Benchmarking Satellite Pretraining for Ophthalmic Image Analysis cs.CVPDF

Lovre Antonio Budimir, Mingya Alexa Gong, Alyssa Foong Quinney, Ivana Matovinović, Yukun Zhou

TL;DR: 本文提出将卫星图像作为医学视觉基础模型（MedVFMs）的新型预训练领域，以缓解医学图像数据稀缺和隐私限制的问题。通过在多种眼科成像模态上对比卫星图像预训练的DINOv3-SAT493m与自然图像预训练的DINOv3-LVD1689m及医学专用基线模型，发现卫星图像预训练在眼科任务上表现更优，尤其在富含血管的正面成像模态中，甚至能在无医学数据情况下匹配或超越医学专用模型。

Details

Motivation: 医学视觉基础模型受限于数据稀缺、隐私问题和开发成本高，而现有方法依赖自然图像预训练导致与医学数据的分布偏移。卫星图像因其视觉特征与医学数据更接近且无隐私限制，被提议作为替代预训练领域。

Result: 在多个眼科成像任务中，基于4.93亿卫星图像预训练的DINOv3-SAT493m优于基于17亿自然图像预训练的DINOv3-LVD1689m，并在高分辨率正面输入任务上匹配或超越了医学专用基线模型DINOv3-RETFound和MAE-RETFound。

Insight: 创新点在于首次系统性地将卫星图像作为医学视觉基础模型的预训练源，利用其与医学图像的视觉相似性（如血管结构）缓解数据分布偏移问题；客观来看，这为医学AI开发提供了可扩展、无隐私顾虑的替代数据源，降低了领域适应成本。

Abstract: Vision Foundation Models (VFMs) have emerged as a promising approach in medical imaging, producing broadly applicable systems that can be efficiently adapted across diverse imaging modalities, anatomical regions, and clinical tasks. However, VFMs require extensive training data, and their progress in medical image analysis is constrained by limited data availability, privacy concerns, and high development costs. To alleviate these constraints, medical VFMs (MedVFMs) are often built upon weights from generalist models pretrained on vast amounts of publicly available natural images, introducing a substantial distribution shift for medical task adaptation. To address this, we propose satellite imagery as a novel pretraining domain for MedVFM development and benchmarking, motivated by its closer visual alignment with medical data and its freedom from the privacy constraints that limit medical datasets. Across multiple ophthalmic imaging modalities, we compare DINOv3-SAT493m pretrained on 493 million satellite images against DINOv3-LVD1689m pretrained on 1.7 billion natural images, together with two medical specialist baselines: DINOv3-RETFound and MAE-RETFound. Our experiments show that satellite imagery is a stronger pretraining source than natural images for ophthalmic tasks, particularly on en face vascular-rich modalities. On several tasks, satellite pretraining matches or exceeds the medical specialists on high-resolution en face inputs, despite using no medical data.

[72] Anchor-Regularized Adaptation for Generalizable AI-Generated Image Detection with DINOv3 cs.CVPDF

Hyeongjun Choi, Juhun Lee, Davide Cozzolino, Luisa Verdoliva, Simon S. Woo

TL;DR: 本文提出了一种名为锚点正则化适应（ARA）的方法，用于提升AI生成图像检测的泛化能力。该方法通过在微调过程中结合冻结的DINOv3表示和低秩适应（LoRA），有效利用像素级伪影线索，同时避免破坏预训练表示的结构，从而在多个基准测试中达到最先进的性能。

Details

Motivation: 现有AI生成图像检测方法依赖训练数据对齐来消除虚假相关性以提升泛化，但研究发现冻结的DINOv3表示即使在数据未对齐时也能表现优异，这揭示了其决策基于真实图像空间的忠实特征，但难以捕捉对齐数据强调的像素级伪影线索。

Result: ARA方法在九个多样且具有挑战性的基准测试中实现了最先进的性能，表明其能够有效结合未对齐和对齐数据的互补监督，提升检测效果。

Insight: 创新点在于提出锚点正则化适应（ARA），通过冻结锚点分类器防止预训练表示失真，并利用低秩适应捕获像素级伪影，从而平衡泛化能力和对细微伪影的敏感性，为多数据源监督提供了新思路。

Abstract: Recent works in AI-generated image detection have shown that careful training data alignment can improve generalization by removing spurious correlations. However, linear probes on frozen DINOv3 representations achieve remarkably strong performance even when trained on misaligned datasets. Motivated by this result, we analyze the underlying rationale and the limits of this generalization. We find that frozen DINOv3 performs well because its decisions rely on features that faithfully represent the space of authentic images. At the same time, its final layer is less effective at capturing the subtle pixel-artifact cues that can be emphasized by aligned training data. We further observe that naively mixing aligned and misaligned data during adaptation improves sensitivity to such cues but at the cost of distorting the pre-trained representation, limiting generalization. To address this issue, we propose Anchor-Regularized Adaptation (ARA). We apply Low-Rank Adaptation to capture pixel-level artifacts while leveraging a frozen anchor classifier to avoid deviations from the original representation structure. This allows the model to exploit pixel-artifact cues without sacrificing generalization. Our method achieves state-of-the-art performance on nine diverse and challenging benchmarks, indicating that ARA enables complementary supervision from misaligned and aligned data for more effective detection.

[73] PersonaDrive: Controllable Trajectory Prediction with Multi-Dimensional Driving Personas cs.CVPDF

Chan Lee, Kimin Yun, Yuseok Bae, Seong Tae Kim, Jung Uk Kim

TL;DR: 本文提出了PersonaDrive框架和Persona-Conditioned Trajectory (PCT)数据集，旨在解决轨迹预测和自动驾驶中缺乏有意义可控性的问题。PCT数据集将驾驶行为模式（Persona）分解为时间紧迫性和乘坐舒适性两个维度，每个维度有三个等级，共形成九种行为模式，并配有自然语言描述和轨迹。PersonaDrive框架能够从语言中学习驾驶行为模式，并生成特定于该模式的轨迹，其核心组件包括Persona-Conditioned Anchor Transform (PCAT)和Persona-Conditioned Multi-Modal Fusion (PCMF)。

Details

Motivation: 现有轨迹预测和端到端自动驾驶方法在城市场景中鲁棒性有所提升，但缺乏有意义的可控性。现有基准要么没有提供行为模式条件标注，要么只支持单一的紧急程度谱（如紧急、正常、放松），无法区分共享相同紧急程度但需要不同驾驶动态的行为模式。

Result: 实验结果表明，PersonaDrive在多维场景中持续优于所比较的基线方法。

Insight: 创新点在于提出了一个二维（时间紧迫性和乘坐舒适性）分解的驾驶行为模式框架和对应的PCT数据集，使得轨迹预测具有更细粒度和有意义的可控性。技术上的创新包括PCAT模块对锚点的分层重塑、PCMF模块的BEV级行为模式融合，以及用于训练的分层引导损失和轴分解多样性损失，以防止对角线模式崩溃。

Abstract: Although recent trajectory prediction and end-to-end autonomous driving methods improve robustness in urban environments, they still lack meaningful controllability. Existing benchmarks either provide no persona-conditioned annotations or support only a single urgency spectrum (i.e., emergency, normal, relaxed), which cannot distinguish personas that share the same urgency level but require different driving dynamics. To address this, we propose (i) the Persona-Conditioned Trajectory (PCT) dataset, which decomposes driving personas along two axes, Temporal Urgency and Ride Comfort, and combines three levels of each to form a grid of nine personas, each paired with natural-language descriptions and trajectories, and (ii) PersonaDrive, a framework that can learn driving personas from language and can generate persona-specific trajectories. PersonaDrive incorporates Persona-Conditioned Anchor Transform (PCAT), which hierarchically reshapes anchors along both axes, and Persona-Conditioned Multi-Modal Fusion (PCMF) for BEV-level persona fusion. Training is supervised by a Hierarchical Guide Loss enforcing axis-aligned physical orderings and an Axis-Decomposed Diversity Loss preventing diagonal mode collapse. Experimental results show that PersonaDrive consistently improves over the compared baselines across multi-dimensional scenarios. The code and PCT dataset are available at https://github.com/VisualAIKHU/PersonaDrive

[74] DCA-MoE: Spatially Adaptive Cross-Layer Fusion and Density-Routed Experts for Crowd Counting cs.CV | cs.IRPDF

Hao Wang

TL;DR: 论文提出DCA-MoE框架，用于人群计数任务，旨在解决视角、头部尺度、遮挡和背景杂乱等严重变化下的可靠局部密度估计问题。该框架采用冻结的DINOv3编码器，并引入了空间自适应层融合（SALF）和密度路由多感受野专家（DR-MoE）模块，使特征融合和感受野选择依赖于内容。

Details

Motivation: 现有许多多级解码器在特征融合时使用空间不变的方式，并对所有位置应用单一感受野模式，这限制了在复杂场景下恢复可靠局部密度的能力。论文旨在通过内容依赖的空间自适应决策来改进这一问题。

Result: 在NWPU-Crowd验证集上，基于DINOv3 ViT-L/16的最强配置获得了31.7 MAE和72.2 RMSE；匹配的ViT-B/16完整模型获得了32.2 MAE和75.9 RMSE。跨数据集结果仍有混合表现，部分组件基线目前报告了单次种子的独立选择最小值。

Insight: 创新点包括空间自适应层融合（SALF）和密度路由多感受野专家（DR-MoE），使融合权重和专家路由（局部、中程、大上下文）依赖于位置内容。客观来看，这种将MoE（专家混合）思想与空间自适应路由结合用于密度估计是新颖的，且框架在保持编码器冻结的情况下实现了性能提升。

Abstract: Crowd counting must recover reliable local density under severe variations in perspective, head scale, occlusion, and background clutter. Although modern counting objectives provide strong spatial supervision, many multi-level decoders still use spatially invariant feature fusion and apply one receptive-field pattern to every location. We propose DCA-MoE, a framework that makes both decisions content dependent while retaining a frozen DINOv3 encoder. Spatially Adaptive Layer Fusion (SALF) predicts position-wise weights over four aligned backbone features, and Density-Routed Multi-Receptive-Field Experts (DR-MoE) assigns each location a soft mixture of local, mid-range, and large-context residual experts. An EBC-style head reconstructs block density, while DMCount supervision and an auxiliary routing-balance term train the decoder without updating the backbone. On the NWPU-Crowd validation split, the strongest paired configuration, based on DINOv3 ViT-L/16, obtains 31.7 MAE and 72.2 RMSE; the matched ViT-B/16 full model obtains a paired 32.2/75.9. Cross-dataset results remain mixed, and several component baselines currently report independently selected minima from a single seed. The evidence therefore supports the feasibility of spatially adaptive fusion and routing, while broader paired and multi-seed evaluation remains necessary for causal attribution.

[75] UC-VLM: Consistency-Driven Learning for AI-Generated Image Detection with Vision-Language Large Models cs.CVPDF

Lei Tan, Shuwei Li, Mohan Kankanhalli, Robby T. Tan

TL;DR: 本文提出了UC-VLM，一个用于AI生成图像检测的统一多阶段框架。该框架仅依赖二值监督，通过自动识别有效指令变体，并复用相同的真实性标签，在一个多阶段训练框架中同时优化视觉判别目标和标签条件生成目标，从而增强对低级视觉取证线索的敏感性并监督文本输出。

Details

Motivation: 现有基于视觉语言大模型的AI生成图像检测器主要微调语言侧，对低级视觉取证线索关注有限，且常依赖人工设计的提示或标注的推理依据，限制了可扩展性。

Result: 在GenImage数据集上达到96.1%的平均准确率，超出先前最佳结果4.6%；在Chameleon数据集上，分别以ProGAN和SDV1.4训练时，准确率达到69.6%和77.9%，分别超过最佳基线11.2%和15.3%。

Insight: 创新点在于提出了一个统一的多阶段二值监督框架，将弱二值监督信号复用于视觉适应和标签条件文本生成，同时利用自动优化的指令减少对提示的敏感性，无需人工标注的推理依据或手工提示。

Abstract: Vision-Language Large Models (VLLMs) are promising for AI-generated image (AIGI) detection because they can produce both a prediction and a natural-language output. However, most existing VLLM-based detectors primarily fine-tune the language side while giving limited attention to low-level visual forensic cues. They also often depend on manually crafted prompts or human-annotated rationales, which limits scalability.We present UC-VLM, a unified multi-stage framework for AIGI detection that relies solely on binary supervision. UC-VLM first identifies effective instruction variants automatically. It then reuses the same binary label within a multi-stage training framework: (i) a visual discrimination objective that strengthens sensitivity to non-semantic forensic cues, and (ii) a label-conditioned generation objective that uses the binary label to supervise textual outputs. This design turns weak binary supervision into a shared supervision signal for both the visual pathway and the language output. Our key novelty is a unified multi-stage binary-supervised framework that consistently reuses the same authenticity labels for visual adaptation and label-conditioned text generation, while leveraging automatically optimized instructions to reduce prompt sensitivity without requiring human-written rationales or hand-crafted prompts.Experiments show that UC-VLM achieves 96.1% average accuracy on GenImage, exceeding the strongest prior result by 4.6%, and obtains 69.6% / 77.9% accuracy on Chameleon under ProGAN / SDV1.4 training, surpassing the best baseline by 11.2% / 15.3%, respectively.

[76] UAV Video Deblurring via Motion-Aware Diffusion: A Path to Robust Target Detection cs.CV | cs.AI | cs.ROPDF

Zhiqiang Hu, Shouren Huang, Masatoshi Ishikawa

TL;DR: 本文提出一种面向无人机视频的运动感知扩散去模糊方法，旨在提升目标检测的鲁棒性。该方法通过自适应潜在尺度选择器动态调整潜在空间分辨率以平衡细节保留与计算效率，并引入多帧对齐与可学习门控模块来融合相关时序信息、抑制未对齐特征，从而有效恢复清晰细节。

Details

Motivation: 无人机在灾害响应、交通监控等场景中至关重要，但其航拍视频常因快速机动、振动和相机平移而产生严重运动模糊，严重影响下游目标检测等任务性能。本文旨在探索一种计算高效且有效的视频去模糊方法，以增强无人机目标检测能力。

Result: 在真实无人机基准测试上的大量实验表明，该方法不仅取得了优异的去模糊性能，还显著提升了目标检测精度，使其高度适用于鲁棒的空中视觉任务。

Insight: 创新点包括自适应潜在尺度选择器（根据运动强度动态调整分辨率以平衡效率与细节）和多帧对齐与可学习门控模块（通过扭曲和门控前帧来融合相关时序信息、抑制无关特征），这些设计兼顾了计算效率与时空一致性，为无人机视频处理提供了可借鉴的解决方案。

Abstract: Unmanned Aerial Vehicles (UAVs) play a crucial role in various scenarios ranging from disaster response to traffic surveillance. However, aerial video footage often suffers from severe motion blur due to rapid flight maneuvers, vibrations, and camera panning, which can significantly degrade downstream tasks such as target detection. Our goal is to explore a computationally-efficient and effective video deblurring approach to enhance UAV target detection performance. To reduce computational cost, we first propose an Adaptive Latent Scale Selector that dynamically adjusts the latent space resolution according to the intensity of UAV motion, thus balancing detail preservation with inference efficiency. To ensure temporal consistency, we introduce a Multi-Frame Alignment and Learnable Gating module to warp and gate the preceding frames, allowing the model to fuse only relevant temporal information and suppress misaligned or uninformative features. Our method can effectively recover sharp details from the UAV video stream. Extensive experiments on real UAV benchmarks demonstrate that our method not only yields superior deblurring performance but also significantly boosts target detection accuracy, making it highly applicable to robust aerial vision tasks.

[77] Boundary-Aligned Contribution Routing for Robust Optical–SAR Object Detection cs.CVPDF

Haifa Zhang, Yijing Wang, Haoyu Wang, Zheng Li, Zhiqiang Zuo

TL;DR: 本文提出了一种边界对齐的贡献路由方法，用于提升光学与合成孔径雷达（SAR）融合在遥感目标检测中的鲁棒性。该方法通过任务驱动的贡献路由，在特征融合前动态调节各模态的贡献，以应对多模态数据间的不完美对应关系，从而抑制负向跨模态迁移。

Details

Motivation: 光学和SAR图像融合虽能互补，但空间、时间和语义上的不完美对应可能导致负向跨模态迁移，损害检测性能。论文旨在从任务效用的角度出发，仅利用检测监督学习条件化的贡献路由，以解决这一问题。

Result: 在M4-SAR和SpaceNet6-OTD数据集上的实验表明，该方法在完整输入下将mAP50提升了0.5-5.9点；在模态缺失情况下，相比基线提升了7.6-41.6点mAP50，并将负迁移率降低了最多12.7个百分点。学习到的路由权重与留一模态效用之间的Spearman相关系数达0.45-0.66。

Insight: 创新点在于提出了无需显式效用监督、质量标签或重构的贡献路由机制，包括针对浅层交互的Feature Router和针对双骨干网络的Dual-Statistic Semantic Router，实现了任务条件化的模态贡献调节，有效提升了融合系统的鲁棒性。

Abstract: Optical imagery provides rich appearance cues, whereas synthetic aperture radar (SAR) offers observations that are less sensitive to illumination and weather, making optical–SAR fusion attractive for remote-sensing object detection. However, the presence of multiple modalities does not guarantee beneficial fusion: imperfect spatial, temporal, and semantic correspondence can make an otherwise intact stream conditionally harmful and induce negative cross-modal transfer. We handle this issue through a model-specific task-utility perspective and learn task-conditioned contribution routing using detection supervision alone. The proposed fusion-boundary-aligned routing regulates each modality’s contribution before the first learned cross-modal feature-value mixing operation. For architectures with frequent shallow interaction, a Feature Router performs cross-conditioned, group-addressable modulation near the input; for dual-backbone architectures, a Dual-Statistic Semantic Router predicts stream-level contribution weights from modality-specific average and maximum statistics before late semantic fusion. The routers require no explicit utility supervision, quality labels, reconstruction, or distillation. Experiments on M4-SAR and SpaceNet6-OTD cover nominal full inputs, controlled correspondence shifts, missing modalities, and four nonzero modality-corruption scenarios. Across the reported clean-training controls, routing improves full-input $\text{mAP}{50}$ by 0.5–5.9 points. Relative to the corresponding modality-dropout baselines, it raises missing-modality $\text{mAP}{50}$ by 7.6–41.6 points and reduces the negative-transfer rate by up to 12.7 percentage points. Spearman correlations between the learned routing weights and model-specific leave-one-modality-out utility range from 0.45 to 0.66, supporting the task-utility interpretation of the routing coefficients.

[78] VGGT-Align: Bridging Local Reconstruction and Global Consistency for Long-Sequence 3D Reconstruction cs.CV | cs.AIPDF

Wei Zhang, Yihang Wu, Songhua Li, Qi Wang

TL;DR: 本文提出VGGT-Align，一种用于长序列3D重建的尺度一致性增强框架。该框架通过场景几何不变锚定（SGIA）模块，利用结构化环境中固有的几何不变量来约束分块间的尺度漂移，将7自由度Sim(3)对齐退化为6自由度刚体变换，从而在源头切断尺度误差的链式传播。此外，还引入了一种轻量级测试时自适应策略，通过多目标自监督微调归一化层参数，以渐进提升序列内分块预测的质量。

Details

Motivation: 解决长序列3D重建中全局几何一致性保持的核心挑战，特别是分块式推理流程中，序列Sim(3)对齐的尺度自由度不受约束，导致估计误差累积并扭曲全局轨迹和点云几何的问题。

Result: 在多个长序列基准测试上实现了最先进的性能，绝对轨迹误差降低了高达32%，并在轨迹稳定性和重建质量上取得了显著提升。

Insight: 核心创新在于提出了场景几何不变锚定（SGIA），利用环境规律性产生的几何不变量（如平面法向量、平行线方向）作为跨分块的尺度约束，这是一种独立于点云配准的、从源头解决尺度漂移的新颖方法。同时，轻量级测试时自适应策略提供了一种无需离线重训练即可渐进优化模型的有效途径。

Abstract: Maintaining global geometric consistency is a central challenge in long-sequence 3D reconstruction, with scale drift being the most critical failure mode. In chunk-based inference pipelines, the scale degree of freedom in sequential Sim(3) alignment is left unconstrained, causing estimation errors to compound multiplicatively and distort global trajectories and point cloud geometry. We present a scale-consistency enhancement framework built on a key insight: in structured environments such as driving scenes, geometric quantities arising from environmental regularity remain inherently invariant across temporal segments, and discrepancies in their per-chunk measurements directly expose inter-chunk scale drift. We propose Scene Geometric Invariant Anchoring (SGIA), which extracts dominant geometric invariants from each chunk’s predicted point cloud via coarse-to-fine robust estimation and exploits their cross-chunk consistency to establish scale constraints independent of point cloud registration, explicitly degenerating 7-DoF Sim(3) alignment into 6-DoF rigid-body transformation and severing chain-wise scale error propagation at its source. We further introduce a lightweight test-time adaptation strategy that fine-tunes only normalization-layer parameters via multi-objective self-supervision, progressively improving intra-chunk predictions along the sequence. Both modules are plug-and-play and require no offline retraining. Experiments on multiple long-sequence benchmarks demonstrate state-of-the-art performance, reducing absolute trajectory error by up to 32% with significant gains in trajectory stability and reconstruction quality. Code: https://github.com/WZ-CS/VGGT-Align

[79] TinyDETR-Pose: Towards End-to-End Real-Time Single-Stage 6DoF Object Pose Estimation with Lightweight Transformers cs.CVPDF

Paul Julius Kühn, Duc Anh Nguyen, Saptarshi Neil Sinha, Michael Weinmann, Arjan Kuijper

TL;DR: 本文提出了TinyDETR-Pose，一个轻量级、端到端的单阶段框架，用于在资源受限的硬件上进行实时6自由度物体姿态估计。该框架基于高效的LW-DETR架构，将检测和姿态估计构建为一个集合预测问题，通过附加的MLP头直接回归旋转、单目深度和投影物体中心，无需PnP、NMS或迭代姿态优化。

Details

Motivation: 解决在资源受限硬件上实现实时6DoF物体姿态估计的挑战，因为现有基于精确对应关系和优化的方法通常依赖不可微的PnP/RANSAC阶段或昂贵的迭代优化，而近期基于基础模型的方法推理成本过高，不适合边缘部署。

Result: 在YCB-V数据集上，TinyDETR-Pose取得了85.9的ADD-S AUC，与其他基于DETR的单阶段姿态估计方法相比，参数减少了高达72.7%。在NVIDIA Jetson Nano上使用TensorRT实现了约4.5毫秒每帧的推理延迟，达到实时性能。

Insight: 创新点在于将检测与姿态估计统一为端到端的集合预测问题，避免了PnP、NMS等后处理步骤；采用统一的ADD-S损失处理物体对称性，无需针对不同物体的特定损失调度；设计了基于类别和2D空间线索的对称安全匈牙利匹配器，以稳定处理对称性和深度模糊性。其轻量化设计证明了基于Transformer的6D姿态估计在边缘部署的实用性。

Abstract: Real-time 6DoF object pose estimation on resource-constrained hardware remains challenging, as accurate correspondence-based and refinement pipelines typically rely on non-differentiable PnP/RANSAC stages or costly iterative refinement, while recent foundation-model-based approaches incur inference costs that are prohibitive for edge deployment. We present TinyDETR-Pose, a lightweight, end-to-end, single-stage framework that jointly detects objects and regresses their full 6D pose in a single forward pass. Built on the efficient LW-DETR architecture, TinyDETR-Pose formulates detection and pose estimation as a set-prediction problem and attaches dedicated MLP heads for rotation, monocular depth, and projected object center regression to each decoder query, eliminating the need for PnP, NMS (non-maximum suppression), or iterative pose refinement. Object symmetries are handled through a ADD-S loss applied uniformly to all objects, without the need for object-specific loss schedules or separate geodesic/ADD supervision. In addition, predictions are assigned to ground truth using a symmetry-safe Hungarian matcher based on class and 2D spatial cues, yielding stable assignment under symmetry and depth ambiguity. On YCB-V, TinyDETR-Pose achieves a comparable ADD-S AUC of 85.9, while requiring up to 72.7% fewer parameters than other DETR-based single-stage pose-estimation approaches. Due to its compact design, TinyDETR-Pose runs in real time and achieves an inference latency of only ~4.5 ms per frame on an NVIDIA Jetson Nano using TensorRT, demonstrating that accurate end-to-end transformer-based 6D pose estimation can be made practical for edge deployment.

[80] SAGE-OR: Semi-supervised Adaptive Scene Graph Generation for Operating Rooms cs.CVPDF

Brandon Leblanc, Charalambos Poullis

TL;DR: SAGE-OR是一个用于手术室的半监督自适应场景图生成框架，它采用解耦的特征表示与关系推理范式，利用冻结的基础模型进行隐式定位，并通过轻量级图变换器在缓存特征上进行关系推理，无需定位监督。

Details

Motivation: 现有手术场景图生成方法依赖密集的多模态监督和专用硬件，导致数据集构建成本高昂且仅限于模拟环境，本文旨在消除定位监督，实现更实用、可适应新手术环境的方法。

Result: 在4D-OR基准测试中，核心模型达到76% F1分数，与全监督基线相当；通过无监督手部增强提升至86%，仅比需要密集多模态监督的SOTA方法低4个百分点。

Insight: 创新点在于用解耦的特征表示-推理范式替代传统检测-推理范式，利用通用分割提示实现半监督学习和无监督上下文增强，并通过轻量级图变换器实现高效推理，适合手术室边缘硬件部署。

Abstract: Current surgical scene graph generation methods depend on dense multi-modal supervision and specialized hardware (synchronized RGB-D sensors, calibration rigs), making dataset construction expensive and restricting all existing benchmarks to simulated environments. We propose SAGE-OR, a feature-centric framework that replaces the traditional detect-then-reason paradigm with a decoupled representation-reasoning paradigm in which localization is derived from frozen foundation models, encoded implicitly in pre-computed features, and used without any localization supervision, while a lightweight graph transformer performs relational reasoning over cached features. We employ a semi-supervised formulation with general-purpose segmentation prompts to eliminate localization supervision while enabling unsupervised context augmentation through additional prompt-driven entities, such as hands, which are absent from annotations. General-purpose prompts are used to induce near-perfect recall, while precision is delegated to downstream attention-based reasoning, enabling simple adaptation to new entities via prompt-level modification. This design enables a lightweight 15M-parameter graph transformer that trains in 1.4 hours and runs relational inference at $\sim$1ms per frame with peak memory under 2GB, suitable for edge hardware used in the operating room; feature extraction runs offline as a separate caching stage (4.27s per frame). On the 4D-OR benchmark, the core model achieves 76% F1, matching the fully supervised 4D-OR baseline while eliminating all localization annotations, and unsupervised hand augmentation raises this to 86%, within 4 points of state-of-the-art (SOTA) methods requiring dense multi-modal supervision, providing a practical pathway for adaptation to new surgical settings without annotation other than relationship and class labels.

[81] A Multi-Annotator Study of Segmentation Noise and Uncertainty in Turbid Underwater Images cs.CVPDF

Galadrielle Humblot-Renaux, Vasiliki Ismiroglou, Malte Pedersen

TL;DR: 本文首次对真实水下场景中的分割任务进行了系统的多标注者研究，涉及超过100名参与者，并控制了不同的浑浊度水平。研究发现水下数据集面临与其他视觉任务类似的标注挑战，而浑浊度会引入额外的系统性误差。研究还探讨了导致标签噪声的主要因素，并探索了在浑浊水下环境中提高标注质量的方法，包括特权信息、个体努力和标注者集成。

Details

Motivation: 水下图像由于需要领域专业知识、能见度条件差以及难以在不可访问环境中建立可靠真实值，特别容易受到标签不确定性和标注者分歧的影响，但这一问题在水下图像领域尚未得到充分探索。

Result: 研究通过多标注者实验发现，浑浊度会引入系统性误差，并识别了影响标注质量的关键因素，如特权信息和标注者努力程度。

Insight: 创新点在于首次系统研究水下图像分割中的标注不确定性问题，并提供了多标注者数据集和标注策略（如标注者集成）以改善标注质量，为水下视觉任务的可靠评估提供了新视角。

Abstract: Label uncertainty and annotator disagreement are common challenges in the field of computer vision, yet their study has largely been confined to the medical domain or to generic image-recognition datasets. Underwater datasets are particularly susceptible to these issues due to the need for domain expertise, degraded visibility conditions, and the inherent difficulty of establishing reliable ground truth in inaccessible environments. Despite these challenges, annotation uncertainty in underwater imagery remains largely unexplored. In this work, we present the first systematic multi-annotator study of segmentation in real underwater scenes, with over 100 participants, and across varying, controlled levels of turbidity. We show that underwater datasets face many of the same annotation challenges as other vision tasks, while turbidity introduces additional systematic errors. We further investigate the main factors driving label noise and explore ways to improve annotation quality in turbid underwater environments, including privileged information, individual effort and annotator ensembles. All (meta-) data collected in this study will be available on the project page: https://vap.aau.dk/tubcertainty

[82] Decomposing Whole Slide Image Report Generation with Graph-Constrained Multiple Instance Learning Workflows cs.CVPDF

Antony Gitau, Martyna Borak, Bjørn-Jostein Singstad, Martin Paulson, Karl Thomas Hjelmervik

TL;DR: 该论文提出了一种用于全切片图像（WSI）病理报告生成的分解框架，该框架将报告生成任务分解为视觉识别、结构化推理和语言生成三个子任务。具体而言，它利用冻结的Virchow2图块嵌入，通过多实例学习（MIL）分类头回答器官特异性诊断问题，然后使用器官条件图将这些答案约束组装成结构化推理链，最后由语言模型生成病理报告。

Details

Motivation: 解决现有端到端视觉-文本模型在WSI报告生成中混淆了视觉识别、结构化推理和语言生成各自的贡献与失败模式的问题，旨在提高生成过程的透明度、可控性和可解释性。

Result: 在REG2026测试集（2028张切片）上，所提工作流的链式Jaccard分数达到0.702。消融实验表明，移除基于图的链构建、使用单一通用图或让语言模型自由构建链，性能分别下降至0.420、0.398和0.371。使用相同的报告生成器，图结构化的推理链将报告分数从0.330提升至0.495。在未微调的350张外部TCGA WSI上，正确器官图被选中的概率为64.0%，排名前三的概率为86.6%；提供正确器官图后，与TCGA粗粒度原发诊断标签的一致性从61.8%提升至92.6%。

Insight: 主要创新点在于将复杂的WSI报告生成任务分解为可解释的模块化工作流，并引入了器官条件图来约束和引导从视觉特征到诊断推理链的结构化组装。这不仅能提升最终报告的质量，还允许对生成过程中的特定阶段（如器官路由）进行错误定位和分析，增强了模型的透明度和鲁棒性，特别是在面对领域偏移时。

Abstract: Whole-slide image (WSI) report generation requires recognizing spatially distributed pathological features and organizing them into a coherent diagnostic narrative. Although direct vision-to-text models can yield fluent reports, they obscure the contributions and failure modes of visual recognition, structured reasoning, and language generation. We propose a decomposed framework in which frozen Virchow2 tile embeddings are aggregated by multiple-instance learning (MIL) classification heads that answer organ-specific diagnostic questions. An organ-conditioned graph constrains the assembly of these answers into a structured reasoning chain, which a language model realizes as a pathology report. On the REG2026 held-out set of 2,028 slides, the proposed workflow achieved a chain-Jaccard score of 0.702. Performance fell to 0.420 without graph-based chain construction, 0.398 when the organ-specific graphs were replaced by a single organ-agnostic graph, and 0.371 when the language model constructed the chain freely from MIL predictions. Using the same report generator, graph-structured chains improved the report score from 0.330 to 0.495. On 350 external TCGA WSIs spanning the seven REG organs without fine-tuning, the expected organ graph was selected in 64.0% of cases and ranked among the top three in 86.6%. Providing the correct organ graph increased agreement with coarse TCGA primary-diagnosis labels from 61.8% to 92.6%, identifying organ routing as a main bottleneck under domain shift. Overall, organ-conditioned, graph-constrained chain assembly improves structured reasoning and report generation while enabling stage-specific error localization.

[83] CBX-Bench: A Human-Aligned MLLM Council for Benchmarking Concept Bottleneck Model Explanations cs.CVPDF

Yusuf Meric Karadag, Gulay Oklan, Seref Baris Cagliyan, Umut Ozdemir, Emre Akbas

TL;DR: 本文提出了CBX-Bench，一个用于评估概念瓶颈模型（CBM）解释质量的基准测试和排行榜。为了解决CBM解释缺乏定量评估标准的问题，作者构建了一个由多模态大语言模型（MLLM）组成的“委员会”，该委员会能够为图像及其CBM解释生成质量评分。该委员会的评估能力通过一项包含2700个人类判断的研究进行了验证和校准。

Details

Motivation: 概念瓶颈模型的核心动机是提供可解释性，但目前主要仍通过下游分类准确率和孤立的定性示例进行评估，缺乏大规模、定量的评估方法。这主要是因为大规模获取真实概念标注不可行，且概念列表缺乏共识。

Result: 基于人类研究建立的参考标准，由五个开源MLLM组成的委员会恢复了超过70%的严格人类偏好排序，在人类标注者完全一致的条目上，这一比例上升到83%。基于此，作者发布了公开的基准测试CBX-Bench，用于对CBM解释进行数据集级别的质量评分和排名。

Insight: 主要创新点在于提出了一个利用MLLM委员会来量化评估CBM解释质量的方法，并通过人类研究进行校准，从而创建了一个与人类对齐、可扩展的评估基准。这为超越准确率和定性示例的模型可解释性评估提供了新思路和工具。

Abstract: Concept Bottleneck Models (CBMs) are designed to make visual classification interpretable by expressing predictions through human-understandable concepts. Although interpretability is the central motivation for CBMs, they are still largely evaluated as predictive models by downstream classification accuracy, supplemented by isolated qualitative examples. This highlights a pressing need for quantitative measures, a challenge complicated by the infeasibility of ground-truth concept annotation at scale and the open nature of concept lists due to a lack of consensus. To fill this gap, we develop a multimodal large language model (MLLM) council that, given an image and its CBM explanation, produces an explanation quality score. To ground and validate the council, we first conduct a human study to establish a ground-truth reference for CBM explanation quality: for an image, annotators compare explanations from two of LF-CBM, VLG-CBM, and CBM-Suite and choose the more useful one, or mark them as equally good or equally bad, yielding 2700 judgments over 900 image-comparison items on CUB-200, ImageNet-100, and Places365. Against this human reference, our five-model council, consisting of open-weight MLLMs, recovers over 70% of strict human preference rankings, rising to 83% on items where human annotators unanimously agree. Building on this validated council, we introduce CBX-Bench, a public benchmark and leaderboard: authors of new CBMs can submit their model’s explanations, and CBX-Bench scores them with the council and maintains dataset-level rankings of explanation quality. CBX-Bench thus provides a human-aligned, scalable evaluation of CBM explanations beyond accuracy and isolated qualitative examples. The benchmark is available at https://github.com/meric-karadag/cbx-bench.

[84] NumerosityVLM: A Cognitively Inspired Benchmark for Interpreting Numerosity Representations in Vision-Language Models cs.CV | cs.AIPDF

Yiming Fu, Fangjun Li, Xiujin Liu, Ruidong Ma, Hang Yu

TL;DR: 该论文提出了一个名为NumerosityVLM的认知启发式诊断基准，用于评估视觉语言模型（VLMs）对数量感知（numerosity）的表征能力。该基准包含10,800张合成图像，通过正交操控物体大小、空间排列和数量，并逐步消融纹理、形状和颜色，以分离数量感知与其他视觉因素的纠缠。在零样本设置下评估了七个VLMs，发现模型架构是性能差异的主要解释因素，且数量信号在视觉编码器的早期阶段就已线性可分，而性能差异主要与语言模型组件相关。

Details

Motivation: 现有计数基准将数量感知与相关的视觉因素（如物体大小、排列）纠缠在一起，导致当前视觉语言模型（VLMs）在人类婴儿在语言习得前就已具备的数量感知这一认知能力上表现不佳，缺乏深入理解。

Result: 在提出的NumerosityVLM基准上对七个VLMs进行零样本评估，多因素分析显示模型架构解释了最大的性能方差比例（部分ω²=0.325），远超视觉条件的影响。分层探测进一步表明，线性可分的数量信号在视觉编码器的早期阶段一致出现。

Insight: 论文的创新点在于设计了一个认知启发的、严格控制变量的诊断基准，能够正交解耦数量感知与其他视觉因素，从而更纯粹地评估VLMs的核心数量表征能力。客观分析认为，其方法揭示了模型架构（尤其是语言模型组件）对数量感知任务的关键影响，以及数量信息在视觉编码器中的早期编码特性，这为理解VLMs的认知能力提供了新的诊断工具和见解。

Abstract: Vision-language models (VLMs) achieve strong performance on high-level multimodal tasks, yet numerosity perception, a cognitive ability that emerges in human infants before language acquisition, remains poorly understood in current models, as existing counting benchmarks entangle numerosity with correlated visual factors. We introduce a cognitively inspired diagnostic benchmark, NumerosityVLM, comprising 10,800 synthetic images across six controlled conditions. The benchmark orthogonally manipulates object size, spatial arrangement, and numerosity, while progressively ablating texture, shape, and color. Evaluating seven VLMs in a zero-shot setting, multi-factor analysis reveals that model architecture explains the largest proportion of performance variance (partial $ω^{2}=0.325$), far exceeding visual conditions. Layer-wise probing further shows that linearly separable numerosity signals consistently emerge at early stages of the vision encoder, while performance differences across evaluated models are primarily associated with the language model component. Code and data are publicly available at https://github.com/fuy3/NumerosityVLM-Benchmark, and https://huggingface.co/datasets/fuy3/NumerosityVLM.

[85] AlignJEPA: Predictive Vision-Language Alignment for Remote Sensing Foundation Models cs.CVPDF

Md Aminur Hossain, Omkumar Vaghasiya, Rajeev Ranjan Dwivedi, Vinod Kurmi, Biplab Banerjee

TL;DR: 本文提出AlignJEPA，一种受JEPA启发的预测性视觉-语言对齐框架，用于遥感基础模型。该方法通过仅训练一个轻量级的预测对齐网络，利用掩码视觉令牌预测遥感文本嵌入，并结合语义预测与双向对比检索进行训练，为地球观测基础模型与语言的对齐提供了一条参数高效的路径。

Details

Motivation: 现有遥感基础模型与自然语言的弱对齐限制了自然语言存档搜索、图文检索和问题条件分析等应用，因此需要一种高效的视觉-语言对齐方法。

Result: 在BigEarthNet.txt数据集上训练和评估，用于自然语言Sentinel检索；在RSICD上进行跨数据集适应性评估；使用RSVQA作为闭集表示探针。

Insight: 创新点在于采用预测性对齐而非仅依赖全局图像-文本对比对齐，通过掩码感知多尺度预测对齐器聚合细粒度、区域和全局的可见令牌，并使用交叉尺度Transformer联合建模，实现了参数高效的遥感视觉-语言对齐。

Abstract: Remote sensing (RS) foundation models provide transferable Earth observation representations across sensors, resolutions, and geographies, yet most remain weakly aligned with natural language, limiting natural-language archive search, image-text retrieval, and question-conditioned analysis. We propose AlignJEPA, a JEPA-inspired predictive vision-language alignment framework for remote sensing foundation models. AlignJEPA uses a pretrained AnySat visual encoder and a RemoteCLIP text encoder while training only a lightweight predictive alignment network. Instead of relying on global image–text contrastive alignment alone, the framework predicts remote-sensing text embeddings from masked visual foundation-model tokens. Its mask-aware multi-scale predictive aligner aggregates visible tokens at fine, regional, and global scales, jointly models them with a cross-scale Transformer, and projects the resulting representation into the text space using learned query pooling. Training combines semantic prediction with bidirectional contrastive retrieval. We train and evaluate AlignJEPA on BigEarthNet.txt for natural-language Sentinel retrieval, evaluate cross-dataset adaptation on RSICD, and use RSVQA only as a closed-set representation probe. AlignJEPA provides a parameter-efficient route for aligning Earth observation foundation models with language.

Shengchuan Gao, Teng Hu, Bohao Feng, Luchen Li, Wenqiang Wang

TL;DR: 本文提出了一种同步感知的加速框架，用于高效的音视频生成。该方法通过分析音视频跨模态注意力中结构化交互模式，引入保护性稀疏注意力策略，在保持关键同步令牌高保真计算的同时稀疏化冗余注意力交互，从而提升推理效率并维持音视频质量与同步性。

Details

Motivation: 现有音视频生成模型在统一扩散过程中合成同步视频和音频，但长视频令牌序列导致跨去噪步骤的重复注意力计算，推理成本高昂。现有为视频生成设计的加速技术（如量化、注意力稀疏化）直接应用于音视频模型会忽视模态间交互，可能破坏音视频同步。

Result: 该方法在保持视频质量、音频质量和音视频同步性的同时，显著提高了推理效率。

Insight: 创新点在于观察到双向音视频交叉注意力揭示了模态间的结构化交互模式（高响应集中于少数声音相关的视觉和时序区域），并据此设计了同步感知的保护性稀疏注意力策略，在加速过程中显式考虑跨模态依赖性。

Abstract: Recent audio-visual generation models can synthesize synchronized video and sound in a unified diffusion process, but their inference cost remains high because long video token sequences require repeated attention computation across denoising steps.A variety of acceleration techniques have been developed for video generation models, including low-bit quantization, attention sparsification, and feature caching.However, since these methods are originally designed for video generation, directly applying them to audio-visual models overlooks the interactions between the audio and video branches and may therefore disrupt audio-video synchronization.We present a synchronization-aware acceleration framework for efficient audio-visual generation.Our key observation is that bidirectional audio-video cross-attention reveals structured interactions between the two branches, with high responses often concentrated on a few sound-related visual and temporal regions.Guided by this interaction pattern, we introduce a protected sparse attention strategy that preserves high-fidelity computation for synchronization-critical tokens while sparsifying redundant attention interactions.By explicitly accounting for cross-modal dependence during acceleration, our method improves inference efficiency while keeping video quality, audio quality, and audio-video synchronization.

[87] GLaQ: Grounding Latent Queries in Visual Evidence for Multimodal Reasoning cs.CVPDF

Zesheng Yang, Lingling Zhang, Xinyu Zhang, Cheng Zhang, Pengyu Li

TL;DR: 论文提出了GLaQ（Grounded Latent-Query）框架，旨在解决多模态大语言模型在思维链推理中难以保留和重用细粒度视觉证据的问题。该方法用一组基于原始视觉标记的、上下文条件化的固定查询，替代了自回归的潜在状态序列，从而在答案生成阶段直接、协调地访问源视觉证据。

Details

Motivation: 现有基于思维链的多模态模型在文本推理步骤中难以保留细粒度视觉证据。外部工具增强方法需要预定义工具和额外推理时间处理，而内部连续视觉潜在推理方法因自回归结构导致潜在状态信息重复，无法有效捕获互补的视觉细节。

Result: 在五个细粒度视觉理解和感知基准测试中，GLaQ-7B模型相比其基础模型提升了5.99%到9.66%，并且在所有比较的视觉潜在方法中处于领先地位，表明其方法达到了SOTA水平。

Insight: 核心创新在于提出了一个基于视觉标记的、固定的、上下文条件化的潜在查询框架，替代了自回归的潜在状态展开。这种方法允许在推理过程中直接、协调地访问原始视觉证据，无需外部视觉操作或自回归潜在展开，从而更有效地恢复图像中的局部化证据。

Abstract: Chain-of-thought reasoning has substantially improved the problem-solving capabilities of multimodal large language models. Fine-grained visual evidence, however, remains difficult to preserve and reuse across text-based reasoning steps. To address this limitation, tool-augmented thinking-with-images methods maintain visual access externally by revisiting or manipulating the image, but require predefined tools and additional inference-time processing. As an internal alternative, continuous visual latent reasoning retains intermediate computation in hidden states. However, its prevailing autoregressive construction makes each latent state depend on its predecessors, so later states may repeat information already present in the latent sequence rather than capture complementary visual details. We introduce GLaQ, a grounded latent-query framework that replaces sequential latent rollout with a fixed set of context-conditioned queries grounded in the original visual tokens. The grounded queries are reinjected for answer generation, providing direct and coordinated access to source visual evidence. We train GLaQ with localized-view supervision followed by reinforcement learning under task-level rewards. Across five benchmarks for fine-grained visual understanding and perception, GLaQ-7B gains 5.99–9.66% over its base model and leads all compared visual latent methods, suggesting that direct query-to-image grounding can recover localized evidence from the full image without external visual operations or autoregressive latent rollouts.

[88] CrossView: Can Vision-Language Models Reason Across Cameras? cs.CVPDF

Sahil Shah, S P Sharan, Harsh Goel, Manvik Pasula, Adithya Hebbalae

TL;DR: 该论文提出了CrossView，一个针对多摄像头视频问答的基准测试，涵盖自动驾驶、安防监控、第一人称/第三人称视频和机器人等多个领域。研究发现，现有模型（包括GPT-5.2和Qwen3-VL）在该基准上表现不佳，尤其是在需要跨视角联合推理的任务上，凸显了多摄像头推理与单摄像头问题的本质区别。

Details

Motivation: 现实世界（如自动驾驶、安防系统）依赖于多摄像头网络，而现有视频理解基准和模型主要针对单摄像头设置，缺乏系统评估模型跨多视角进行上下文整合、遮挡解析和证据融合等关键推理能力的基准。

Result: 评估显示，专有模型（如GPT-5.2）和开源模型（如Qwen3-VL）在CrossView基准上的准确率均较低，且开源模型表现显著落后；模型性能与其联合处理多视角的能力强相关，这确立了CrossView作为多摄像头视频推理的严格基准地位。

Insight: 论文的创新点在于首次系统性地定义了多摄像头视频推理的独特挑战（如上下文扩展、遮挡解析、视角重要性判断），并构建了相应的基准测试；从客观角度看，该研究揭示了当前视觉语言模型在跨摄像头联合推理能力上的重大局限，为未来模型设计提供了关键的评估方向。

Abstract: Video understanding benchmarks have long centered on single-camera settings, where modern multi-modal language models achieve strong performance across image and video tasks. Yet, the real world runs on multi-camera networks: autonomous vehicles, security systems, and robots all gather data across many simultaneous views. We argue that this is not simply “more” of the single-camera problem; it is fundamentally different. Multi-camera reasoning requires handling context that scales with the number of views, resolving occlusions visible from only a subset of cameras, judging which views matter, and integrating evidence across perspectives that may overlap or diverge. Current models struggle with exactly these challenges, yet no benchmark systematically targets them. We introduce CrossView, a multi-camera video question-answering benchmark spanning autonomous driving, security surveillance, egocentric/exocentric video, and robotics. Evaluation of proprietary models, such as GPT-5.2, and open-source models, like Qwen3-VL, reveals consistently low accuracy, with open-source models trailing by a wide margin. Performance scales strongly with a model’s ability to jointly process multiple viewpoints, positioning CrossView as a rigorous benchmark for multi-camera video. We open-source our code and dataset at https://utaustin-swarmlab.github.io/CrossView.

[89] RigidBench: Evaluating Rigid-Body Physics in Video Generation Models cs.CV | cs.LGPDF

Swarnim Jain, Shangzhe Wu

TL;DR: 本文提出了RigidBench，一个基于物理模拟器的基准测试，用于评估视频生成模型在刚体物理运动预测方面的能力。该基准包含五个任务，提供逐帧的掩码、深度、6自由度轨迹和接触信息，用于分离评估运动、几何、身份、背景稳定性和视觉相似性等多个方面。作者评估了八个模型，发现不同指标的排名差异很大，并利用基准的训练数据对Wan 2.2 TI2V-5B模型进行了微调和分析。

Details

Motivation: 当前视频生成模型的评估指标（如SSIM）通常将运动、几何、身份等不同方面的错误混合在一起，无法有效衡量预测物体是否按正确的物理规律运动。因此，需要一种能够分离评估这些不同方面的基准测试。

Result: 在RigidBench上评估了八个视频生成模型，结果显示没有模型在所有十项测量指标上都领先，且模型间的平均SSIM与3D轨迹误差呈强正相关（r=0.89）。对Wan 2.2 TI2V-5B进行全微调后，其3D轨迹误差降低了约20%，而SSIM几乎不变。

Insight: 创新点在于提出了一个基于模拟器、提供精确物理状态（如6-DoF轨迹）的基准测试，能够分离评估视频生成的多个关键方面。客观分析表明，该基准不仅可用于模型排名，其附带的高质量训练数据还能有效用于模型微调和内部表征分析（如通过教师强制探测和针对性干预揭示物体位置信息在扩散Transformer中的表示和使用）。

Abstract: Video models are increasingly used to predict what happens next in a scene, yet the metrics commonly used to compare their outputs say little about whether the predicted objects move correctly. Motion, geometry, identity, background stability, and visual similarity can fail independently, but whole-frame scores often mix these errors together. We introduce RigidBench, a simulator-grounded benchmark that compares a generated continuation with a reference rollout from the same initial frame and motion description. Its five rigid-body tasks vary objects, materials, viewpoints, and indoor and outdoor scenes, with per-frame masks, depth, 6-DoF trajectories, and contacts available for scoring. We evaluate eight models on the same 100 examples with ten measurements that keep these aspects separate. The resulting rankings depend strongly on what is measured: no model leads on all ten, and across model means, higher SSIM accompanies larger 3D trajectory error (r = 0.89). RigidBench also includes 5,000 training videos with exact simulator state, which we use to fine-tune and analyze Wan 2.2 TI2V-5B. Full fine-tuning reduces 3D trajectory error by about 20% with almost no change in SSIM, while teacher-forced probes and targeted interventions show that object position is represented throughout Wan’s diffusion transformer and used by its denoising computation.

[90] AlloEgo-VLM: Disambiguating Allocentric and Egocentric Reference Frames in Vision-Language Models cs.CVPDF

Kuan-Lin Chen, Tzu-Ti Wei, Chao-Chi Liao, Yu-Chee Tseng, Jen-Jee Chen

TL;DR: 本文研究了视觉语言模型（VLMs）在理解空间语义时面临的歧义性挑战，并提出了AlloEgo-VLM框架来解决此问题。作者构建了包含以自我为中心和以他者为中心视角的新数据集AlloEgo-View，并基于此开发了能够消解参考系歧义的框架，该框架可通过监督微轻松集成到现有VLMs中，并在机器人仿真平台上验证了其在实际物体搜索任务中的可行性。

Details

Motivation: 自然语言描述空间关系时常省略明确的参考系，导致语义歧义，这可能对具身AI机器人造成严重错误。现有VLMs由于在参考系和物体方向性上训练不足，常产生不一致的响应。

Result: 实验突显了当前VLMs在处理视角特定查询时的局限性，并证明了AlloEgo-VLM强大的消歧能力。

Insight: 创新点在于构建了结构化标注视角特定空间关系的数据集AlloEgo-View，并提出了一个可集成到现有VLMs中的框架，以明确区分自我中心和他者中心参考系，从而提升模型的空间语义理解鲁棒性。

Abstract: This study investigates the challenge of ambiguity faced by Vision-Language Models (VLMs) in understanding spatial semantics. Spatial cognition, shaped by cognitive psychology, spatial science, and cultural context, often assigns directionality to objects. However, natural language descriptions of spatial relations frequently omit explicit reference frames, leading to semantic ambiguity and potentially serious errors for embodied AI robots. Existing VLMs, due to insufficient training on reference frames and object orientations, often produce inconsistent responses. To address this issue, we construct a new dataset, AlloEgo-View, comprising (image, query, view-specific answer) triplets that capture key object relations from both allocentric and egocentric perspectives. The view-specific descriptions follow a structured spatial representation that annotate detailed scene descriptions, reference and target objects, their orientations, reference frames, and view types. Building on AlloEgo-View, we develop AlloEgo-VLM, a framework to disambiguate allocentric and egocentric reference frames, even under ambiguous queries, and to be easily integrated into existing VLMs via supervised fine-tuning. Furthermore, we deploy our framework onto an embodied robotic platform within NVIDIA Isaac Sim to validate its real-world feasibility in open-ended object searching tasks. Experiments highlight the limitations of current VLMs in handling view-specific queries and demonstrate the strong disambiguation ability of AlloEgo-VLM.

[91] Catching Hallucinated Citations in Video-LLM Question Answering: A Self-Verification Pipeline and Verifier Ablation Study cs.CV | cs.AIPDF

Yogesh Kumar

TL;DR: 本文针对视频问答系统中视觉语言模型常产生虚假时间戳引用的幻觉问题，提出了一种自验证流水线。该流水线通过检索增强语言模型生成带时间戳的答案，并独立验证每个引用帧，以检测不支持的虚假声明。研究比较了不同验证设计，发现使用小型自然语言推理模型能稳定捕获79%的对抗性虚假声明。

Details

Motivation: 解决视频问答系统中视觉语言模型因时间戳引用而导致的欺骗性幻觉问题，即模型高置信度生成未得到视频帧支持的虚假声明，从而误导用户信任。

Result: 在对抗性虚假前提问题上，基于小型自然语言推理模型的验证器能稳定捕获79%的伪造声明，且不影响真实声明；而直接询问视觉模型的方法完全失败（捕获率0%），基于重描述和通用LLM法官的方法则不稳定（捕获率在0%到100%间波动）。

Insight: 创新点在于提出闭环自验证流水线以检测视频问答中的幻觉引用，并通过消融实验揭示：直接视觉模型询问因奉承性而失效，而小型专用NLI模型能提供稳定且可解释的验证，优于通用LLM法官。

Abstract: Video question answering systems built on vision-language models often produce timestamped claims with high confidence even when unsupported by the cited frame. This deceptive hallucination arises because timestamps imply grounding without ensuring correctness, increasing user trust but not accuracy. We introduce a pipeline that closes this loop. A retrieval-augmented language model drafts answers with per-claim timestamp citations, and each cited frame is independently re-examined before being shown to the user. We compare against a plain baseline and ablate three verification designs, evaluated on both Apple Silicon (MLX) and Google Colab (HF Transformers, CUDA). Directly asking the vision model whether a frame supports a claim fails completely (0% catch rate on 40 claims) due to sycophancy. Blind re-captioning plus a general LLM judge improves results but is unstable, oscillating between 0% and 100% flagged depending on prompt phrasing. Replacing that judge with a small natural language inference model yields a stable, interpretable verifier that catches 79% of fabricated claims on adversarial false-premise questions while leaving true claims untouched. We release the full pipeline, evaluation harness, and implementations for both Apple Silicon and Colab. Code is available at https://github.com/yogesh-iitj/grounded-video-qa.

[92] EgoGazeLite: On-Device Egocentric Gaze Prediction for Token-Efficient Multimodal LLM Video Input cs.CV | cs.AIPDF

Matteo Stoiber, Niels Buus Lassen

TL;DR: 本文提出EgoGazeLite，一种轻量级双过程注视预测器，用于解决在消费级智能眼镜等可穿戴设备上，因缺乏专用眼动追踪硬件而无法实现基于注视的视频裁剪压缩的问题。该方法能在移动设备上实时运行，通过预测的注视区域裁剪视频，大幅减少多模态大语言模型处理视频所需的视觉令牌数量，同时保持下游视频描述任务的质量。

Details

Motivation: 动机在于，现有基于注视的视频压缩方法（如GazeLLM）依赖专用眼动追踪硬件，这在消费级智能眼镜上不可用。因此，需要构建一个仅依赖软件的、既准确又足够轻量的注视预测器，以在移动设备的算力和功耗预算内实现令牌高效的多模态大语言模型视频输入。

Result: 在两个多模态大语言模型、三个自动化指标和两个LLM评判者的评估下，使用预测注视区域裁剪的视频与使用真实注视区域裁剪的视频在描述质量上无显著差异，在全部十种测试情况下均得到等效性确认。EgoGazeLite模型参数量为1570万，计算量为6.71 GFLOPs，在消费级加速硬件上能以实时速度（21.6毫秒/帧）端到端运行完整的注视预测与裁剪流程。

Insight: 创新点在于设计了一个轻量级、双过程的软件注视预测器，成功移除了对专用眼动追踪硬件的依赖，实现了在移动设备上实时、令牌高效的、基于注视的自我中心视频理解。从客观角度看，其核心创新是将高精度的注视预测任务成功轻量化，并验证了软件预测与硬件测量在最终任务效果上的等效性，为边缘设备上的高效多模态视频处理提供了可行的技术路径。

Abstract: The use of multimodal LLMs (MLLMs) for egocentric video understanding with wearable devices is constrained by the token budget. Memory and compute cost scale with the number of visual tokens, and high-resolution video quickly becomes expensive to transmit and process at scale. Prior work (GazeLLM) addresses this by cropping the video around the camera wearer’s gaze. This reduces the number of visual tokens by about tenfold while maintaining or improving the quality of full-resolution descriptions. However, this compression strategy depends on dedicated eye-tracking hardware, which is unavailable on consumer smart glasses. Building a software-only substitute poses a joint constraint: the predictor must be accurate enough to preserve downstream description quality, yet light enough to run on-device, within the power and compute budget of a smartphone. We address this with EgoGazeLite, a lightweight dual-process gaze predictor for egocentric video. Across two MLLMs, three automated metrics, and two LLM judges, predicted-gaze crops show no significant difference from ground-truth-gaze crops. Equivalence is confirmed in all ten cases. EgoGazeLite achieves this at 15.7M parameters, 6.71 GFLOPs, and runs the full gaze-and-crop pipeline end-to-end in real time (21.6 ms/frame) on consumer accelerator hardware. Together, these results remove the need for eye-tracking hardware for token-efficient, gaze-conditioned egocentric video understanding with MLLMs.

[93] Scalable Black-Box Model Attribution for Images cs.CVPDF

Asaf Livne, Amir Jevnisek, Shai Avidan

TL;DR: 本文提出了一种名为RPA（Raw-Patch Attribution）的轻量级CNN方法，用于解决生成模型的归属问题，即在仅给定图像的情况下确定其由哪个生成模型产生。该方法在严格的黑盒设置下运行，具有数据效率高、计算成本与候选模型数量无关、对图像压缩、模糊和缩放等真实世界操作保持鲁棒性等优点。

Details

Motivation: 随着生成模型的快速扩散，模型归属问题日益突出：仅凭一张图像，能否确定其由哪个模型生成？现有方法通常假设更复杂的生成器需要更复杂的归属器，但本文挑战了这一假设。

Result: RPA在25类的DRAGON数据集上达到98.0%的准确率，在27类的OpenFake数据集上达到92.9%的准确率，超越了先前工作，能够以更高的准确率归属更多模型。

Insight: 创新点在于证明了简单的轻量级CNN（RPA）在严格黑盒设置下即可高效解决模型归属问题，无需复杂设计。训练得到的特征提取器具有多功能性：同一表示可以无监督地恢复模型谱系、标记和分组未见过的生成器，并通过少量样本适应新模型而无需重新训练。

Abstract: The rapid proliferation of generative models raises the model attribution problem: given only an image, can we determine which model produced it? Existing methods have grown as elaborate as the generators they target, on the as- sumption that a more sophisticated model demands a more sophisticated attributor. We show it does not. RPA (Raw- Patch Attribution) attributes images in the strictest black- box setting with a lightweight CNN. Despite its simplicity, it attributes more models at higher accuracy than prior work, reaching 98.0% on 25-class DRAGON and 92.9% on 27- class OpenFake; it is data-efficient and runs at a cost inde- pendent of the number of candidate models; and it stays ro- bust to the compression, blur, and resizing images undergo in the wild. Training for closed-set attribution yields a ver- satile feature extractor: the same representation recovers model lineage without supervision, flags and groups unseen generators, and admits new models through few-shot adap- tation rather than retraining.

Shuaishuai Cao, Meng Tang, Shuwei Peng, Xuan Liu, Min Huang

TL;DR: 本文提出了HAFR-Net，一种用于超高分辨率遥感图像语义分割的渐进式细化框架。该网络通过异质性引导的阶段自适应融合、频率残差适配器和混淆感知三先验解码器，自适应地组织和保守地细化预训练分层编码器的多阶段特征表示，而不是用单一的解码器变换替换它们。

Details

Motivation: 当前利用预训练分层编码器进行VHR遥感图像分割时面临挑战：邻近区域需要在细节和语义上下文之间取得不同平衡，激进的任务特定变换会扰动有用的预训练特征，且传统的语义监督提供的结构指导有限。

Result: 在匹配的Swin-B训练和单尺度推理协议下，HAFR-Net在ISPRS Vaihingen、ISPRS Potsdam、LoveDA和OpenEarthMap数据集上分别达到了84.12%、87.86%、55.17%和67.70%的mIoU，相比匹配的UPerNet基线分别提升了0.55、0.95、1.55和1.84个百分点，实现了SOTA性能。

Insight: 创新点在于提出了一个保守的细化框架，核心是HG-SAF模块根据局部特征变化预测密集的阶段权重进行自适应融合，FRA模块通过有界、零初始化的残差分支注入频率信息以保持参考表示，以及CATP解码器利用边界、对象性和类关系先验进行正则化。这提供了超越仅基于内容路由的、更一致的空间重加权能力，并改善了边界和细长结构的精度。

Abstract: Semantic segmentation of very-high-resolution (VHR) remote sensing imagery increasingly benefits from strong pretrained hierarchical encoders, yet exploiting their multi-stage representations remains difficult. Nearby regions demand different balances between fine detail and semantic context, aggressive task-specific transformations perturb useful pretrained features, and conventional semantic supervision provides limited structural guidance. We present HAFR-Net, a progressive refinement framework that adaptively organizes and conservatively refines hierarchical representations instead of replacing them with a monolithic decoder transformation. Heterogeneity-Guided Stage-Adaptive Fusion (HG-SAF) predicts dense stage weights conditioned on local feature variation. A Frequency-Residual Adapter (FRA) then injects frequency information through a bounded, zero-initialized residual branch that keeps the fused representation as its reference. A Confusion-Aware Tri-Prior Decoder (CATP) finally regularizes the prediction with boundary, objectness, and training-derived class-relation cues. Under a matched Swin-B training and single-scale inference protocol, HAFR-Net attains 84.12%, 87.86%, 55.17%, and 67.70% mIoU on ISPRS Vaihingen, ISPRS Potsdam, LoveDA, and OpenEarthMap, improving the matched UPerNet baseline by 0.55, 0.95, 1.55, and 1.84 percentage points, respectively. Controlled analyses further show consistent spatial reweighting beyond content-only routing, improved boundary and thin-structure accuracy over matched spatial and spectral alternatives, and reduced confusion on pre-declared class pairs.

[95] Gaussian-JEPA: Joint-Embedding Predictive Learning for 3D Gaussian Splats cs.CVPDF

Bin Ren, Qi Ma, Yue Li, Zongyan Han, Yidi Li

TL;DR: 本文提出Gaussian-JEPA，一种用于3D高斯溅射（3DGS）表示的联合嵌入预测学习方法。该方法通过预测被遮挡的高斯令牌块的表示，而非直接重建其属性，旨在学习可重用且对输入变化鲁棒的3D高斯特征表示。

Details

Motivation: 现有基于3DGS的自监督方法主要重建被掩码的高斯属性，这使监督依赖于特定的采样实现，且需要输入空间解码器。本文旨在探索一种潜在的预测目标，以学习对高斯令牌（具有耦合属性和异构空间支持）更有效的表示。

Result: 在多种任务上评估，包括高斯重采样、部分观测和可渲染形状补全，以及向部件分割和对象分类的迁移。相比匹配的重建预训练，Gaussian-JEPA在重采样输入下更一致，在部分观测下保留更多实例信息，并为高斯补全提供更强的冻结特征。

Insight: 核心创新在于将联合嵌入预测架构（JEPA）应用于3D高斯表示，通过在线编码器和EMA编码器生成多尺度目标，并结合互补的目标投影和特征空间接地进行潜在监督，避免了直接重建高斯属性，从而学习到更鲁棒和可迁移的表示。

Abstract: 3D Gaussian Splatting (3DGS) represents 3D content with anisotropic primitives that jointly encode geometry and appearance. Fixed-budget encoders consume sampled observations of Gaussian assets, so the same object may be observed through different primitive realizations. Existing self-supervised methods mainly reconstruct masked Gaussian attributes, tying supervision to one sampled realization and requiring an input-space decoder. Latent prediction offers an alternative, but its application to Gaussian tokens requires targets that accommodate coupled attributes and heterogeneous spatial support. We introduce Gaussian-JEPA, which predicts representations of held-out Gaussian token blocks from visible context. An online encoder processes the context, while a shared exponential-moving-average encoder supplies stop-gradient features for multi-scale targets. Complementary target projections and feature-space grounding provide latent supervision without reconstructing Gaussian attributes. We evaluate the features under Gaussian resampling, partial observations, and renderable shape completion, together with transfer to part segmentation and object classification. Compared with matched reconstruction pretraining, Gaussian-JEPA is more consistent across resampled inputs, retains more instance information under partial observations, and provides stronger frozen features for Gaussian completion. These results support latent prediction as an effective objective for reusable 3D Gaussian representations. Code is on the project page (https://amazingren.github.io/Gaussian-JEPA/).

[96] WorldRover: A Scalable Synthetic Video Data Engine for World Exploration with Rich Annotations cs.CV | cs.GRPDF

Xiaojie Xu, Zhengyuan Lin, Runyi Li, Yihao Liu, Kaipeng Zhang

TL;DR: WorldRover是一个可扩展的合成视频数据引擎，用于生成具有丰富注释（如深度、相机轨迹、光流等）的世界探索视频。其核心是Unreal Engine管线，能够离线渲染分钟级路线，并支持第一人称、第三人称和360度全景视角在不同环境状态下的重放。基于此引擎构建的WorldRover-10M数据集为学习可探索世界的连贯表示提供了监督信号。

Details

Motivation: 学习生成或重建可探索世界需要超越RGB的视频数据，包括相机运动、场景几何、时序对应关系以及控制信号。真实捕获难以提供密集几何和长程对应关系，而现有合成资源很少能在同一帧中结合这些信息并支持可控的视点和外观变化。

Result: 论文构建了WorldRover-10M数据集，其序列为RGB图像配对提供了度量深度、相机轨迹和轨迹衍生的动作信号。第三人称子集还提供了密集光流、具有可见性的长程2D/3D点轨迹以及独立于相机轨迹的角色轨迹。

Insight: 主要创新点在于提出了一个可扩展的合成数据生成引擎，能够系统性地为长时程世界探索任务生成多视角、多模态的丰富注释数据，将世界探索转化为一个可扩展的数据生成问题，为构建和维护可探索世界的连贯表示模型提供了关键监督。

Abstract: Learning to generate or reconstruct explorable worlds requires video paired with more than RGB: camera motion, scene geometry, temporal correspondence and, for interactive models, control signals. Real capture can provide some of these signals, but dense geometry and long-range correspondence usually rely on estimation or specialised instrumentation. Rendering provides these quantities directly, yet existing synthetic resources rarely combine them on the same frames while also supporting controlled changes of viewpoint and appearance. We introduce WorldRover, a data engine for generating richly annotated, long-range explorations of artist-built environments. At its core, WorldRover-Engine is an Unreal Engine pipeline that executes and offline-renders minute-scale routes while preserving their full trajectories and scene geometry. The same exploration can be replayed from first-person, third-person, and 360 panoramic cameras under different environmental states. Using WorldRover-Engine, we construct WorldRover-10M, whose sequences pair RGB with metric depth, camera trajectories, and trajectory-derived action signals throughout each exploration. Third-person subsets additionally provide dense optical flow, long-range 2D/3D point tracks with visibility, and a character trajectory distinct from the camera trajectory. The engine can render a traversal from first-person, third-person and 360 panoramic viewpoints, under different environmental states or with a neutral white material, while preserving the route and scene geometry. WorldRover therefore turns long-horizon world exploration into a scalable data-generation problem, providing supervision for models that must build, maintain, and revisit coherent representations of an explorable world.

[97] Counterfactual Sensitivity Is Not Repairability: Auditing Replay Probes for Video Evidence cs.CVPDF

Rama AlHamidi, Rasul Khanbayov, Erchin Serpedin, Hasan Kurban

TL;DR: 本文提出CARVE，一种黑盒反事实探测方法，用于评估视频问答代理对检索视觉证据的依赖程度。该方法通过比较代理在原始帧被语义破坏（DESTROY）与模拟重放（SHAM）下的答案变化，量化证据依赖度。在LVBench基准测试中，CARVE成功筛选出部分问题，将代理准确率提升3.26个百分点。

Details

Motivation: 视频问答代理在回答前会检索视觉证据，但最终答案不一定依赖于检索内容，需要一种黑盒测试方法来审计代理对检索证据的敏感度。

Result: 在VideoExplorer风格代理上，DESTROY操作比SHAM操作导致答案变化的频率高29.3个百分点，效果显著且可复现。在LVBench的1,258个问题中，CARVE筛选出538个问题，将准确率提升3.26点。

Insight: 创新点在于提出匹配的SHAM与DESTROY重放对比机制，实现无需模型内部信息的黑盒审计；CARVE可作为路由信号而非直接的基础事实分类器，提升代理回答的可靠性。

Abstract: Tool-using video agents retrieve visual evidence before answering, but the final answer is not forced to depend on what was retrieved. The natural black box test is counterfactual: destroy the semantic content of the frames the agent retrieved and check whether the answer changes, against a matched sham that re-executes the identical pipeline on those same frames. We introduce CARVE, a black-box counterfactual probe that compares answer changes under matched SHAM and DESTROY replays. Across three independent k=3 runs on a frozen VideoExplorer-style agent, DESTROY changes the answer 29.3 percentage points more often than SHAM, yielding a large and reproducible aggregate effect. Question-level scores are less stable, and increasing the replay budget from k=3 to k=10 reduces ties but weakens the original zero-threshold routing policy. At k=3, CARVE selects 538 of 1,258 LVBench questions and improves accuracy by 3.26 points, with higher fallback yield than most matched random subsets. The score shows only a weak association with annotated temporal coverage, so CARVE is best understood as a routing signal rather than a direct grounding classifier. Our implementation is available at https://github.com/KurbanIntelligenceLab/CARVE.

[98] BASeg: Boundary-Aware Remote Sensing Segmentation with Structural Penalties cs.CVPDF

Yuexi Song, Kailai Sun, Zhuoyu Wang, Mingyi He, Paul Pu Liang

TL;DR: 本文提出了一种用于遥感图像语义分割的边界感知框架BASeg，其核心是结合了马氏距离和角度感知的边界损失函数MABL，以增强对物体边界和形状一致性的建模。该框架还引入了全局视觉状态空间模块和跨特征融合模块来捕获长程上下文依赖和细粒度局部细节。此外，研究构建了一个包含10个城市的全球基准数据集GCD-25k，并在多个基准测试中验证了方法的有效性。

Details

Motivation: 当前遥感语义分割方法难以捕捉细粒度物体特征和边界细节，且现有数据集缺乏城市形态多样性，生成式图像的分割研究也相对不足。

Result: 在四个遥感分割基准测试上的广泛实验表明，BASeg始终优于现有方法，mIoU最高提升2.8%，并能生成更准确的物体边界分割结果。将MABL集成到多种现有分割架构中，也一致地提升了它们在多个数据集上的性能。

Insight: 主要创新点在于提出了一个可即插即用的马氏距离-角度边界损失函数MABL，它通过联合建模结构重要性和边界方向来显式增强边界一致性。从客观角度看，其构建的全球多城市基准数据集GCD-25k，以及将全局上下文建模与局部细节融合的框架设计，也具有借鉴价值。

Abstract: Semantic segmentation is a core computer vision task in the remote sensing field, accelerating advancements in ur- ban development, agriculture, ecology, water resources, and environmental monitoring. However, recent methods usually struggle to capture fine-grained object features and bound- ary details. Besides, current widely used datasets often lack city morphology diversity and segmentation on generative im- ages remains largely unexplored. To address these issues, we propose a Mahalanobis-Angle Boundary Loss (MABL) that explicitly enhances boundary and shape consistency. MABL jointly models structural importance and boundary orientation through Mahalanobis distance-based weighting and angle- aware penalty. It can be readily integrated into diverse seg- mentation architectures and consistently improves their accu- racy. Built upon MABL, we introduce BASeg, a boundary- aware remote sensing segmentation framework with Struc- tural Penalties. BASeg integrates a Global Visual State Space module (GSM) with a Cross-Feature Fusion module (CFM) to capture both long-range contextual dependencies and fine- grained local details. Additionally, we establish a global 10- city benchmark dataset (GCD-25k) to facilitate accurate build- ing and road segmentation. Extensive experiments on four remote-sensing benchmarks demonstrate that BASeg consis- tently outperforms existing methods, achieving up to a 2.8% improvement in mIoU while producing more accurate object boundary segmentation across diverse scenes. Moreover, integrating MABL into multiple existing segmentation archi- tectures consistently improves performance across datasets, demonstrating its robustness and broad applicability.

[99] Training-Free Long-Term Multi-Object Tracking for Sports Video Analytics cs.CVPDF

Tomasz Stanczyk, Seongro Yoon, Francois Bremond

TL;DR: 本文提出了McByte++，一个无需训练的检测跟踪框架，用于解决体育视频中长期多目标跟踪的挑战，如频繁遮挡和快速相机运动。该方法集成了轻量级掩码传播、条件相机运动补偿和在线重识别，在提升身份保持能力的同时显著提高了运行效率。

Details

Motivation: 解决体育视频中由于频繁遮挡、快速相机运动和运动员重复出现导致的长时期多目标跟踪难题。

Result: 在SoccerNet-tracking和SportsMOT基准测试中，McByte++在在线设置下比原版McByte提升了高达3.0 HOTA和6.1 IDF1，结合离线全局关联后性能进一步提升，且速度提升可达一个数量级。

Insight: 创新点在于将轻量级掩码传播、条件相机运动补偿和在线重识别集成到统一流程中，实现了无需训练或数据集特定调优的高效跟踪，通过替换繁重的分割组件和优化运动建模来平衡性能与速度。

Abstract: Long-term multi-object tracking in sports remains challenging due to frequent occlusions, rapid camera motion, and repeated player reappearances. We introduce McByte++, a training-free tracking-by-detection framework that integrates lightweight mask propagation, conditional camera motion compensation, and online re-identification within a unified pipeline. Compared to its predecessor, McByte++ substantially improves runtime efficiency while enhancing identity preservation. On SoccerNet-tracking and SportsMOT benchmarks, McByte++ achieves up to +3.0 HOTA and +6.1 IDF1 improvements over the original McByte in the online setting, with further gains when combined with offline global association. Replacing heavy segmentation components and optimizing motion modeling yields up to an order-of-magnitude speed increase. All results are obtained without detector retraining or dataset-specific tuning. Code will be made available at https://github.com/tstanczyk95/McBytePlusPlus.

[100] ConceptFormer: Learning Adaptive Latent Concepts for Query-Document Alignment in Visual Document Retrieval cs.CV | cs.IRPDF

Peng Chunyi, Xu Zhipeng, Yan Yukun, Liu Zhenghao, Yu Shi

TL;DR: ConceptFormer是一个用于视觉文档检索的潜在概念表示学习框架，它将查询相关证据建模为连续的、查询条件化的潜在概念，从而在局部视觉证据和语义相关性之间建立显式桥梁，无需文本中间表示或原始视觉标注。

Details

Motivation: 现有视觉文档检索方法依赖文本描述或局部视觉区域作为证据代理，可能忽略复杂视觉结构或提供不完整、不准确的证据表示，ConceptFormer旨在解决这些局限性。

Result: 在多个视觉文档检索基准测试中，ConceptFormer在平均NDCG@10指标上分别比最强的视觉检索基线和最强的基于OCR的文本检索基线相对提升了16.7%和22.1%。

Insight: 创新点在于引入查询条件化的潜在概念作为中间表示，动态确定概念数量，有效连接局部视觉证据与语义相关性，使检索器能同时捕获细粒度文本线索和复杂文档级视觉结构。

Abstract: Visual document retrieval is a critical component of multimodal retrieval-augmented generation, aiming to identify query-relevant pages from document collections where evidence is distributed across text, layout, charts, and visual structures. Recent efforts toward finer-grained supervision primarily rely on textual descriptions or localized visual regions as evidence proxies. However, such supervision signals may either overlook complex visual structures or provide incomplete and inaccurate representations of the underlying evidence. To address these limitations, we propose ConceptFormer, a latent concept representation learning framework for visual document retrieval. ConceptFormer models query-relevant evidence as continuous, query-conditioned latent concepts that explicitly bridge localized visual evidence and semantic relevance, without requiring either textual intermediate representations or direct reliance on raw visual annotations. During training, ConceptFormer employs a strong vision-language model to dynamically determine the number of latent concept tokens and uses these concepts as an intermediate representation to bridge the semantic gap between queries and documents, thereby guiding the learning of the embedding space. Experiments on diverse visual document retrieval benchmarks demonstrate that ConceptFormer achieves 16.7% and 22.1% relative improvements in average NDCG@10 over the strongest visual retrieval baseline and the strongest OCR-based text retrieval baseline, respectively. Further analysis reveals that latent concepts effectively connect localized visual evidence with semantic relevance, enabling the retriever to capture both fine-grained textual cues and complex document-level visual structures while preserving strong retrieval alignment. Codes and data are available at https://github.com/Neuir/ConceptFormer.

[101] What You Ask is What You Ground: Bridging Question Intent to Temporal Evidence for Grounded VideoQA cs.CVPDF

Jinhwan Seo, Kyubeom Han, Jumin Lee, Junhyug Noh, Sung-eui Yoon

TL;DR: 本文针对Grounded VideoQA中的问题不变性定位失败模式，提出GroundFormer模型，通过可学习的通信令牌在定位前将视频特征与问题意图关联，并采用因子化MIL交叉注意力和高斯平滑技术，结合分层多模态对比损失，显著提升了NExT-GQA和STAR数据集上的问题判别性时间定位性能。

Details

Motivation: 解决Grounded VideoQA中模型对不同问题预测几乎相同时间段的失败模式，即问题不变性定位，这源于先前设计中模态隔离和问题注入不足的结构性限制。

Result: 在NExT-GQA和STAR数据集上达到最先进的Grounded VideoQA性能，显著改善了问题判别性时间定位。

Insight: 创新点包括：通过可学习通信令牌实现定向视觉-语言交互以条件化视频特征；因子化MIL交叉注意力在候选级监督下耦合答案选择与时间证据；高斯平滑将峰值注意力转换为时间连贯段；分层多模态对比损失对齐视频、问题和答案嵌入。

Abstract: We study a critical yet overlooked failure mode in Grounded Video Question Answering: question-invariant grounding, where models predict nearly identical temporal segments for different questions about the same video. We trace this behavior to two structural limitations in prior common designs: (i) modality isolation that fixes video representations before they receive question semantics, and (ii) weak question injection inside the grounding module. To address this, we propose GroundFormer, which conditions video features on question intent before localization via learnable communication tokens that mediate directed visuo-lingual interaction. On top of the question-conditioned features, a factorized MIL cross-attention couples answer selection with temporal evidence under candidate-level supervision, while Gaussian smoothing converts peaked attention into temporally coherent segments. We further introduce a hierarchical multi-modal contrastive loss that aligns video, question, and answer embeddings across a two-pass training pipeline. GroundFormer achieves state-of-the-art grounded VideoQA performance on NExT-GQA and STAR, substantially improving question-discriminative temporal grounding.

[102] Beyond Single Object: Learning 3D Relations with Large Language Models cs.CV | cs.AI | cs.CLPDF

Kohsuke Ide, Ryousuke Yamada, Yue Qiu, Xianzheng Ma, Yoshihiro Fukuhara

TL;DR: 本文针对现有3D-LLM专注于单物体/场景描述、难以进行细粒度多物体比较的不足，提出了一个用于多物体间详细推理的框架。该框架包含一个需要精细多物体比较的指令数据集MO3D、一个建模物体间/内关系并保持局部几何的轻量级Patch-Interaction Transformer模型Multi-3DLLM，以及两个用于评估几何理解实用性的应用基准（形状匹配与变化描述）。

Details

Motivation: 解决现有3D-LLM在详细、跨物体的比较推理能力上的根本性不足，旨在提升模型对多个物体间关系的细粒度理解和几何感知能力。

Result: 在提出的MO3D数据集上，Multi-3DLLM超越了所有基线模型（包括最近的3D-LLM和2D-VLM），并在单物体分类任务上表现出正向迁移效果。

Insight: 创新点在于构建了专注于多物体比较的指令数据集和轻量级的Patch-Interaction Transformer，通过建模物体间关系来学习几何推理，并设计了面向实际应用的基准任务来评估模型的几何理解能力。

Abstract: We address a fundamental gap in 3D-LLMs: existing models focus on single-object/scene description, struggling with detailed, inter-object comparison. We propose a framework for detailed object-level reasoning across multiple objects with three components: (1) MO3D (Multi-Object in 3D), an instruction dataset requiring fine-grained multi-object comparison; (2) Multi-3DLLM, using a minimal Patch-Interaction Transformer (PIT) that models inter-/intra-object relationships while preserving local geometry; (3) Mini-apps, two application-driven benchmarks (Shape Mating, Change Captioning) that probe geometric understanding for practical use. Recent 3D-LLMs and 2D-VLMs perform poorly on these tasks, lacking both comparison-centric design and geometric awareness. In contrast, Multi-3DLLM trained on our mixture data learns geometric reasoning, surpasses all baselines on MO3D, and provides positive transfer to single-object classification.

[103] Anatomical and Physical Supervision for CT-less PET Attenuation Correction: BIC-MAC 2026 Challenge cs.CVPDF

Petros Chatzitoulousis, George K. Matsopoulos

TL;DR: 本文介绍了参加BIC-MAC 2026挑战赛的提交方案，旨在通过多模态伪CT合成实现无CT的PET衰减校正。该方法基于标准nnU-Net架构，结合解剖学和物理学监督，以提升伪CT质量和下游PET重建效果。具体创新包括使用冻结的TotalSegmentator特征提取器进行解剖学监督、基于多角度衰减投影的可微衰减校正因子投影损失进行物理学监督，并利用SynthRAD挑战赛的预训练权重初始化网络。

Details

Motivation: 解决无CT扫描时PET图像衰减校正的难题，通过伪CT合成替代缺失的CT数据，以提高PET重建的准确性和临床实用性。

Result: 在BIC-MAC 2026挑战赛的基准测试中，该方法展示了结合解剖学监督、衰减物理和高效nnU-Net扩展的有效性，具体定量结果未在摘要中明确提及，但暗示性能有所提升。

Insight: 创新点在于将解剖学监督（通过预训练分割模型和结构约束）与物理学监督（可微投影损失）相结合，集成到nnU-Net流程中，无需重大架构修改，即可优化伪CT合成和PET校正的整体性能。

Abstract: This report describes our submission to the Big Cross-Modal Attenuation Correction (BIC-MAC) 2026 Challenge for CT-less PET attenuation correction through multimodal pseudo-CT synthesis. We build upon a standard nnU-Net architecture and combine anatomical and physical supervision to improve both pseudo-CT quality and downstream PET reconstruction. Anatomical supervision is introduced through a frozen TotalSegmentator feature extractor, anatomy-guided structural constraints and patch sampling, while physical supervision is achieved using a differentiable attenuation correction factor projection loss based on multi-angle attenuation projections. Furthermore, the network is initialized with pretrained weights obtained from training on the SynthRAD Challenge MR-to-CT dataset. Minimal architectural modifications are applied, while performance improvements are pursued across the nnU-Net pipeline, including preprocessing, plans, and supervision design, among other components. Our final submission demonstrates the effectiveness of combining anatomical supervision, attenuation physics, and efficient nnU-Net scaling for CT-less PET attenuation correction.

[104] Identifying Confusion Trends in Concept-based XAI for Multi-Label Classification cs.CV | cs.AIPDF

Haadia Amjad, Ronald Tetzlaff

TL;DR: 本研究探索了基于概念的可解释人工智能（CXAI）在多标签分类任务中的应用。通过在MS-COCO数据集上训练VGG16和ResNet50模型，并应用CRP和CRAFT两种CXAI方法生成概念级解释，分析了模型的学习弱点、概念混淆和数据集偏差。

Details

Motivation: 在医疗和自动驾驶等高风险领域部署深度神经网络时，模型不仅需要准确，还需具备可解释性以建立用户信任。针对包含背景噪声和大量标注的复杂图像任务，需要评估CXAI方法的适用性和问题解决能力。

Result: 在MS-COCO数据集的20个最多标注类别上评估了两种CXAI方法。分析表明，CXAI能够揭示DNN的学习弱点，更高的概念区分度可以减少标签和概念混淆，环境概念能够暴露数据集引起的偏差。

Insight: 论文的创新点在于将CXAI系统应用于多标签分类场景，并实证分析了概念解释如何揭示模型泛化能力和数据集偏差。从客观角度看，该研究为评估CXAI在复杂视觉任务中的实用性提供了方法论框架和具体发现。

Abstract: Deep Neural Networks (DNNs) deployed in high-risk domains, such as healthcare and autonomous driving, must be not only accurate but also understandable to ensure user trust. In real-world computer vision tasks, these models often operate on complex images containing background noise and are heavily annotated. To make such models explainable, Concept-based Explainable AI (CXAI) methods need to be assessed for their applicability and problem-solving capacity. In this work, we explore CXAI use cases in multi-label classification by training two DNNs, VGG16 and ResNet50, on the 20 most annotated labels in the MS-COCO dataset (Microsoft Common Objects in Context). We apply two CXAI methods, CRP (Concept Relevance Propagation) and CRAFT (Concept Recursive Activation FacTorization), to generate concept-level explanations and investigate the overall evaluations. Our analysis reveals three key findings: (1) CXAI highlights learning weaknesses in DNNs, (2) higher concept distinctiveness reduces label and concept confusion, and (3) environmental concepts expose dataset-induced biases. Our results demonstrate the potential of CXAI to enhance the understanding of model generalizability and to diagnose bias instigated by the dataset.

[105] FlowDance: Music-Driven Dance Video Generation with Parallel Pose and RGB Streams cs.CVPDF

Genying Li, Boda Lin, Jiachen Li, Zijian Jia, Haojie Zheng

TL;DR: FlowDance是一个音乐驱动的舞蹈视频生成框架，通过并行的姿态和RGB流，将显式的运动建模与参考人物外观保持的视觉合成相结合。它引入了时间步感知的姿态注入和持久身份注入技术，以改善结构引导和长视频中的外观一致性。

Details

Motivation: 解决音乐驱动舞蹈视频合成中的挑战，包括学习音乐到运动的对应关系、保持人物身份的人体动画、时间连贯性以及视觉逼真的视频生成。

Result: 在舞蹈动作生成和音乐驱动的舞蹈视频合成任务上均取得了强劲的性能，但摘要中未提及具体的基准测试或与SOTA的定量比较。

Insight: 核心创新在于并行处理姿态和RGB流的架构设计，以及用于改善去噪过程结构引导和长序列外观保持的两种注入机制。同时，构建了一个高质量、多模态的野外舞蹈视频数据集来支持该任务。

Abstract: Music-driven dance video synthesis aims to animate a reference person according to a given music clip. The task is challenging because it requires a model to jointly learn music-to-motion correspondence, identity-preserving human animation, temporal coherence, and visually realistic video generation. We present FlowDance, a music-driven dance video generation framework that integrates explicit motion modeling with reference-preserving visual synthesis through parallel pose and RGB streams. We further introduce timestep-aware pose injection to adapt structural guidance across denoising steps and persistent identity injection to preserve the reference appearance over long video. To support this task, we further build a popularity-curated, high-resolution in-the-wild dance video dataset with synchronized music, RGB videos, 3D body motion, camera parameters, and projected 2D pose annotations. Extensive experiments show that FlowDance achieves strong performance in both dance motion generation and music-driven dance video synthesis.

[106] ChainSpace: A Chained-Reasoning Paradigm for Spatial Intelligence cs.CVPDF

Xiaohan Zhang, Feng Gu, Xudong Rao, Xuhao Pan, Tao Wei

TL;DR: 本文提出了ChainSpace，一种用于空间智能的链式推理范式，旨在解决现有以数据为中心的方法将空间推理视为独立问答实例、导致基于捷径的答案和持续性空间理解监督有限的问题。该范式将空间问题组织成逻辑约束且联合一致的链，后续问题依赖于先前轮次建立的空间约束。基于此，作者构建了手动标注的真实世界多轮基准测试ChainSpace-Bench和基于模拟器的链式监督生成框架ChainSpace-Pipeline。实验表明，新基准能暴露孤立问题准确率无法捕捉的链级错误，且使用少量模拟器生成的链式数据训练的模型在新基准上达到开源模型最佳性能，并能很好地迁移到其他外部空间智能基准。

Details

Motivation: 现有空间智能方法通常将空间推理处理为独立的问答实例，这允许模型通过捷径回答问题，并为持续性空间理解提供了有限的监督。为了促进对连贯空间状态的维护，需要一种能结构化多轮交互、保持状态一致性的新范式。

Result: 在提出的ChainSpace-Bench基准上，实验暴露了模型在链级推理上的失败，这些失败未被孤立问题准确率捕获。使用ChainSpace-Pipeline生成的相对少量链式数据训练的模型，在ChainSpace-Bench上取得了开源模型中的最佳性能，并且在多个外部空间智能基准（如VQA-v2、GQA、CLEVR）上具有竞争力的迁移表现。

Insight: 核心创新在于将空间智能建模为一个状态保持的多轮链式推理过程，通过逻辑约束的问答链来强制模型维持连贯的空间状态。这提供了更忠实（faithful）的评估范式和更数据高效的学习方法。从客观角度看，其链式监督生成框架（利用模拟器）为获取高质量、结构化的训练数据提供了一种有前景的途径。

Abstract: Spatial intelligence requires foundation models to maintain coherent spatial state across interactions with the physical world. However, existing data-centric approaches typically treat spatial reasoning as independent question-answer instances, enabling shortcut-based answering and providing limited supervision for persistent spatial understanding. To address this, we introduce ChainSpace, a chained-reasoning paradigm that structures spatial reasoning as a state-preserving multi-round process. In this paradigm, spatial questions are organized into logically constrained and jointly consistent chains, where later questions depend on spatial constraints established in earlier rounds. Following this principle, we instantiate ChainSpace-Bench, a manually annotated real-world multi-round benchmark with a Chain-Aware Metric, and ChainSpace-Pipeline, a simulator-based chain-structured supervision generation framework for spatial intelligence training. Experiments show that ChainSpace-Bench exposes chain-level failures that are not captured by isolated question accuracy. Additionally, with a relatively small amount of simulator-generated chained data, models trained by ChainSpace-Pipeline achieve the best performance among open-source models on ChainSpace-Bench and transfer competitively to multiple external spatial intelligence benchmarks. These results establish ChainSpace as an effective paradigm for more faithful evaluation and more data-efficient learning of spatial intelligence.

[107] PWLR: Pairwise Witness Local Rejection for Boundary-Aware Out-of-Distribution Detection cs.CV | cs.LGPDF

Chengyao Jia, Ruixuan Wang

TL;DR: 本文提出了一种名为PWLR（Pairwise Witness Local Rejection）的新方法，用于提升图像分类器的分布外（OOD）检测能力，特别是在处理与分布内（ID）类别边界接近的near-OOD样本时。该方法利用多模态大语言模型（MLLM）离线生成描述性的局部视觉线索短语，这些线索用于区分一对混淆的ID类别。通过仅使用ID数据筛选可靠的局部验证器，并在推理时结合全局类别得分与成对的局部证据进行校准，从而做出更准确的OOD判断。

Details

Motivation: 当前基于视觉语言的OOD检测方法虽然利用类别语义、局部提示或LLM生成的离群概念有所改进，但很少将语言作为混淆ID类别之间的明确边界证据。本文旨在解决near-OOD样本靠近ID类别边界时检测困难的问题。

Result: 在ImageNet-100的far-OOD、cleaner/challenging OOD和near-OOD基准测试上的实验表明，PWLR方法在多个骨干网络上持续改进了强大的视觉语言基线模型，提升了OOD检测性能。

Insight: 核心创新在于利用MLLM生成成对的、基于局部视觉线索的类别区分短语作为边界证据，并通过ID数据筛选和校准将其与全局分类得分结合。这提供了一种将语言模型生成的细粒度、可解释的局部证据系统性地整合到OOD检测决策中的新范式。

Abstract: Out-of-distribution (OOD) detection remains challenging for image classifiers, especially when near-OOD samples lie close to in-distribution (ID) class boundaries. Recent vision-language detectors improve OOD detection through class semantics, local prompting, or LLM-generated outlier concepts, but seldom use language as explicit boundary evidence between confusing ID classes. We propose Pairwise Witness Local Rejection (PWLR), which uses an MLLM offline to describe visible local cues that favor one ID class over a specific rival class. These cue phrases are then screened with ID-only data under a frozen vision-language backbone, so that only reliable local verifiers are kept. At inference, PWLR first retains a small set of globally plausible classes, then checks whether any of them is locally supported against its most relevant rivals, and finally combines this pairwise local evidence with the global class score through calibration. Experiments on ImageNet-100 far-OOD, cleaner/challenging OOD and near-OOD benchmarks show that PWLR consistently improves strong vision-language baselines across multiple backbones. Source code will be released.

[108] MITE-Net: SWaP-Optimized 4K Video Tiny Target Perception for Embodied Edge SAR cs.CVPDF

Mingshuo Xu, Mu Hua, Jigen Peng, Qi Wang, Shigang Yue

TL;DR: 本文提出了MITE-Net，一个针对边缘设备（如无人机）在严格尺寸、重量和功耗（SWaP）约束下进行4K视频微小目标感知的优化框架。该框架包含一个级联架构、专门的数据集和硬件基准测试。MITE-Net结合了仿生、无需学习的运动区域提议网络和超轻量级检测头，旨在解决高分辨率图像下采样导致特征丢失与分片处理延迟过高的问题。

Details

Motivation: 动机在于解决嵌入式搜索与救援（SAR）任务中，边缘设备因严格的SWaP限制而面临的实时高分辨率微小目标感知瓶颈。传统方法如降采样会导致严重特征损失，而基于分片的处理则带来难以接受的延迟。

Result: 在NVIDIA Jetson AGX Xavier边缘设备上，MITE-Net直接处理4K海事图像，实现了30.33 FPS的帧率和100%的搜索成功率，功耗仅为3.19 W（能效为9.51 FPS/W），在目标召回率和能效上大幅超越YOLO基线模型。但在城市场景的UAVID-Tiny数据集上，其无学习仿生前端对复杂背景处理能力不足，超轻量级头部也缺乏复杂特征的表示能力。

Insight: 创新点包括：1）提出了一种SWaP优化的级联架构MITE-Net，结合了仿生运动提议和超轻量检测头；2）构建了标准化的4K微小目标评估数据集SAR-Tiny（包括SeaDroneSee-Tiny和UAVID-Tiny）；3）提供了硬件级的严格基准，为未来端到端SAR架构设计提供了指导。从客观角度看，其将仿生无学习前端与极小参数检测头结合的思路，为边缘设备上的高效感知提供了新范式，但结构局限性在复杂场景中凸显，表明需要更平衡的设计。

Abstract: Real-time tiny target perception in high-resolution imagery is critical for embodied Search-and-Rescue (SAR) missions. However, strict Size, Weight, and Power (SWaP) constraints on edge devices like UAVs create a bottleneck: traditional image downsampling causes severe feature loss, while slice-based processing incurs prohibitive latency. To address this gap, this paper introduces a comprehensive framework encompassing a novel architecture, specialized datasets, and hardware-level benchmarks. First, we propose MITE-Net, a SWaP-optimized cascaded architecture, which couples a bio-inspired, learning-free Tiny Target Motion-Based Region Proposal Network (TTM-RPN) with a sub-0.14M-parameter R-CNN-like head. Second, to standardize 4K tiny target evaluation, we construct the SAR-Tiny Datasets by relabeling two challenging UAV datasets: SeaDroneSee-Tiny (dynamic maritime scenes, tiny targets predominantly of 64-256 pixels ) and UAVID-Tiny (cluttered urban scenes, extremely tiny targets, less than 64 pixels). Third, we benchmark against state-of-the-art YOLO models on an edge device, NVIDIA Jetson AGX Xavier, where MITE-Net directly processes 4K maritime imagery, achieving a 100% search success rate at 30.33 FPS. Consuming merely 3.19 W (9.51 FPS/W), MITE-Net vastly outperforms YOLO baselines in target recall and energy efficiency. Conversely, UAVID-Tiny evaluations expose a compound structural limitation: the learning-free bionic front-end struggles against urban backgrounds, while the ultra-lightweight head lacks representational capacity for complex features. Ultimately, this work delivers an efficient onboard perception paradigm and a rigorous baseline guiding future end-to-end SAR architectures.

[109] CardiacMamba: Fair and Robust RGB-RF Fusion for Remote Heart Rate Estimation via State Space Modeling cs.CV | cs.AIPDF

Bo Zhao, Zheng Wu, Yiping Xie, Zitong YU

TL;DR: 本文提出了一种名为CardiacMamba的公平且鲁棒的RGB-RF融合框架，用于远程心率估计。该框架通过状态空间建模整合了面部光学信号（RGB）和射频心脏运动信号（RF），以解决传统RGB方法易受光照、运动伪影和肤色差异影响的问题。

Details

Motivation: 远程光电容积描记术（rPPG）虽然能通过面部视频进行非接触式心率监测，但仅依赖RGB的方法对光照变化、运动伪影和肤色依赖的光学反射非常敏感，因此需要融合更鲁棒的信号源（如RF）来提升公平性和鲁棒性。

Result: 在EquiPleth数据集上，CardiacMamba实现了最先进的性能：平均绝对误差（MAE）为0.96 bpm，均方根误差（RMSE）为3.06 bpm，皮尔逊相关系数为0.97。同时，它将观察到的深浅肤色间的MAE差距缩小至0.26 bpm，并在RGB退化或RF缺失条件下保持鲁棒性。

Insight: 论文的创新点包括：引入时序差分Mamba模块（TDMM）来增强细微的RF时序变化，基于双向状态空间模型（SSM）的交互机制以对齐异构的RGB-RF动态，以及通道级快速傅里叶变换（CFFT）模块进行通道域频谱细化。这些设计有效提升了多模态融合的公平性和鲁棒性。

Abstract: Remote photoplethysmography (rPPG) enables non-contact heart rate (HR) monitoring from facial videos, but RGB-only methods are vulnerable to illumination changes, motion artifacts, and skin-tone-dependent optical reflectance. We propose CardiacMamba, a fair and robust RGB-RF fusion framework that integrates optical facial cues and radio-frequency cardiac motion cues through state space modeling. CardiacMamba introduces a Temporal Difference Mamba Module (TDMM) to enhance subtle RF temporal variations, a bidirectional SSM-based interaction mechanism to align heterogeneous RGB-RF dynamics, and a Channel-wise Fast Fourier Transform (CFFT) module for channel-domain spectral refinement. On the EquiPleth dataset, CardiacMamba achieves state-of-the-art performance with 0.96 bpm MAE, 3.06 bpm RMSE, and 0.97 Pearson correlation, while reducing the observed light-dark skin-tone MAE gap to 0.26 bpm and maintaining robustness under RGB degradation and RF-missing conditions

[110] Beyond Visual CoT: Internalized Visual Thinking for Proactive Video Reasoning cs.CV | cs.AI | cs.CL | cs.LG | cs.MMPDF

Xiaoyu Zhu, Xinke Deng, Suresh Taddewadikar, Arnab Kumar Mondal, Zhongyu Jiang

TL;DR: 本文提出了一种名为内部化视觉思维（IVT）的后训练框架，用于主动视频推理。IVT通过联合优化文本预测和未标记视频的下一嵌入预测，使模型在训练时学习视觉思维，在推理时直接生成答案，无需生成中间图像，从而显著降低延迟。

Details

Motivation: 解决视觉思维链（Visual CoT）在主动视频推理中引入的显著推理开销问题，探索模型是否能在训练时学习视觉思维，在推理时直接推理。

Result: 在六个评估设置中，IVT均优于直接答案微调，且推理路径相同；与显式Visual CoT相比，IVT达到相当或更好的性能，并将平均端到端延迟降低5倍以上。

Insight: 创新点在于将预测性世界建模内部化到训练中，使多模态推理器更准确高效；摘要宣称的突破是推理时无需显式像素空间生成，即可实现有效的主动视频推理。

Abstract: Multimodal large language models increasingly use visual chain-of-thought (Visual CoT) to reason about spatial, temporal, and embodied environments. By generating intermediate reasoning images, Visual CoT provides an intuitive mechanism for visual foresight but introduces substantial inference overhead, which is particularly problematic for proactive video reasoning. We ask whether models can learn to think visually during training while reasoning directly at inference. We introduce Internalized Visual Thinking (IVT), a post-training framework that jointly optimizes textual prediction and next-embedding prediction over unlabeled videos. Given a partially observed video, IVT predicts latent representations of future frames together with the target textual answer, encouraging the model to capture motion, object transitions, interactions, and latent intent. At inference, IVT generates the answer directly without synthesizing or re-encoding future frames. We conduct controlled studies across target representations, decoder designs, prediction horizons, data mixtures, training curricula, and predictive objectives. IVT improves over direct-answer fine-tuning on all six evaluation settings while retaining the same inference pathway. Compared with explicit Visual CoT, IVT achieves comparable or better performance and reduces average end-to-end latency by more than 5x. Together, our findings suggest that explicit pixel-space generation at inference time, as used in visual chain-of-thought, may not be necessary for effective proactive video reasoning. Predictive world modeling can be internalized during training to produce multimodal reasoners that are both more accurate and substantially more efficient.

Yubo Zhang, Yiyao Liu

TL;DR: CM-MAE是一个自监督的视觉-无线预训练框架，旨在通过RGB帧和64波束接收功率向量实现跨场景表示迁移。其核心包括软对比对齐损失和掩码联合解码器，以学习稳健的跨模态表示，并通过差分速率微调规则快速适应新场景。

Details

Motivation: 解决在视角、流量、光照和传播几何变化时，单一部署场景下学习的表示在其他场景中失效的问题，实现跨场景的稳健表示迁移。

Result: 在DeepSense 6G数据集上，软对齐损失将匹配线性探针迁移平均准确率从24.88%提升至29.49%；融合微调在未见场景6-8上达到77.38%的Top-1准确率，经归纳归一化适应后可提升至78.69%。

Insight: 创新点包括软对比对齐损失（避免相似方向响应的非相同样本被错误分离为负样本）和掩码联合解码器（通过模态丢弃重建视觉块和无线角度簇），以及差分速率微调规则（实现编码器慢更新、融合头快适应）。

Abstract: Synchronized camera and wireless measurements observe the same scene through different physical channels. The central difficulty is that a representation learned in one deployment can fail when viewpoint, traffic, illumination, and propagation geometry change. This paper presents CM-MAE, a self-supervised vision–wireless pretraining framework for cross-scenario representation transfer. The evaluated real-data model uses only RGB frames and the measured 64-beam received-power vector available in DeepSense 6G; it does not use ray-traced paths, calibrated depth, or beam-index labels during pretraining. Its central pretraining term is a \emph{soft contrastive alignment loss}. Instead of making the synchronized image–wireless pair the only positive pair, this loss builds a target distribution from similarities between measured beam-power profiles, so nonidentical samples with similar directional responses are not forced apart as false negatives. A masked joint decoder provides the complementary local objective by reconstructing hidden visual patches and wireless angular clusters under modality dropout. After pretraining, a differential-rate fine-tuning rule lets a new fusion head adapt quickly while the encoders move slowly. Under a sequence-disjoint DeepSense 6G protocol, adding the soft alignment loss improves a matched linear-probe transfer average from 24.88% to 29.49%. Mild fusion fine-tuning reaches 77.38% Top-1 accuracy on unseen Scenarios 6–8, and optional transductive normalization adaptation reaches 78.69%. Since the fusion setting uses the contemporaneous 64-beam power vector at inference, these results should be read as representation-transfer diagnostics, not as proactive beam-prediction or reduced-sweeping claims.

[112] Comprehensive Benchmarking of Deep Learning Architectures for Lung Cancer Histopathology cs.CV | cs.AIPDF

Hadi Hasan, Safaa Salman, Lama Sleem, Ralph Mouawad, Ali Chehab

TL;DR: 本研究提出了一个用于肺癌组织病理学图像分析的两阶段深度学习框架，包含多类组织分类和像素级区域分割任务，并对各阶段的最先进架构进行了系统性比较。

Details

Motivation: 肺癌是全球癌症相关死亡的主要原因，而组织病理学诊断常受观察者间差异和繁重人工阅片工作量的影响。尽管深度学习在计算病理学中展现出潜力，但将组织分类和区域分割整合在统一分析框架内的综合性基准研究仍然有限。

Result: 在组织分类任务中，YOLO11在包含39,000张图像的数据集上取得了最佳性能，准确率达98.38%；在区域分割任务中，DeepLabV3+在GlaS腺体分割基准上获得了最高的IoU（0.80）和Dice分数（0.89），而YOLO11-seg以约14倍更少的参数实现了可比的IoU（0.79）。

Insight: 论文的创新点在于构建了一个端到端的统一分析框架，并进行了广泛的架构基准测试，为自动化组织病理学图像分析提供了准确、计算高效且可复现的基线。特别是发现YOLO11系列模型在分类和分割任务中均表现出色，且在分割任务中实现了参数效率与性能的平衡。

Abstract: Lung cancer remains the leading cause of cancer-related mortality worldwide, while histopathological diagnosis is often affected by inter-observer variability and the substantial workload associated with manual slide examination. Although deep learning has shown considerable potential in computational pathology, comprehensive benchmarks that integrate tissue classification and region segmentation within a unified analytical framework remain limited. This study presents a two-stage deep learning framework for multi-class tissue classification and pixel-level histopathological region segmentation, accompanied by a systematic comparison of state-of-the-art architectures at each stage. For tissue classification, six models, a custom convolutional neural network, VGG16, DenseNet, MobileNetV3, a custom Vision Transformer, and YOLO11, are evaluated on a combined dataset of 39,000 images derived from LC25000 and LungHist700. The models distinguish between adenocarcinoma, squamous cell carcinoma, and normal lung tissue. YOLO11 achieves the best classification performance, with an accuracy of 98.38%, a five-fold cross-validation accuracy of 98.21 +/- 0.35%, and a macro F1-score of 0.98. For region segmentation, U-Net, ResNet-encoder U-Net, DeepLabV3+, and YOLO11-seg are evaluated using the GlaS gland segmentation benchmark. DeepLabV3+ obtains the highest Intersection over Union of 0.80 and a Dice score of 0.89, while YOLO11-seg achieves a comparable Intersection over Union of 0.79 using approximately 14x fewer parameters. The best-performing classification and segmentation models are subsequently integrated into an end-to-end framework, providing an accurate, computationally efficient, and reproducible baseline for automated histopathological image analysis.

[113] CLARA: Clip-Level Multimodal Alignment with VLM-Derived Rationales for Hateful Video Detection cs.CV | cs.MMPDF

Yuchen Zhang, Shuang Dai, Zeyu Fu, Yunfei Long, Ravi Shekhar

TL;DR: 本文提出CLARA，一种用于仇恨视频检测的片段级多模态框架。该方法将视频建模为细粒度片段序列，通过专家混合编码器实现自适应多模态对齐，结合局部-全局片段对比目标建模短期线索与长期依赖，并利用VLM生成的原理提供高级语义指导。在三个仇恨视频数据集上的实验表明，CLARA持续优于现有最先进方法。

Details

Motivation: 随着视频社交平台的快速增长，仇恨视频检测变得日益重要，但相比文本或静态多模态内容，仇恨视频检测研究不足且更具挑战性，因为仇恨含义常产生于语音、音频和视觉内容等多模态线索的复杂交互，且这些信号通常短暂、隐式且具有时间依赖性，传统视频级表示难以捕捉。

Result: 在三个仇恨视频数据集上的广泛实验表明，CLARA持续优于最先进方法（SOTA）。进一步的消融研究和参数分析验证了每个组件的有效性。

Insight: 创新点包括：将视频建模为片段序列以精确捕捉时间局部化仇恨信号；引入专家混合片段编码器实现自适应多模态对齐；设计局部-全局片段对比目标联合建模短期线索与长期时间依赖；通过门控Transformer集成VLM生成的原理提供高级语义指导。

Abstract: Hateful video detection has become increasingly important with the rapid growth of video-centric social media platforms, given the serious risks that hate speech poses to both individual well-being and social cohesion. Compared with text or static multimodal content, hateful video detection remains underexplored and significantly more challenging, as hateful meaning often arises from complex interactions among multimodal cues, including speech, audio, and visual content. Moreover, such signals are often brief, implicit, and temporally dependent, making them difficult to capture using conventional video-level representations. In this work, we propose CLARA, a clip-level multimodal framework for hateful video detection. Instead of treating a video as a single instance, CLARA models it as a sequence of fine-grained clips, enabling more precise capture of temporally localized hateful signals. We introduce a Mixture-of-Experts clip encoder for adaptive multimodal alignment, a local-global segment contrastive objective to jointly model short-term cues and long-range temporal dependencies, and VLM-derived rationales integrated via a gated Transformer to provide high-level semantic guidance. Extensive experiments on three hateful video datasets demonstrate that CLARA consistently outperforms state-of-the-art methods. Further ablation studies and parameter analyses validate the effectiveness of each component.

[114] Spatial Temporal Synergy: Balancing Change and Invariance in Text Driven 3D Human Motion Editing cs.CVPDF

Shaohui Lin, Zhenwu Shi, Jingyu Gong, Jiao Xie, Yu Zhou

TL;DR: 本文提出了一种名为CIME的统一框架，用于解决文本驱动3D人体运动编辑中平衡文本响应性’变化’与运动惯性’不变性’的挑战。该方法将变化与不变性解耦到空间姿态和时间节奏两个维度，通过全监督正负学习机制处理空间姿态，并引入黎曼非均匀积分流形映射模块处理时间节奏，以实现高保真的运动编辑。

Details

Motivation: 现有基于扩散的方法在文本驱动人体运动编辑中难以平衡文本驱动的’变化’与保持原始运动结构’不变性’，常因粗糙的空间约束和僵化的均匀时间假设导致空间运动失真和内在物理节奏破坏，特别是在变长度编辑时。

Result: 在MotionFix和STANCE Adjustment数据集上的大量实验表明，CIME在编辑对齐度和结构保真度方面达到了最先进的性能，验证了其统一架构的有效性。

Insight: 创新点在于将变化与不变性系统地解耦到空间和时间两个维度，并分别设计了针对性的解决方案：空间上采用包含分层回顾特征监督、细微运动保持和基于三元组的语义对齐的全监督正负学习机制；时间上引入基于运动学感知非均匀时间戳的黎曼非均匀积分流形映射模块，以高保真地再现编辑后运动的物理节拍。

Abstract: Text-driven human motion editing aims to modify existing motion sequences according to natural language instructions while maintaining the structural consistency of the original motion. Existing diffusion-based approaches struggle to balance text-responsive “change” and inertial “invariance”. They often rely on coarse spatial constraints and rigid uniform time assumptions, leading to spatial motion distortions and the destruction of intrinsic physical rhythms during variable-length editing. To handle these challenges, we propose Change and Invariance Motion Editing (CIME), a unified framework that comprehensively decouples change and invariance into spatial pose and temporal rhythm dimensions. For spatial poses, our method integrates an omni-supervised positive-negative learning mechanism comprising hierarchical retrospective feature supervision, subtle motion preservation, and triplet-based semantic alignment. For temporal rhythms, we introduce the Riemannian Non-uniform Integral Manifold Mapping (RNIMM) module, which achieves high-fidelity reproduction of physical beats in the edited text via kinematics-aware non-uniform timestamps. Extensive experiments on the MotionFix and STANCE Adjustment datasets demonstrate that CIME achieves state-of-the-art performance in editing alignment and structural fidelity, validating the effectiveness of our unified architecture. Our source codes and models have been released at: github.com/ZhenwuShi/CIME.git

[115] A Plug-and-Play 2D Motion Interface for Real-World Motion Language Models cs.CVPDF

Kaname Yokoyama, Norimichi Ukita

TL;DR: 本文提出了一种即插即用的2D运动接口，使得预训练的3D运动语言模型能够直接处理2D运动输入，而无需修改或微调原始模型。该方法在多个公共数据集上取得了与3D输入相当的性能，并在单目视频的真实世界评估中展示了2D运动相对于3D运动的实用性。

Details

Motivation: 现有运动语言模型通常依赖3D运动数据，但从单目视频中获取准确的3D运动具有挑战性，这限制了模型在真实世界中的应用。本文旨在解决这一瓶颈，使模型能够利用更易获取的2D运动输入。

Result: 在多个公共数据集上的实验表明，该方法在多个MoLM上达到了与3D运动输入相当的性能，并且优于从头开始用2D运动训练的模型。在构建的单目真实世界视频评估数据集上，通过引入真实视频适配器，证明了在评估的单目姿态估计设置下，2D运动比3D运动更有用。

Insight: 核心创新在于设计了一个无需修改或微调预训练3D MoLM的即插即用2D接口，这为将运动语言模型部署到真实世界（依赖单目视频输入）提供了一条实用路径。其构建的真实世界视频评估数据集和适配器也颇具参考价值。

Abstract: Motion Language Models (MoLMs) typically understand human motions by tokenizing 3D motion and processing the resulting tokens using a language model. However, obtaining accurate 3D motions from monocular videos is challenging, limiting their real-world applicability. To address this issue, we introduce a plug-and-play 2D Motion Interface that enables 3D-pretrained MoLMs to accept 2D motion inputs without modifying or fine-tuning the original models. Experiments on public datasets show that our method achieves performance comparable to 3D motion inputs across multiple MoLMs and outperforms training MoLMs from scratch on 2D motions. We further construct a monocular real-world video motion evaluation dataset and introduce a real-video adapter, demonstrating the usefulness of 2D motions over 3D motions under the evaluated monocular pose-estimation setting. These results suggest that 2D motion provides a practical interface for deploying MoLMs in real-world motion understanding settings. Code is available at https://github.com/irajisamurai/2D-Motion-Interface.

[116] Depth-guided Multi-view Exposure Bracketing for HDR Robot Vision cs.CVPDF

Jinnyeong Kim, Juhyung Choi, Woohyeok Kim, Sunghyun Cho, Seung-Hwan Baek

TL;DR: 该论文提出了一个用于评估多传感器机器人系统中高动态范围（HDR）感知的大规模数据集，并引入了一种名为深度引导多视图曝光包围（DMEB）的单次HDR成像方法。该方法通过在多个低比特深度相机上分配不同曝光度，并利用深度信息进行置信度感知融合，以应对极端光照条件。

Details

Motivation: 解决在极端光照条件下实现可靠的单次HDR成像的长期挑战，并填补多传感器机器人系统中缺乏全面HDR感知评估基准的空白。

Result: 在作者构建的大规模真实世界和合成数据集上的评估表明，DMEB方法建立了一个强有力的参考基准，并展示了该传感器配置在各种多相机和深度传感器系统中实现鲁棒HDR感知的潜力。

Insight: 创新点在于提出了一个专门用于机器人HDR感知的综合性数据集，以及一种结合多视图不同曝光和深度引导融合的单次HDR成像方法，为多传感器系统在复杂光照下的感知提供了新思路。

Abstract: Achieving reliable single-shot high dynamic range (HDR) imaging under extreme illumination conditions remains a long-standing challenge, yet no comprehensive benchmark exist for evaluating HDR perception in multi-sensor robotic systems. To fill this gap, we introduce a large-scale dataset collected via a custom robotic vision platform and an iPhone 13 Pro: 121 real-world scenes spanning modest and ultra-high dynamic range conditions, alongside 20 synthetic video sequences from the CARLA simulator. As a reference pipeline for this dataset, we propose Depth-guided Multi-view Exposure Bracketing (DMEB), a single-shot HDR method that distributes drastically different exposures across multi-view low-bit-depth cameras and fuses them via depth-guided confidence-aware fusion. Evaluations on our dataset show that DMEB establishes a strong reference point and highlight the promise of this sensor configuration for robust HDR perception in diverse multi-camera and depth sensor system.

[117] Representation Is Not Enough: Body-Localized Thermal Evidence for Contactless Stress and Craving Sensing in Opioid Use Disorder cs.CV | cs.LGPDF

Sachin Deb, Harshit Sharma, Asif Salekin

TL;DR: 本文提出FABLE-Therm，一种弱监督架构，用于从无接触热成像视频中检测阿片类药物使用障碍（OUD）患者的压力和渴求信号。该模型通过保留身体区域、时间和编码器特定表征的局部证据，解决了无接触传感中的弱监督证据定位问题。

Details

Motivation: 无接触生理监测移除了可穿戴设备，也失去了对压力反应发生位置和时间的监督信号，这使其成为一个弱监督的证据定位问题。在OUD早期康复中，持续使用可穿戴设备困难，而压力是复发的主要诱因，因此需要无接触的传感方法。

Result: 在固定的热成像视频上，FABLE-Therm在留出参与者上达到了0.938的AUROC，其学习到的表征可迁移到自我报告的渴求检测上，首次证明了渴求可以从无接触热视频中恢复。

Insight: 创新点在于提出了一种弱监督的局部证据融合架构（FABLE-Therm），理论解释了局部融合优于特征拼接和预测平均的原因。研究还表明，仅改进表征不足以实现公平部署，需考虑人群异质性，并提供了首个队列结构的无接触热成像OUD基准。

Abstract: Removing wearables from physiological monitoring also removes their supervision: the signal indicating where and when a stress response occurred. Contactless stress sensing therefore becomes a weakly supervised evidence-localization problem, where a clip-level label must be traced to the body regions and moments that produced it. We address this with FABLE-Therm, a weakly supervised architecture that preserves localized evidence across body regions, time, and encoder-specific representations until the final decision. FABLE-Therm fuses frozen foundation-model encoders at the embedding level, with theory explaining why localized fusion can outperform feature concatenation and prediction averaging. We study this problem in opioid use disorder (OUD), where stress is a major relapse trigger and sustained wearable use can be difficult during early recovery. Using fixed thermal video, FABLE-Therm achieves 0.938 AUROC on held-out participants, and its learned representation transfers to self-reported craving, providing, to our knowledge, the first evidence that craving can be recovered from contactless thermal video. Localized evidence also enables participant-level analysis of deployment failure. We find that improving representation alone is insufficient for equitable deployment: additional data from the underserved group would recover only about half of the cohort gap, while the remainder reflects person-to-person heterogeneity. This modality-agnostic decomposition applies to models with identifiable subpopulations. Together with the first cohort-structured contactless thermal OUD benchmark, our results show that preserving localized evidence supports both accurate sensing and principled analysis of who a model fails and why.

[118] Beyond Similarity Matching: Structured Reasoning for Open-Vocabulary Referring Segmentation in 3DGS cs.CVPDF

Yizhao Wang, Xinfa Wang, Jingbo Wang, Jingbo Wang, Guantao Zhang

TL;DR: 本文提出了QAGaussian，一种用于3D高斯泼溅（3DGS）中开放词汇指代分割的查询自适应神经推理框架。该方法通过查询条件化的多尺度高斯槽学习、关系感知的槽图推理和粒度自适应路由，解决了现有基于相似度匹配方法在处理属性、参照物、空间关系和细粒度部件时的问题，显著提升了分割性能。

Details

Motivation: 现有基于3DGS的开放词汇指代分割方法主要依赖全局文本-区域相似度，在处理涉及属性、参照对象、空间关系和细粒度部件的查询时表现较弱，容易导致目标-参照混淆、粒度不匹配、部件-整体泄漏和关系违反等问题。

Result: 在独立基准测试中，QAGaussian无需目标数据集微调，取得了47.2的平均mIoU和63.2的平均F1分数，比最强的3DGS指代基线分别提升了2.7个mIoU点和2.9个F1点。同时，部件mIoU从38.6提升至43.4，关系mIoU从44.4提升至50.8，目标-参照混淆从10.8降低至7.4，实现了SOTA性能。

Insight: 核心创新在于将指代分割从简单的相似度匹配提升为结构化推理。具体包括：1）查询条件化的多尺度高斯槽学习，使候选区域感知输入表达；2）构建语言条件化边加权的、关系感知的槽图进行证据传播；3）粒度自适应路由器软融合多级掩码分支，并结合关系约束的细化来保证一致性。这为3DGS中的开放词汇分割提供了一种有效的神经建模策略。

Abstract: Open-vocabulary referring segmentation in 3D Gaussian Splatting (3DGS) requires a neural model to select Gaussian primitives according to free-form language expressions. Existing 3DGS-based methods usually rely on global text-region similarity, which is weak for queries involving attributes, reference objects, spatial relations, and fine-grained parts. This often causes target-reference confusion, granularity mismatch, part-whole leakage, and relation violations. We propose QAGaussian, a query-adaptive neural reasoning framework for language-guided Gaussian primitive selection. QAGaussian first learns query-conditioned multi-scale Gaussian slots as differentiable candidates whose receptive fields are shaped by the input expression. It then builds a relation-aware slot graph with language-conditioned edge weighting to propagate target-reference, attribute, part-whole, and contextual evidence. A granularity-adaptive router softly combines region-level, object-level, part-level, attribute-aware, and relation-aware mask branches, followed by relation-constrained refinement for spatial, part-whole, attribute, and geometric consistency. QAGaussian is pretrained only on Mosaic3D-5.6M for Gaussian-text alignment and evaluated on independent benchmarks without target-dataset fine-tuning. It achieves 47.2 Avg. mIoU and 63.2 Avg. F1, outperforming the strongest 3DGS referring baseline by 2.7 mIoU points and 2.9 F1 points. It also improves Part-mIoU from 38.6 to 43.4, Rel-mIoU from 44.4 to 50.8, and reduces target-reference confusion from 10.8 to 7.4. These results demonstrate that query-conditioned slot learning, relation-aware graph reasoning, and adaptive routing provide an effective neural modeling strategy for open-vocabulary referring segmentation in 3DGS. The code is available at https://github.com/zqeslwyz/QAGaussian.

[119] Deep Thought Alignment: Trajectory-Level Latent Distillation for Video Reasoning cs.CV | cs.AI | cs.CLPDF

Ao Shen, Yongheng Zhang, Yinghui Li, Manning Wang, Di Yin

TL;DR: 本文提出了一种名为Latent-OPD的新方法，用于视频推理任务中的模型蒸馏。该方法在传统的策略蒸馏基础上，引入了轨迹级别的潜在表示蒸馏，并配合渐进式教师前瞻策略，旨在将大型视频推理模型的推理能力更有效地迁移到更小、更高效的模型上。

Details

Motivation: 视频推理的大型多模态模型因处理海量视觉信息而计算成本高昂，需要将大模型能力蒸馏到小模型。现有策略蒸馏方法仅匹配输出token分布，无法直接约束推理过程中形成的潜在表示，而视频推理常依赖跨多帧的证据积累。

Result: 在六个视频推理基准测试上的实验表明，Latent-OPD始终优于仅使用输出的策略蒸馏方法。在帧数有限、视频较长或需要复杂证据聚合的任务中，改进尤为显著。

Insight: 核心创新点在于轨迹级别的潜在蒸馏，聚焦于轨迹末端能总结累积证据和推理上下文的隐藏状态。此外，渐进式教师前瞻策略将学生中后期层与教师更深层对齐，这是一种新颖的层间对齐机制，有助于提升模型在资源受限下的推理效率。

Abstract: Large Multimodal Models (LMMs) for video reasoning have long been hindered by the high computational cost of processing vast amounts of visual information. This dilemma motivates the transfer of the reasoning capabilities of large models to smaller, more efficient ones. On-Policy Distillation (OPD) offers a promising solution by matching output-token distributions along student-generated trajectories. However, video reasoning often depends on evidence accumulated across multiple frames. In this context, output-level supervision only captures information expressed through token predictions and does not directly constrain the latent representations formed during reasoning. To address this limitation, we propose Latent-OPD, which augments OPD with trajectory-level latent distillation. Specifically, our method focuses on the position at the end of each trajectory, where hidden states effectively summarize the accumulated visual evidence and reasoning context. Furthermore, we introduce a progressive teacher-lookahead strategy, which aligns middle-to-late student layers with increasingly deeper teacher layers. Experiments on six video reasoning benchmarks show that Latent-OPD consistently outperforms output-only OPD. Notably, the improvements are particularly pronounced in scenarios with limited frames, long videos, or tasks requiring complex evidence aggregation. These results establish Latent-OPD as a highly effective approach to frame-efficient video reasoning.

[120] SafeGesture: Evaluating Fine-Grained Hand Gesture Understanding in Vision-Language Models through Scenario-Conditioned Safety Interpretation cs.CVPDF

Taegang Kim, Saleh Afroogh, Junfeng Jiao

TL;DR: 论文提出了SafeGesture基准，用于评估视觉语言模型在安全关键操作场景下对精细手部手势的理解能力。该基准将六种HaGRID手势与八种操作场景配对，构建了4800个测试项，并对Qwen2.5-VL-7B、LLaVA-NeXT-7B、InternVL2-8B、Phi-3.5-Vision和GPT-4o等模型进行了评估。研究发现模型存在感知与推理的解耦现象，即手势识别准确率高但安全推理准确率低，且主要瓶颈在于场景条件下的安全推理而非手势识别。

Details

Motivation: 现有开源和前沿视觉语言模型在通用图像理解上表现良好，但其在安全关键操作场景中解释精细手部手势的能力尚未得到充分检验。论文旨在评估模型是否能从手部手势推断出场景适当的安全行动。

Result: 在SafeGesture基准上，GPT-4o手势识别准确率达98.4%，但安全推理准确率仅为53.3%；Qwen2.5-VL分别为84.9%和39.5%，两者差距达45.0和45.4个百分点。仅GPT-4o在宏观F1分数上超过了无视觉输入的场景多数策略（58.3%），但所有模型安全准确率均未超过56.2%。视觉输入将安全准确率提升了11.2至30.2个百分点。

Insight: 论文的创新点在于构建了首个专注于安全关键场景下手势理解的基准SafeGesture，揭示了视觉语言模型中感知与安全推理的显著解耦现象。客观分析表明，模型性能的主要瓶颈在于场景条件下的安全推理能力，而非手势识别本身，这为未来模型在细粒度、安全敏感任务上的改进指明了方向。

Abstract: Open-weight and frontier vision-language models (VLMs) perform well on general image understanding, but their ability to interpret fine-grained hand gestures in safety-critical operational contexts remains largely unexamined. We introduce SafeGesture, a benchmark that evaluates whether a model can infer scenario-appropriate safety actions from hand gestures. It pairs six HaGRID gestures with eight operational scenarios for 4,800 items and evaluates Qwen2.5-VL-7B, LLaVA-NeXT-7B, InternVL2-8B, Phi-3.5-Vision, and GPT-4o. Results reveal a perception-reasoning decoupling: GPT-4o achieves 98.4% gesture accuracy but 53.3% safety accuracy, while Qwen2.5-VL reaches 84.9% and 39.5%, yielding gaps of 45.0 and 45.4 percentage points. Four of five models rarely or never use the uncertainty label, and failure directions differ substantially across models. Accuracy also obscures label bias: a scenario-majority policy with no visual input reaches 58.3%, above every evaluated model, while only GPT-4o exceeds this prior under macro-F1. Visual input improves safety accuracy by 11.2 to 30.2 percentage points, but providing the ground-truth gesture as text improves performance by only 0.4 to 3.2 points, and no model exceeds 56.2%. These results indicate that the main bottleneck is scenario-conditioned safety reasoning rather than gesture recognition.

[121] TokenSTFormer: A Tokenized Spatial-temporal Attention Model for Holistic Motion Analysis in Adolescent Idiopathic Scoliosis Screening cs.CV | cs.AI | cs.LGPDF

Dong Chen, Kenneth M. C. Cheung

TL;DR: 本文针对青少年特发性脊柱侧凸（AIS）筛查，提出了一个名为TokenSTFormer的新型模型，该模型通过标记化时空语义来增强特征表示和收敛性。研究还构建了ScoliGait数据集，包含1,516个步态视频片段及其对应的X光记录。模型在关键指标上达到了最先进的性能，准确率为0.79，展示了利用步态视频的整体运动特征和基于注意力的模型进行可扩展、成本效益高的AIS筛查的潜力。

Details

Motivation: 传统AIS筛查方法存在主观解释、依赖专业知识和可扩展性低等局限性，为解决这些问题，研究旨在开发一种基于计算机视觉的自动化筛查方法。

Result: TokenSTFormer在ScoliGait数据集上实现了最先进的性能，准确率达到0.79，超越了原始的Vision Transformer编码器。

Insight: 创新点在于提出了一个标记化时空注意力的模型（TokenSTFormer），通过将时空语义标记化来改进特征学习；同时，构建了首个结合步态视频与X光记录的AIS筛查数据集（ScoliGait），为利用整体运动特征进行筛查提供了新途径。

Abstract: Adolescent Idiopathic Scoliosis (AIS) is a prevalent spinal deformity in adolescents that, if left untreated, can result in severe health outcomes. Traditional screening methods are limited by subjective interpretation, reliance on professional expertise and low scalability. To address these challenges, we present ScoliGait dataset, which comprises 1,516 gait video clips paired with corresponding X-ray records. We also introduce TokenSTFormer, a novel model that tokenizes spatial and temporal semantics to enhance feature representation and convergence. Our model achieves state-of-the-art performance, surpassing vanilla Vision Transformer encoder across key metrics, including accuracy of 0.79. This study highlights the potential of leveraging holistic motion features derived from gait video and attention-based models for scalable, cost-effective AIS screening, paving the way for future clinical applications in scoliosis detection.

[122] Graph Neural Assisted Actor-Critic for Latency-Efficient Edge Vision System cs.CV | cs.AIPDF

Alam Noor, Luis Almeida, Kai Li, Jiyan Wu, Miguel Gutiérrez Gaitán

TL;DR: 本文提出了一种图卷积神经网络辅助的深度强化学习系统模型（GCN-Assisted A2C），用于优化无人机边缘视觉系统中的视频传输延迟。该方法通过识别可疑对象的最优像素相关区域，仅传输帧中的子组像素相关区域而非整个视频帧，从而降低延迟。结合拉格朗日对偶形式和梯度下降来确保收敛性并避免约束违反。

Details

Motivation: 无人机机载视觉系统在监控等任务中需将视频流传输至地面服务器，但现有视频传输技术通常导致显著的延迟，影响操作员辅助的实时性。因此，需要一种方法在保证检测准确性的同时优化传输延迟。

Result: 实验结果表明，与其他深度强化学习模型和先进模型相比，GCN-Assisted A2C在无人机视觉系统中有效降低了视频帧传输延迟和误检率。

Insight: 创新点在于将图卷积神经网络（GCN）与A2C深度强化学习模型结合，利用GCN探索像素特征相关组的隐藏表示，并监督A2C模型选择子组像素区域以优化传输延迟。此外，采用拉格朗日对偶与梯度下降相结合的方法，解决了延迟优化中的收敛问题和约束违反问题，为边缘视觉系统的低延迟传输提供了新思路。

Abstract: UAV on-board vision systems are widely used for different activities, including monitoring in no-fly zones. In this case, the vision-equipped UAV streams a video to a ground server where an operator assists its activities. The latency of video transmission has a profound impact on the effectiveness of the operator assistance. However, most techniques available for video transmission still incur significant latency costs. In this paper, we propose a graph convolutional neural network-assisted (GCN-Assisted A2C) deep reinforcement learning (DRL) system model to find the optimal pixel-correlated area of a suspicious object. We combine the Lagrangian dual form with gradient descent to prevent lack of convergence and over- and under-penalization constraint violation during latency optimization. The proposed system model sends a sub-group pixel-correlated area of the frame from the UAV to the server rather than the transmission of the whole video frame. The proposed framework utilizes the GCN model to explore hidden representations of feature-correlated groups of pixels. Moreover, the GCN supervises the A2C model, which selects a subgroup to enhance transmission latency, thus supervising the training of UAV actions in A2C. Experimental results show that GCN-assisted A2C reduces video frame transmission latency together with false detection rate in UAV vision systems over other DRL and state-of-the-art models.

[123] Matched Outcomes, Divergent Gaze: How Foveated MLLMs Search Compared to Humans cs.CV | cs.AI | cs.CL | cs.HC | cs.MMPDF

Mohamed Amine Kerkouri, Marouane Tliba, Aladine Chetouani, Ulas Bagci, Alessandro Bruno

TL;DR: 本研究比较了三种通用多模态大语言模型（MLLMs）与人类在目标导向视觉搜索任务（COCO-Search18）中的眼动扫描路径。研究发现，在目标存在判断和目标获取效率上，模型表现达到或超过人类水平，但模型的注视过程（扫描路径）与人类存在本质差异，表现为低熵、大幅度、高度自一致的扫描模式，而非人类典型的序列化搜索过程。

Details

Motivation: 研究动机在于探究多模态大语言模型在接收与人类相同的中央凹输入时，其视觉搜索过程是否与人类相似，这关系到模型能否作为人类视觉的可靠模型以及注意力对齐评分的有效性。

Result: 在COCO-Search18基准测试中，模型在目标检测（接近天花板水平）和目标获取效率（首次眼跳即到达目标的频率高于人类）上匹配或超越人类表现，但扫描路径的熵值低、幅度大、自一致性远高于人类间一致性，表明搜索过程非人类化。

Insight: 创新点在于将模型评估解耦为决策、目标获取效率和注视过程三个维度，揭示了当前基于答案对齐和显著性的度量无法捕捉到搜索过程的时序差异。这表明零样本模型适用于结果和空间问题，但不适用于需要模拟人类时序搜索过程的任务。

Abstract: Human visual search is serial: the fovea must land on a candidate to confirm it, and those landings form a scanpath. Whether multimodal large language models (MLLMs), given the same foveated input, search as humans do bears on their use as models of human vision and on attention-alignment scores. We compare three general-purpose MLLMs with human eye-movement scanpaths on goal-directed search (COCO-Search18), driving each model fixation by fixation through an identical, human-matched foveated view and assessing it along three axes: the decision of target presence, the efficiency of reaching the target, and the gaze process itself. The axes dissociate. On the decision and on target acquisition the models match or exceed humans, detecting present targets near ceiling and reaching them on the first saccade more often than people do. The gaze process is not human. Under the human-matched condition, all three share one signature: low-entropy, large-amplitude, self-consistent scanpaths that agree with themselves far more closely than two humans agree with each other. That is consistent with a single-pass, non-serial architecture rather than a limit of acuity. Matched retinal input reproduces where humans look but not how the looking unfolds in time, and no degradation regime recovers human-like search at human-like success. The gap sits on a process axis that answer-alignment and saliency metrics do not measure. Because they miss it, such metrics cannot certify human-like vision, and zero-shot models suit outcome and spatial questions but not temporal, process-level ones.

[124] KeyID: Decoupled Drafting and Keyframe Editing for Identity-Preserving Video Generation cs.CV | cs.GR | cs.MMPDF

Jianjie Luo, Yiming Zhong, Haoming Shen, Yupeng Xiao, Zhenguo Yang

TL;DR: KeyID是一个无需训练的、身份保持视频生成框架，通过将视频动态合成与身份注入解耦来解决现有方法在复杂长序列动作中难以保持身份一致性的问题。它包含两个组件：参考感知视频生成（生成与参考对齐但身份无关的视频草稿）和身份保持关键帧编辑（通过稀疏关键帧校正和运动插值注入目标身份）。

Details

Motivation: 现有身份保持视频生成方法存在高调优成本或输入级增强有限的问题，在复杂长序列动作中难以维持严格的身份一致性，KeyID旨在解决这些限制。

Result: 在ACM Multimedia 2026 IPVG Grand Challenge的Track 2（顺序动作）中获得了亚军，在官方挑战基准上通过自动和人工评估验证了其优于先前工作。

Insight: 创新点在于从密集帧级监督转向稀疏关键帧级精化，有效解决了提示遵循与身份保真度之间的容量冲突；其模块化设计允许无缝扩展到多主体参考和复杂顺序动作生成，无需额外训练。

Abstract: Identity-preserving video generation (IPVG) requires synthesizing videos that are faithful to both reference subjects and text prompts. Existing methods are often hindered by high tuning costs or limited input-level enhancements, struggling to maintain rigid identity consistency during complex, long-sequence actions. To address these limitations, we propose KeyID, a training-free IPVG framework that decouples the synthesis of video dynamics from the injection of identity. Specifically, KeyID comprises two components: (1) Reference-Aware Video Generation, which produces an identity-agnostic video draft aligned with multiple references, and (2) Identity-Preserved Keyframe Editing, which integrates the target identity via sparse keyframe correction and subsequent motion interpolation. By shifting from dense frame-level supervision to sparse keyframe-level refinement, KeyID effectively resolves the capacity conflict between prompt adherence and identity fidelity. Crucially, our modular design allows seamless extension to multi-subject references and complex sequential action generation without additional training. KeyID outperforms prior works and is validated by automatic and human evaluations on the official challenge benchmark, ultimately securing the runner-up position in the Track 2 (Sequential Action) of the ACM Multimedia 2026 IPVG Grand Challenge. Source code is available at https://github.com/WISLab-GDUT/KeyID.

[125] Beyond Clear Skies: Synthetic Seasonal and Weather Variations for Real-World Drone Detection cs.CVPDF

Tamara R. Lenhard, Andreas Weinmann, Tobias Koch

TL;DR: 本文介绍了SynDroneVision-Weather (SDV-W)数据集，这是一个针对无人机检测任务、系统性地扩展了SynDroneVision (SDV)的合成数据集。它旨在通过可控的合成渲染技术，生成涵盖多种恶劣天气（如雨、雪、雾）和季节变化的高分辨率标注图像，以解决真实世界数据中此类条件覆盖不足的问题，从而提升无人机检测模型在复杂环境下的鲁棒性。

Details

Motivation: 在真实世界部署中实现可靠的无人机检测，需要覆盖包括恶劣天气和季节变化在内的完整操作设计域的训练数据。然而，大规模获取和标注此类数据成本高昂，因为恶劣天气条件难以控制、复现和系统性地采样，导致现有数据集对此类条件的覆盖非常有限。

Result: 实验表明，在代表性的YOLO模型和真实世界数据集上，使用SDV-W数据能够提升检测器在恶劣外观变化下的可靠性，减少漏检和误报。SDV-W最有效的使用方式是作为通用合成无人机检测数据的补充。

Insight: 论文的核心创新点是提出了一个系统性的、可控的合成数据生成框架（SDV-W），专门针对恶劣天气和季节变化这一特定领域偏移。其客观价值在于通过保留原始SDV的场景和轨迹配置，实现了“干净-恶劣”条件的配对比较，从而能够量化特定条件对检测器性能的退化影响，为研究环境鲁棒性提供了标准化的评估基准。

Abstract: Reliable drone detection under real-world deployment conditions requires training data that spans the full operational design domain, including adverse weather and seasonal appearance variation. However, acquiring and annotating such data at scale remains highly resource-intensive, as adverse-weather conditions are inherently difficult to control, reproduce, and sample systematically. Existing datasets therefore typically provide only limited coverage of such conditions. Conversely, synthetic data offers a scalable alternative: environmental variation becomes controllable, while modern game-engine-based pipelines provide realistic rendering and automatic annotations. Leveraging this potential, we introduce SynDroneVision-Weather (SDV-W), an systematic extension of SynDroneVision (SDV) targeting adverse-weather and seasonal domain shifts in urban drone detection. SDV-W comprises 55,187 annotated high-resolution images from three urban environments, rendered across three seasonal configurations and diverse weather conditions, including rain, snow, and fog at multiple severity levels. By preserving SDV’s scene and trajectory configuration, SDV-W enables matched clean-adverse comparisons and quantification of condition-specific detector degradation. Across representative YOLO models and real-world datasets, we show that SDV-W improves detector reliability under adverse appearance shifts, reduces missed detections and false alarms, and is most effective as a complement to general-purpose synthetic drone-detection data. SDV-W will be publicly released upon paper acceptance.

Tianchen Deng, Xuefeng Chen, Shuang Wu, Qu Chen, Jiajun Zhu

TL;DR: 本文提出了GaussianDWM++，一种基于语言接地的3D高斯驾驶世界模型，旨在统一场景理解、编辑和多模态生成。该方法通过基础特征高斯分词器将视觉语言特征蒸馏到3D高斯基元中，构建紧凑的开放词汇高斯语义场，并利用几何感知高斯适配器将密集高斯基元聚合为世界令牌，同时引入基于KL的高斯-图像分布对齐目标以提升表示兼容性。

Details

Motivation: 现有驾驶世界模型主要关注条件场景生成，缺乏显式的3D场景理解、语言接地推理和可控的4D编辑能力，且常用点云、占据或BEV表示难以实现文本信息与底层3D场景结构的细粒度对齐。

Result: 在广泛的驾驶基准测试中，该方法在场景理解、视觉接地、规划导向推理和可控4D生成任务上均达到了最先进的性能。

Insight: 创新点包括基础特征高斯分词器实现视觉语言特征到3D高斯基元的直接蒸馏，几何感知高斯适配器结合重要性感知分层选择与文本条件交叉注意力，以及KL对齐目标提升高斯世界令牌与基础图像令牌的兼容性，从而支持指令可控的场景编辑。

Abstract: Driving World Models (DWMs) have recently advanced rapidly with generative models, yet most existing methods mainly focus on conditional scene generation and lack explicit 3D scene understanding, language-grounded reasoning, and controllable 4D editing capabilities. Moreover, commonly used point cloud, occupancy, or BEV representations make it difficult to achieve fine-grained alignment between textual information and the underlying 3D scene structure. To address these limitations, we propose a foundation-feature Gaussian driving world model that unifies scene understanding, language-grounded reasoning, controllable 4D editing, and multi-modal generation within a single framework. Specifically, we introduce a foundation-feature Gaussian tokenizer that directly distills Qwen/SigLIP visual-language features into 3D Gaussian primitives, building a compact open-vocabulary Gaussian semantic field. We further design a geometry-aware Gaussian adapter that combines importance-aware hierarchical selection with text-conditioned Perceiver-style cross-attention to aggregate dense Gaussian primitives into compact world tokens. To improve representation compatibility, we introduce a KL-based Gaussian–image distribution alignment objective that aligns Gaussian world tokens with foundation image tokens. Based on the aligned Gaussian representation, our framework further supports instruction-controllable scene editing, including weather-conditioned generation and dynamic vehicle manipulation. Extensive experiments on broader driving benchmarks demonstrate that our method achieves state-of-the-art performance across scene understanding, visual grounding, planning-oriented reasoning, and controllable 4D generation tasks. We will release the code and datasets publicly on Github.

[127] Seeing Before Answering: Training-Free Visual Layer Profiling for Vision-Language Models cs.CVPDF

Ruchen Liu, Yi Yang, Yiming Xu, Michael Ying Yang, Monika Sester

TL;DR: 本文研究发现，LLaVA风格的视觉语言模型（VLMs）默认从视觉主干网络的固定层（通常是倒数第二层）提取视觉令牌传递给语言模型的做法是脆弱的。作者提出了一种无需训练、基于表示几何的视觉层性能预测方法，即视觉数据集熵（VDE），它仅需少量未标注样本即可有效识别最优视觉层，从而避免昂贵的逐层推理搜索。

Details

Motivation: 动机在于发现现有VLMs中传递视觉令牌的默认层选择是次优的，且最优层会随任务和视觉主干变化。通过穷举搜索寻找最优层成本过高，因此需要一种高效、无需训练的方法来预测不同视觉层的有效性。

Result: 在2个VLMs和7个图像/视频基准测试的14个模型-任务对中，默认层在13个情况下表现不佳。提出的VDE方法（基于预投影器侧计算）能够准确追踪逐层精度，其排名靠前的层在SigLIP-based LLaVA-Video上覆盖了所有任务的最优层，并为CLIP-based Video-LLaVA提供区域级指导。

Insight: 主要创新点在于将单模态层分析中的矩阵熵度量（即VDE）成功迁移并应用于冻结视觉编码器的VLM场景，用于预测视觉层性能。研究发现，视觉几何信息在跨模态投影后依然保留与性能相关的趋势，使得基于少量样本的VDE成为一种可解释、无需训练的有效选择策略。

Abstract: LLaVA-style Vision-Language Models (VLMs) pass visual tokens from a fixed late layer of the vision backbone, typically the penultimate one, to the language model. We first show that this hidden convention is fragile: across 2 VLMs and 7 image and video benchmarks, the default layer is sub-optimal in 13 of 14 model-task pairs, and the best layer shifts with both task and visual backbone. Finding that layer by exhaustive layer-wise inference is prohibitively expensive, and no better fixed default exists. We therefore ask whether layer usefulness can instead be predicted from representation geometry. We study matrix-based entropy, introduced for unimodal layer analysis, which we compute over sample-level visual embeddings as Visual Dataset Entropy (VDE); and Gromov-Wasserstein (GW) distance, introduced for encoder-level VLM model selection, which we repurpose as a layer-wise visual–language alignment signal. Transferring these to LLaVA-based models is not obvious a priori: the vision tower is frozen while the multimodal projector is trained, so we profile both sides of the projector. We find that VDE transfers, and GW does not. Computed from 100 unlabeled task samples without downstream inference, pre-projector VDE tracks layer-wise accuracy and its top-ranked layers cover the oracle best layer on every task for the SigLIP-based LLaVA-Video, while giving region-level guidance for the CLIP-based Video-LLaVA. Post-projector profiles show that the projector reshapes visual geometry but does not erase the performance-relevant trend, leaving $\mathrm{VDE}_{\mathrm{pre}}$ the stronger signal. GW instead flattens after projection and is best read as an alignment diagnostic rather than a selector. VDE thus offers an interpretable, training-free policy that narrows the visual-layer search to a handful of candidates for limited downstream verification.

[128] CoM$^3$eT: A foundation model for medical image analysis through federated, multidimensional context integration cs.CV | cs.LGPDF

J. Raphael Schäfer, Kai Geissler, Till Nicke, Chiara Tappermann, Karoline Heber

TL;DR: CoM$^3$eT是一个医学视觉基础模型，通过注意力机制建模多维上下文，统一了病理学和放射学、稀疏与密集预测任务以及二维与高维输入。它在涵盖多种数据集和任务的公开竞赛中表现优异，并通过参数高效微调和联邦学习实现了高性能且资源友好的部署。

Details

Motivation: 现有医学基础模型通常局限于单一专业领域（如病理学或放射学）和单一输出类型（如分类或分割），缺乏通用性。本文旨在开发一个能统一不同医学专业、任务和输入维度的通用医学视觉基础模型。

Result: 在涵盖5个断层扫描、4个全标本和3个二维数据集的公开竞赛中，CoM$^3$eT在稀疏/密集预测和报告生成任务上超越了其他医学基础模型。仅微调少于2.5%的参数即可达到与全微调相当的性能，且在联邦学习设置下，其性能可与集中式数据训练相媲美。

Insight: 主要创新点在于通过统一的注意力机制架构整合多维上下文，实现了跨医学专业和任务类型的泛化。其参数高效微调策略和联邦学习兼容性为在资源受限环境中部署高性能医学AI模型提供了实用方案。

Abstract: Medical foundation models improve generalization when training AI models with limited labeled data, but remain confined to a single specialty, such as pathology or radiology, and to either sparse or dense outputs, such as classification or segmentation. Here, we present CoM$^3$eT (Co-representation Multidimensional Multitask Medical Transformer), a medical vision foundation model that unifies pathology and radiology, sparse and dense predictions, and two- and higher-dimensional inputs by modeling multidimensional context with attention. CoM$^3$eT outperformed other medical foundation models in an open competition spanning five tomographic, four whole-specimen, and three two-dimensional datasets, covering sparse and dense prediction tasks as well as report generation. When adapted across diverse clinical applications, training fewer than 2.5% of parameters achieved performance comparable to full fine-tuning, enabling research without access to high-performance GPU clusters. Applied to federated learning across hospitals, this approach achieved performance comparable to pooled-data training over internet connections and with consumer-grade hardware.

[129] TransAnyText: Translating Arbitrary Text in E-commerce Images via Structured Visual Generation cs.CVPDF

Xiaoan Liu, Lichen Ma, Zipeng Guo, Yu He, Xiaoyan Su

TL;DR: TransAnyText是一个用于电商图像文本翻译的结构化视觉代码框架，它将图像文本翻译任务重新定义为从源图像和目标语言生成可渲染的HTML补丁。该框架通过视觉语言模型（VLM）处理视觉理解、跨语言翻译和结构化视觉生成，并利用扩散模型进行背景修复和像素级细化，最终通过确定性渲染合成最终图像。论文还提出了一个三阶段的后训练框架（SFT、PWSD、RLVR）来优化模型，并引入了TransAnyDataset和TransAnyBench数据集与基准。

Details

Motivation: 解决跨境电商中产品图像、横幅和详情页的多语言翻译需求，现有方法难以同时实现准确翻译、忠实视觉身份保持和易于编辑的输出。

Result: 在电商图像翻译任务上，与级联流水线、开源端到端模型和闭源图像编辑系统相比，TransAnyText展现出具有竞争力的性能，提供了有效、可控和可编辑的解决方案。

Insight: 创新点在于将图像文本翻译重构为生成可渲染的HTML补丁的结构化视觉代码框架，实现了语义生成与像素渲染的解耦，并通过三阶段后训练框架（SFT、PWSD、RLVR）优化任务级性能，提升了翻译准确性和编辑便利性。

Abstract: Cross-border e-commerce image translation is essential for global retail, where product images, banners, and detail pages need to be produced in different languages. Existing methods struggle to achieve accurate translation, faithful visual identity preservation, and easy-to-edit outputs, simultaneously. To address these challenges, we introduce TransAnyText, a structured visual code framework that reformulates image text translation as generating renderable HTML patches from source images and target languages. Our framework decouples semantic generation from pixel rendering: a vision-language model (VLM) handles visual understanding, cross-lingual translation, and structured visual generation, while a diffusion model performs background inpainting and pixel-level refinement, followed by deterministic rendering to synthesize the final image. Based on this formulation, we develop a three-stage post-training framework, where supervised fine-tuning (SFT) establishes the image-to-code mapping, privilege-gap weighted self-distillation (PWSD) improves the learning of style and layout tokens, and reinforcement learning with verifiable rewards (RLVR) further optimizes task-level performance. We further introduce TransAnyDataset and TransAnyBench, a multilingual dataset and benchmark for e-commerce image translation. Extensive experiments demonstrate competitive performance against cascaded pipelines, open-source end-to-end models, and closed-source image editing systems, providing an effective, controllable, and editable solution for cross-border e-commerce image translation.

[130] Defake-o3: From Speculative Rationales to Verifiable Evidence for Explainable AIGI Detection cs.CV | cs.AIPDF

Bowen Deng, Jiahui Zhan, Yikun Ji, Haozhen Yan, Jianfu Zhang

TL;DR: 本文提出了Defake-o3，一种可解释的AI生成图像检测器，旨在从推测性解释转向可验证的证据。该方法结合了交互式视觉搜索和验证器引导的证据对齐，通过迭代放大可疑区域来检查细节，并使用基于人类验证标注训练的‘证据验证器’来提供强化学习奖励。为支持此目标，作者构建了包含局部边界框证据的GroundFake数据集和基于10种最新生成器的FakeFrontier分布外基准。

Details

Motivation: 现有基于MLLM的AI生成图像检测器提供的解释往往是推测性的，依赖于模糊或幻觉的伪影，无法捕捉最新生成器的细微局部缺陷，且缺乏可视觉验证的证据，因此需要开发既准确又可解释、可靠的检测器。

Result: 在GroundFake、FakeFrontier及其他分布外基准上的实验表明，Defake-o3在检测准确性和解释质量上均有提升，能产生更局部化、可验证且更具说服力的证据。

Insight: 创新点在于将交互式视觉搜索与基于人类验证的强化学习奖励机制结合，推动检测器生成可验证的证据；同时构建了专门用于可解释检测的数据集和评估协议，为可解释AI检测提供了新的方法论和基准。

Abstract: The rapid progress of image generation models calls for AI-generated image (AIGI) detectors that are not only accurate but also explainable and reliable. While MLLM-based detectors can provide natural language explanations, existing methods often generate speculative rationales: they rely on vague or hallucinated artifacts, miss subtle localized flaws from the latest generators, and fail to provide evidence that can be visually verified. We present Defake-o3, an explainable AIGI detector that moves from speculative rationales to verifiable evidence. It combines interactive visual search with verifier-guided evidence alignment: the model iteratively zooms into suspicious regions to inspect fine-grained details, while an Evidence Verifier, trained from human verification annotations, provides reinforcement learning rewards that favor grounded evidence and penalize baseless claims. To support this objective, we construct GroundFake, a dataset designed for grounded explainable detection, with localized bounding-box evidence, human verification based on visual grounding and artifact specificity, corrected reasoning trajectories, and valid/invalid evidence supervision. We further introduce FakeFrontier, an out-of-distribution benchmark built from real images and outputs of 10 recent generators, together with an MLLM-based protocol for evaluating evidence quality and persuasiveness. Experiments on GroundFake, FakeFrontier, and additional out-of-distribution benchmarks show that Defake-o3 improves both detection accuracy and explanation quality, producing more localized, verifiable, and persuasive evidence.

[131] Audio-Visual Segmentation via Depth-Guided Collaborative Modeling cs.CV | cs.AIPDF

Zhaojin Fu, Yuyang Hong, Qi Yang, Zili Wang, Kun Ding

TL;DR: 本文提出了一种名为DGCM-AVS的三模态框架，用于音频-视觉分割任务。该框架通过引入估计的深度信息作为空间结构线索，联合建模音频、视觉和深度信息，以提升视频中发声物体的像素级分割性能。

Details

Motivation: 现有音频-视觉分割方法大多未显式建模相对距离、遮挡等几何线索，限制了跨模态对齐的鲁棒性。受人类感知中空间结构与视听证据自然整合的启发，本文旨在将深度信息作为空间结构线索引入AVS任务。

Result: 在AVS基准数据集AVSS上，DGCM-AVS相比现有最优方法，在M_J和M_F指标上分别实现了10.2%和8.7%的相对提升，达到了新的SOTA水平。

Insight: 创新点在于首次将深度作为显式空间结构线索引入AVS任务，并设计了深度感知动态调制器和深度引导渐进融合模块，以改善相邻物体分离并促进音频与视觉特征的渐进对齐。这为多模态感知研究提供了一个有前景的新方向。

Abstract: Audio-Visual Segmentation (AVS) is a fundamental task in multimodal perception that performs pixel-level segmentation of sounding objects in videos by leveraging both visual and audio cues. It has broad applications in video understanding, human-computer interaction, and autonomous driving. However, most existing AVS methods do not explicitly model geometric cues such as relative distance and occlusion, thereby limiting the robustness of cross-modal alignment. In human perception, spatial structure is naturally integrated with audio-visual evidence to accurately localize sounding objects. Motivated by this, we incorporate estimated depth as a spatial structural cue for AVS and propose DGCM-AVS, a tri-modal framework that jointly models audio, visual, and depth information. Specifically, we design a Depth-Aware Dynamic Modulator to improve the separation of adjacent objects while preserving intra-object feature consistency. Furthermore, we propose Depth-Guided Progressive Fusion, which uses depth as an intermediate bridge to progressively align audio cues with visual features. Compared to state-of-the-art methods, DGCM-AVS achieves relative improvements of 10.2 percent in M_J and 8.7 percent in M_F on the AVSS dataset. We believe our study highlights depth as a promising yet underexplored modality for AVS and may encourage further research in this direction.

[132] StreamOPD: A Post-Training Recipe with Spatio-Temporal Cue Gating for Streaming Video Understanding cs.CVPDF

Keming Wu, Baoyi Wang, Kaichen Zhang, Xiang An, Zuhao Yang

TL;DR: 本文提出StreamOPD，一种用于流式视频理解的后训练方法，结合了可验证的流式视频数据、思维模式下的策略蒸馏和指令模式部署。该方法在StreamingBench上将准确率从77.9%提升至83.9%，接近9B教师模型的性能，并在OVO-Bench（排除幻觉检测子任务）上提升了9.1个百分点。

Details

Motivation: 现有流式视频理解系统依赖推理时内存、检索和压缩，但训练无关的滑动窗口基线已能匹配其性能，因此研究聚焦于无内存的近期窗口协议，探索仅通过后训练能达到的性能上限。

Result: StreamOPD在StreamingBench上达到83.9%准确率（较基线提升6个百分点），接近9B教师模型；在OVO-Bench（排除幻觉检测）上提升9.1个百分点；其扩展ST-CueGate在OVO-Bench（排除幻觉检测）和Video-MME上分别达到71.9%和64.9%，在所有四个基准测试中均优于基础模型。

Insight: 创新点包括：结合可验证流式视频数据与思维模式策略蒸馏的StreamOPD后训练框架；引入时空线索门控（ST-CueGate）通过教师似然比聚合实现响应重加权；策略自蒸馏能保留大部分性能增益并提升幻觉检测能力，表明方法具有可扩展性和鲁棒性。

Abstract: Streaming video understanding demands direct responses from the causally observed prefix of an unfolding video. Existing systems add inference-time memory, retrieval, and compression, yet a training-free sliding-window baseline already matches them. We therefore fix a memory-free recent-window protocol and ask how far post-training alone can go. Reinforcement learning with verifiable rewards fits this regime poorly, encouraging long ``think-then-answer’’ generations, while on-policy distillation (OPD) supplies dense token-level teacher supervision on student trajectories but is stable only when both models train in thinking mode. These observations lead to \textsc{StreamOPD}, a recipe combining verifiable streaming-video data, thinking-mode OPD, and instruct-mode deployment. It raises StreamingBench from $77.9%$ to $83.9%$—within $0.3$ points of the 9B teacher—and improves OVO-Bench excluding its hallucination-detection subtask (HLD) by $9.1$ points under unchanged inference. As a teacher-privilege extension, \emph{Spatio-Temporal CueGate (ST-CueGate)} aggregates cue-versus-no-cue teacher likelihood ratios into a group-relative response score that reweights OPD. It reaches $71.9%$ on OVO-Bench (excluding HLD) and $64.9%$ on Video-MME, and is the only variant that stays above the base model on all four benchmarks. Replacing the teacher with a frozen copy of the student’s initial policy—on-policy self-distillation—retains most of these gains and lifts HLD to $57.0%$, above both the untrained student and the 9B teacher, so abstention loss is not intrinsic to the recipe. We provide a transparent and reproducible reference for open-source streaming-video research.

[133] PosterText: Towards Unified Visual Text Generation and Editing for E-commerce Poster cs.CVPDF

Xiaoan Liu, Lichen Ma, Zipeng Guo, Yu He, Xiaoyan Su

TL;DR: 本文提出PosterText框架，旨在统一电商海报的视觉文本生成与编辑任务。该框架将文本块视为原子单元，支持海报生成、添加、删除和修改四种操作，并可结合参考风格控制。通过四阶段课程学习（包括文本渲染预训练、指令跟随训练、偏好对齐强化学习和空间引导自蒸馏）进行模型训练，并在构建的大规模数据集上验证了其有效性。

Details

Motivation: 现有电商海报设计方法多专注于端到端生成或多阶段流水线，缺乏对现有海报进行灵活且精确编辑的能力。本文旨在统一生成与编辑任务，提升设计的灵活性和可控性。

Result: 在构建的大规模带文本块标注的数据集和综合评测基准上，大量实验表明PosterText在生成和编辑任务上均取得了与现有方法相竞争的性能，验证了框架的有效性。

Insight: 创新点在于将文本块作为原子单元的统一任务定义（Text Patch Generation and Editing），以及结合四阶段课程学习的训练策略，特别是空间引导自蒸馏用于执行细化，这为统一视觉内容生成与编辑提供了新思路。

Abstract: Automated e-commerce poster design requires both high-quality poster generation and flexible editing of existing designs. However, most existing methods either target end-to-end poster generation or follow multi-stage design pipelines, with limited capability for flexible and precise editing of existing posters. To enable unified generation and editing of e-commerce posters, we introduce Text Patch Generation and Editing, a unified task formulation that treats text patches as atomic units and covers four operations: poster generation, patch addition, patch deletion, and patch modification, with optional reference-guided style control. Based on this, we propose PosterText, a unified model trained with a four-stage curriculum, including text rendering pretraining, instruction-following training, reinforcement learning for preference alignment, and spatial guidance self-distillation for execution refinement. We further construct a large-scale dataset with patch-level annotations and a comprehensive benchmark for evaluation. Extensive experiments demonstrate that PosterText achieves competitive performance against existing generation and editing approaches, validating the effectiveness of the proposed framework.

[134] LaGSplat: Inferring Physics-Governed Interactive Simulation from Monocular Video Using Latent Lagrangian Gaussian Splatting cs.CV | cs.LGPDF

Louen Pottier

TL;DR: LaGSplat（Latent Lagrangian Gaussian Splatting）是一个框架，能够从一个或多个单目视频中推断出受物理约束的交互式动态。它允许用户在推理时对拍摄的物体（无论是刚体还是可变形体）施加训练中从未测量、标注或见过的外力。该框架通过将低维潜在状态同时作为学习耗散拉格朗日量的广义坐标和高斯泼溅解码器的条件变量来实现这一目标，并支持实时渲染物体的2D或3D响应。

Details

Motivation: 解决从单目视频中推断物理交互动态的挑战，特别是如何让模型在训练中未见过外力的情况下，仍能对用户施加的任意外力做出合理响应，而传统基于像素空间（CNN）或神经场（NeRF）的解码器无法直接处理这种外力交互。

Result: 在从刚体到可变形体、从自主系统到受迫真实系统的递增难度测试案例上进行了验证，结合了单目视频和传感器测量。实验表明，该方法能够对未见外力产生有界且合理的响应，而未经约束的预测器则会发散。

Insight: 创新点在于将潜在状态双重角色化（作为拉格朗日广义坐标和高斯泼溅解码器条件），并通过显式移动点的高斯泼溅解码器归纳偏置，实现外力从图像空间到潜在广义力的映射，从而在有限广义坐标的耗散欧拉-拉格朗日方程约束下，平衡了通用性与对未见外力的合理响应。

Abstract: We present LaGSplat (Latent Lagrangian Gaussian Splatting), a framework that infers interactive, physics-governed dynamics from one or a few monocular videos. At inference it lets a user push on the filmed object, rigid or deformable, with an external force that was never measured, annotated, or seen during training. This is possible because a low-dimensional latent state $\mathbf{q} \in \mathbb{R}^d$ plays two roles at once: it is the generalised coordinate of a learned dissipative Lagrangian and the conditioning variable of a Gaussian Splatting decoder. The inductive bias of this decoder, whose primitives are explicit points $μ_i(\mathbf{q})$ that move with the object, is what lets a force $f$ applied in the image pull back into a latent generalised force $J(\mathbf{q})^\top f$ and enter the equations of motion, which pixel-space (CNN) or neural-field (NeRF) decoders cannot do. We validate LaGSplat on test cases of increasing difficulty, from rigid to deformable and from autonomous to forced real systems, combining monocular video and sensor measurements. We further demonstrate interactive use: forces of arbitrary magnitude and direction can be applied to the reconstructed object at any time, its response rendered in real time, in 2D or 3D. Assuming a dissipative Euler-Lagrange equation over a few generalised coordinates trades generality for a bounded, plausible response to unseen forces, where an unconstrained predictor diverges.

Feng Xie, Jiagao Hu, Fuhao Li, Zepeng Wang, Yuxuan Chen

TL;DR: GRNEdit是一个轻量级两阶段通用视频编辑框架，从二进制证据的新视角出发，将编辑语义重新定义为对单个比特的局部保留或翻转决策。它通过紧凑编码器将离散源编码转换为连续证据信号，并利用GRN主干网络进行全局组合，在仅使用0.6M训练数据和少于3%条件参数的情况下，实现了高效的视频编辑性能。

Details

Motivation: 现有基于指令的通用视频编辑方法通常依赖资源密集的条件机制（如重型分支或昂贵的源拼接），本文旨在探索一种更高效的方式来建模编辑意图。

Result: 在OpenVE-Bench基准测试中，GRNEdit-2B和GRNEdit-8B分别获得4.03和4.18分；其中2B模型性能超越多个14B开源编辑器，8B模型与领先开源编辑器表现相当。

Insight: 创新点包括：1) 将编辑语义重新定义为比特级的保留/翻转二元决策；2) 引入空指令条件作为源重建监督，强化证据利用和内容保留；3) 两阶段框架通过对比编辑状态与源保留状态的差异来修正未解决的目标比特决策，实现高效编辑。

Abstract: Instruction-based general video editing seeks to unify diverse editing operations within a single, intuitive interface. Existing approaches often rely on resource-intensive conditioning, using either heavyweight branches or costly source concatenation. Is there any efficient way to model editing intent? Thus, we introduce GRNEdit, a lightweight two-stage framework. GRN inspires our approach by encoding visual semantics through combinations of bits. Through task-specific fine-tuning, we take this representation further and recast editing semantics as local retain-or-flip decisions over individual bits. Source information is consequently modeled as coordinate-wise evidence supporting the observed binary states, while the GRN backbone remains responsible for resolving their global composition into coherent generative semantics. In Stage I, a compact encoder translates discrete source codes into continuous evidence signals, which GRN assimilates throughout binary refinement. Inspired by null-prompt training for classifier-free guidance, we further assign the null condition an editing-specific meaning: an empty instruction denotes no edit and is supervised through source reconstruction. This identity pathway not only implicitly strengthens evidence utilization and content preservation in Stage I, but also produces a source-preserving state in the same representation space as the edited state. Stage II can therefore directly compare each edited state with its source-preserving counterpart and use their discrepancy to revise unresolved target-bit decisions. Trained on only 0.6M pairs with less than 3% conditioning parameters, GRNEdit-2B and GRNEdit-8B achieve scores of 4.03 and 4.18 on OpenVE-Bench. The 2B model outperforms multiple 14B open-source editors, while the 8B model performs on par with leading open-source editors.

[136] Unlocking Motion in Expressions: Temporal Calibration for Referring Video Object Segmentation cs.CVPDF

Yiwen Jiang, Zhengtong Zhu, Ruixin Zhang, Jiaqing Fan

TL;DR: 本文提出了一种用于指代视频目标分割（RVOS）的表达式驱动运动校准（EMC）框架，旨在通过显式建模自然语言描述中的运动语义，自适应地调整运动信息在时序决策中的使用。该方法包含运动信号处理（MSP）、运动影响校准（MIC）和语义时序阶段构建（STSC）三个模块，在六个标准基准测试上验证了其优越性。

Details

Motivation: 现有RVOS方法通常在统一的跨模态时序建模框架中引入运动信息，但未显式建模表达式对运动语义的依赖，难以根据不同的语义需求自适应调整运动信息的使用。

Result: 在Ref-YouTubeVOS、Ref-DAVIS17、MeViS（valid/valid$^u$）、A2D-Sentences和JHMDB-Sentences六个标准基准测试上进行了广泛评估，验证了所提方法的优越性。

Insight: 创新点在于显式地从表达式中提取可解释的运动控制信号，并通过校准模块调整运动线索的贡献，同时构建与表达式相关的紧凑时序阶段候选空间，实现了运动语义的自适应利用。

Abstract: Referring Video Object Segmentation (RVOS) aims to segment referred objects at the pixel level in video sequences based on natural language descriptions. Existing methods typically introduce motion information within a unified cross-modal temporal modeling framework, where language cues are used for target localization and segmentation. However, the dependency of expressions on motion semantics is not explicitly modeled, making it difficult to adaptively adjust the use of motion information according to different semantic requirements. To address these issues, we propose an Expression-driven Motion Calibration (EMC) framework for RVOS that explicitly unlocks and leverages the motion semantics within expressions. The proposed method extracts interpretable motion control signals from expressions via a Motion Signal Processing (MSP) module, and employs a Motion Influence Calibration (MIC) module to adjust the contribution of motion cues during temporal decision making. In addition, a Semantic Temporal Stage Construction (STSC) module is introduced to build expression-relevant temporal stages, providing a compact temporal candidate space for motion calibration. Through extensive evaluation on six standard benchmarks, including Ref-YouTubeVOS, Ref-DAVIS17, MeViS (valid/valid$^u$), A2D-Sentences, and JHMDB-Sentences, the superiority of our method is validated. We will release the code on https://github.com/Jeven7/EMC.

[137] SIGMA-Lane: Scale-pyramId Gated MAmba for Temporally Consistent Video Lane Detection cs.CV | cs.AIPDF

Tiancheng Zhang, Mengmeng Wang, Yan Gao, Xiangjie Kong, Guojiang Shen

TL;DR: SIGMA-Lane是一种用于视频车道线检测的模型，旨在解决严重车辆遮挡导致的时间一致性中断问题。它通过在多尺度金字塔结构中引入基于状态空间模型（SSM）的遮挡感知门控机制，直接控制观测信息进入和融合到时间记忆中的过程，并结合结构空间检索（SSR）从历史先验中恢复缺失的车道结构，从而提升遮挡下的时间稳定性。

Details

Motivation: 现有视频车道检测模型在严重车辆遮挡时，损坏的观测会污染循环模型的隐藏状态，导致错误持续影响后续帧。现有方法通常仅通过提供障碍物掩码作为辅助输入来间接保护状态更新路径，未能直接处理状态污染问题。

Result: 在VIL-100和OpenLane-V基准测试上的实验表明，SIGMA-Lane在严重遮挡下显著改善了时间稳定性，同时保持了具有竞争力的F1分数和mIoU得分。

Insight: 创新点在于将遮挡问题建模为基于SSM的时间建模中的状态污染，并直接在SSM的写入和残差融合路径上设计遮挡感知门控，实现了对时间记忆的直接保护。此外，结合坐标一致仿射对齐后的SSM一致性双门控时间滤波与结构空间检索（SSR），提供了互补的遮挡鲁棒性策略。

Abstract: Video lane detection requires predictions that remain stable across frames, yet severe vehicle occlusions can break temporal cues. In streaming recurrent models, corrupted observations may enter the hidden state and produce errors that persist into later frames. Existing occlusion-aware refinements usually provide obstacle masks as auxiliary inputs, so the state-update path is only indirectly protected. We propose SIGMA-Lane, which treats this failure mode as state contamination in State Space Model (SSM)-based temporal modeling. SIGMA-Lane places occlusion-aware gates on the SSM write and residual-fusion paths, controlling how current observations enter temporal memory and are fused back after temporal propagation. After coordinate-consistent affine alignment, the model combines two complementary paths: SSM-consistent dual-gating for temporal filtering and Structural Spatial Retrieval (SSR) for recovering missing lane structure from aligned historical priors. Experiments on VIL-100 and OpenLane-V show improved temporal stability under heavy occlusion, with competitive F1 and mIoU scores.

[138] Contrastive Energy Fields for Inference-Time Procedure Planning in Instructional Videos cs.CV | cs.AIPDF

Mohamed Afham, Christoph Reich, Oliver Hahn, Daniel Cremers, Stefan Roth

TL;DR: 论文提出了一种名为CEFITO的程序规划方法，用于教学视频中的推理时间程序规划。该方法通过学习一个动作条件表示空间，将程序规划重新表述为一个任务约束优化问题，从而在推理时显式地排除不相关的动作。

Details

Motivation: 现有程序规划方法使用前馈神经网络或基于扩散的推理直接从潜在表示预测动作序列，但缺乏强制执行任务特定逻辑约束的能力，导致可能预测出不合理或不相关的动作。

Result: CEFITO在两个已建立的程序规划基准测试上实现了最先进的准确率。

Insight: 创新点在于将程序规划重新定义为任务约束优化问题，并学习一个动作条件表示空间，使得在推理时能够显式地排除不相关的动作，从而更好地强制执行任务逻辑约束。

Abstract: Procedure planning seeks to estimate a sequence of actions to transition from an observed initial state to a given goal state. Current procedure planning approaches directly predict action sequences from latent representations using feed-forward neural networks or diffusion-based inference. These paradigms treat every action as plausible, lacking the ability to enforce task-specific logical constraints that render certain actions irrelevant or not plausible. We propose CEFITO, a procedure planning approach that learns a predictor to express an action-conditioned representation space. Based on this representation space, we formulate procedure planning as a task-constrained optimization problem. Unlike prior methods, CEFITO explicitly reasons over the action space by omitting irrelevant actions during inference-time planning. This reformulation enables effective procedure planning and achieves state-of-the-art accuracy on two established procedure planning benchmarks.

[139] Synthetic Data Augmentation for Satellite-Based Analysis of Battle-Damaged Agricultural Fields in Ukraine cs.CV | cs.AIPDF

Marta Sumyk, Oleksandr Kosovan, Iryna Voitsitska

TL;DR: 该论文研究了利用合成数据增强技术来改善乌克兰受战争破坏农田的卫星图像分类问题。由于标记图像稀缺且类别不平衡，作者训练了条件生成对抗网络（GAN）和去噪扩散概率模型（DDPM）来生成额外的‘被轰炸’和‘未被轰炸’农田样本，用于增强训练数据。实验表明，基于平衡DDPM增强的最佳配置显著提升了分类器的多项性能指标。

Details

Motivation: 监测乌克兰战争对农田的破坏对于理解粮食安全威胁、环境稳定性和战后恢复至关重要，但开发基于卫星图像的计算机视觉系统受到标记数据稀缺（尤其是受损农田图像）的限制。

Result: 在仅使用真实数据的测试集上进行评估，最佳配置（基于平衡DDPM增强）将准确率从84%提升至88%，平衡准确率从67%提升至81%，宏F1分数从65%提升至78%，对少数类（未被轰炸）的召回率从41%大幅提升至69%。

Insight: 论文的创新点在于将合成数据增强（特别是DDPM）应用于数据稀缺的战争影响地区地理空间分析，有效缓解了类别不平衡问题，并证明了合成卫星图像在此类应用中的潜力。从客观角度看，该方法为小样本、不平衡的遥感图像分类任务提供了一种实用的数据增强解决方案。

Abstract: Monitoring war-induced damage to agricultural land in Ukraine is important for understanding threats to food security, environmental stability, and post-war recovery. However, the development of computer-vision systems for satellite-based damage analysis is limited by the scarcity of labeled imagery, especially for damaged agricultural fields. This work investigates synthetic data augmentation as a method for improving classification under limited and imbalanced training data. We train class-conditional Generative Adversarial Network (GAN) and Denoising Diffusion Probabilistic Model (DDPM) architectures on real satellite images and use them to generate additional bombed and not-bombed agricultural-field samples. The generated images are used only for training augmentation, while all downstream evaluation is performed on an exclusively real test set. A Vision Transformer classifier is trained under multiple real and synthetic data configurations to measure the practical utility of each generative approach. The best configuration, based on balanced DDPM augmentation, improves accuracy from 84% to 88%, balanced accuracy from 67% to 81%, macro F1 from 65% to 78%, and recall for the underrepresented not-bombed class from 41% to 69%. These results demonstrate the potential of synthetic satellite imagery for data-scarce geospatial applications in war-affected regions.

[140] Sterilizable Scene Graph Generation for Operating Rooms cs.CVPDF

Nick Lemke, Ssharvien Kumar Sivakumar, Antoine P. Sanner, John Kalkhof, Henry John Krumb

TL;DR: 本文提出了一种基于神经细胞自动机（NCA）的轻量级场景图生成框架SG-NCA，用于手术视频中的场景理解。该方法旨在解决现有大型模型在手术室部署时面临的硬件占用、卫生限制、延迟和数据隐私等问题。通过在白内障手术和胆囊切除术视频上进行评估，SG-NCA在性能与基线模型相当的同时，参数量减少了55倍，并成功部署在无风扇边缘设备上，展示了其在手术视频描述等下游应用的潜力。

Details

Motivation: 现有最先进的场景图生成方法依赖于参数量大、计算资源消耗多的深度学习模型，难以满足手术室对硬件体积、卫生协议、低延迟和数据隐私的严格要求。

Result: 在白内障手术和胆囊切除术视频数据集上的评估表明，SG-NCA的性能与现有基线模型相当，同时参数量减少了55倍，并成功部署在符合手术室卫生要求的无风扇边缘设备上。

Insight: 主要创新点在于首次将神经细胞自动机（NCA）框架应用于场景图生成任务，并提出了首个能够学习结构化表示的NCA框架。SG-NCA结合了基于NCA的多类别分割进行高效目标检测与特征提取，以及一个轻量级的关系预测器，实现了模型轻量化与部署友好性。

Abstract: Scene graph generation from surgical video enables a holistic and structured understanding of surgical scenes by modeling objects and their semantic relationships. Despite recent advances, state-of-the-art approaches rely on large, parameter-heavy deep learning models that are impractical for deployment in the operating room (OR) due to hardware footprint, hygiene constraints, latency, and data privacy concerns. To the best of our knowledge, this is the first scene graph generation method built on NCAs and the first NCA framework capable of learning structured representations. We introduce SG-NCA, a lightweight scene graph generation framework based on Neural Cellular Automata (NCA), designed for inference in fanless devices critical for OR hygiene protocols. SG-NCA is the first scene graph generation combining NCA-based multiclass segmentation for efficient object detection and feature extraction with a lightweight relation predictor. We evaluate SG-NCA on videos of cataract surgery and cholecystectomy, demonstrating performance comparable to established baselines while requiring 55x fewer parameters. We showcase deployment on fanless edge devices better suited for the OR and demonstrate downstream applications such as surgical video captioning, highlighting SG-NCA’s potential for affordable, privacy-preserving, and OR-ready intraoperative scene understanding.

[141] Remote-Sensing City Layout Extraction with MLLM cs.CVPDF

Zigan Zhou, Kai Li, Yupeng Deng

TL;DR: 本文提出了一种名为Code-as-City的新方法，利用多模态大语言模型从单张遥感图像中提取可执行的城市布局代码。该方法通过图像模型生成语义先验，再经MLLM多阶段推理恢复道路、土地覆盖和建筑，最终生成可渲染的3D城市布局和语义投影图。

Details

Motivation: 解决传统遥感系统输出（如检测框、语义掩码）无法构成保留对象身份、类型化关系、拓扑和再生规则的可执行布局的问题，旨在将视觉观察转化为可检查、可编辑的城市代码。

Result: 在CityLayout-100数据集的100个场景上评估，完整框架取得了41.1%的平均交并比和48.3%的全局交并比，为视觉到可编辑代码的转换提供了量化证据。

Insight: 创新点在于将城市布局提取建模为受约束的代码生成任务，结合MLLM的多阶段推理与确定性后处理，生成耦合的平面语义投影与3D可渲染布局，实现了可同步编辑的双视图输出。

Abstract: Remote-sensing systems usually describe urban content with detection boxes, semantic masks, or vector boundaries. Such outputs locate classes and support image-plane scoring, yet they do not by themselves constitute an executable layout that retains object identities, typed relations, topology, and regeneration rules. Code-as-City instead casts urban-layout extraction from a single top-down image as constrained code generation with a multimodal large language model (MLLM). An image model first produces an aligned five-class semantic layout prior. Three ordered MLLM passes use the image and this prior to recover roads, land-cover regions and relations, and buildings. Deterministic normalization converts the accumulated records into a city graph and a restricted layout program. Executing the program creates a renderable 3D city layout and an orthographic semantic projection over shared geometry. The projection admits pixel-level comparison with remote-sensing masks, while named objects, relations, and editing operations remain available for synchronized regeneration of both views. Evaluated on the 100 scenes of CityLayout-100, the complete framework obtains 41.1% mean intersection-over-union and 48.3% global intersection-over-union. This result provides quantitative evidence that visual observations can be translated into inspectable, editable city code with coupled planar and 3D outputs.

[142] RISE: Roadside Infrastructure Sequence Understanding across 3D Tracking and Structured Vision-Language Reasoning cs.CV | cs.AIPDF

Yanbo Jiang, Haotian Zheng, Jiahao Wang, Hanxiao Ren, Yitao Xu

TL;DR: RISE是一个用于路边基础设施序列理解的框架，涵盖度量3D跟踪和结构化视觉语言推理。在度量跟踪方面，该方法仅使用图像，结合SAM3视频身份识别和校准引导的掩码一致性进行多视角身份关联，无需LiDAR或任务特定的3D训练即可恢复持久的3D轨迹。在结构化视觉语言推理方面，通过一个人类审核的MLLM流程挖掘高价值片段，并使用受限的完整上下文Oracle构建基于边界框的预测性问答对，形成RISE-VQA数据集。

Details

Motivation: 解决路边多摄像头场景下，无需LiDAR或特定3D训练即可实现持久3D目标跟踪的挑战，并构建一个结构化的视觉语言推理基准来评估模型在空间定位、未来预测和交互推理等方面的能力。

Result: 在来自6个交叉路口的20段人工审核视频片段上，生成的3D轨迹在定义的多视角评估范围内达到了66.9的MOTA分数。构建的RISE-VQA数据集包含来自16个交叉路口、61个路边视角的557个片段中的33,910个问答对。实验表明领域适应和时间上下文带来一致收益，但也揭示了在空间定位、未来定位和交互推理方面存在持续挑战。

Insight: 创新点在于提出了一种仅依赖图像和校准信息、无需LiDAR或任务特定3D训练的跨视角3D跟踪方法，其校准条件化的几何设计使其可泛化到不同布局的交叉路口。同时，通过结构化流程构建了一个基于边界框的预测性视觉问答基准（RISE-Bench），采用交叉路口留出评估和确定性任务特定指标，为评估模型的复杂推理能力提供了新工具。

Abstract: We present RISE (Roadside Infrastructure Sequence Understanding and Evaluation), a framework spanning metric 3D tracking and structured vision-language reasoning in roadside sequences. For metric tracking, our image-only method combines SAM3 video identities with calibration-guided mask agreement for multi-view identity association, recovering persistent 3D tracks without LiDAR or task-specific 3D training. Its calibration-conditioned geometry allows the procedure to be instantiated at different calibrated multi-camera intersections without layout-specific retraining. On 20 human-reviewed clips from six intersections, the generated tracks achieve 66.9 MOTA within the defined multi-view evaluation scope. For structured vision-language reasoning, a human-reviewed MLLM pipeline mines high-value clips and uses a constrained full-context Oracle to construct bbox-grounded predictive QA without exposing future evidence to evaluated models. The resulting RISE-VQA dataset contains 33,910 QA pairs from 557 clips across 16 intersections and 61 roadside views. Its intersection-held-out RISE-Bench evaluates semantic choices, coordinates, future boxes, and interaction sets with deterministic task-specific metrics. Experiments show consistent benefits from domain adaptation and generally from temporal context, while revealing persistent challenges in spatial grounding, future localization, and interaction reasoning.

[143] MLLM-Guided Semantic Correction for Text-to-Video Generation cs.CV | cs.AIPDF

Junhao Chen, Zheqi Lv, Keting Yin, Shengyu Zhang, Zhou Zhao

TL;DR: 本文提出了一种无需训练、可解释的中间生成校正框架，用于解决文本到视频生成中的语义错误问题。该框架将多模态大语言模型（MLLM）的反馈直接集成到扩散采样循环中，通过语义评估信号在视频合成过程中进行轨迹校正，从而优化生成内容。

Details

Motivation: 当前基于扩散模型和Transformer的文本到视频生成模型常出现语义错误，如缺失对象、属性错误或动作不匹配，而现有方法在采样前优化或采样后细化，缺乏在生成过程中检测和校正语义偏差的探索。

Result: 在多个基准测试上的广泛实验验证了该方法的有效性，无需修改模型参数即可提升语义对齐、视觉保真度和时间一致性。

Insight: 创新点在于提出了一个训练免费的中间生成校正框架，包含语义评估监督器和语义修改助手两个关键模块，通过可控的潜在轨迹干预在推理过程中校正语义漂移，实现基于MLLM反馈的持续自反思优化。

Abstract: Recent advances in diffusion models and Transformer architectures have led to significant progress in text-to-video generation. However, these models often suffer from semantic errors such as missing objects, incorrect attributes, or mismatched actions. Although some semantic correction methods perform optimization before sampling or refinement after sampling, how to detect and correct semantic deviations during the video generation process remains underexplored. In this paper, we introduce a training-free, interpretable mid-generation correction framework that integrates multimodal large language model (MLLM) feedback directly into the diffusion sampling loop. Our framework achieves diffusion trajectory correction by injecting semantic evaluation signals during video synthesis, enabling the model to optimize the generated content through continuous self-reflection. We propose two key modules: a Semantic Assessment Supervisor that generates intermediate preview frames for semantic evaluations and deviation diagnostics, and a Semantic Modification Assistant that corrects semantic drift during inference via a controllable latent trajectory intervention. Our method improves semantic alignment, visual fidelity, and temporal consistency without modifying model parameters. We validate the effectiveness of our approach through extensive experiments across multiple benchmarks.

[144] FLEET: Token-Based Feature Extraction for Event Camera-based Reinforcement Learning cs.CVPDF

Tristan Gottwald, Maximilian Schier, Melanie Schaller, Bodo Rosenhahn

TL;DR: 本文提出FLEET（基于事件的令牌化特征学习），一种直接处理事件相机异步数据流的特征提取器。它利用随机傅里叶特征和交叉注意力将可变长度事件序列压缩为固定大小的潜在表示，从而在强化学习任务中实现端到端学习，无需辅助损失或轨迹数据预训练。

Details

Motivation: 现有方法未能充分利用事件相机异步、高频、稀疏的特性：基于CNN的方法将事件聚合为稀疏网格，耦合了计算成本与传感器分辨率并模糊了时间信息；而现有生成基线依赖轨迹数据进行预训练。

Result: 在一个新的高吞吐量基准测试上，FLEET超越了当前最优（SOTA）性能，并在观测频率变化方面表现出更强的鲁棒性。

Insight: 创新点在于通过令牌化机制直接处理原始事件流，利用随机傅里叶特征和交叉注意力实现计算成本与传感器分辨率的解耦，从而更有效地保留事件相机的时间稀疏特性，支持端到端强化学习。

Abstract: Event cameras generate asynchronous, high-frequency data streams offering spatially sparse information at lower latency than traditional cameras.In principle, these properties should be ideal for the design of control policies.However, reinforcement learning research in this field remains limited as existing approaches fail to fully exploit the sensor’s properties.CNN-based methods negate the sensors benefits by aggregating events into sparse grids. This couples compute cost to sensor resolution and blurs the temporal information. Meanwhile, existing generative baselines rely on the availability of trajectory data to pretrain the model. We propose FLEET (Feature Learning from Events via Efficient Tokenization), a feature extractor that processes event sequences directly. Leveraging random Fourier features and cross-attention, our architecture compresses variable streams into fixed-size latent representations. This decouples inference cost of the feature extractor’s backbone from the sensor’s resolution, enabling end-to-end learning without auxiliary losses. We validate FLEET on a new, high-throughput benchmark. The results demonstrate that our sequence-based approach surpasses SOTA performance and exhibits superior robustness to variations in observation frequencies.

[145] Automatic Cephalometric Landmark Localization on CBCT-Derived Digitally Reconstructed Radiographs for Skeletal Malocclusion Classification cs.CVPDF

Benjamin Hou, Konstantinia Almpani, Janice S. Lee, Zhiyong Lu

TL;DR: 本文提出了一个基于Vision Transformer的模型CephViT，用于自动定位二维侧位头影测量标志点，并评估了其在骨骼性错颌畸形分类中的应用。模型在公开数据集上训练和测试，并在私有CBCT扫描数据上通过生成数字重建放射影像（DRR）进行验证，结果表明使用DRR自动定位的标志点进行分类的性能与手动标注参考标志点相当。

Details

Motivation: 手动标注头影测量标志点对于颅面评估至关重要，但耗时耗力且难以扩展，因此需要自动化解决方案。

Result: 在公开侧位头影测量数据集上，CephViT的平均径向误差为1.28 +/- 1.42 mm，在3.0 mm阈值下的成功检测率为92.0%。在私有CBCT数据上，使用DRR自动定位标志点进行骨骼性错颌畸形分类的准确率为70.0%，与手动标注参考标志点的准确率68.3%相当。

Insight: 创新点在于将Vision Transformer应用于头影测量标志点定位任务，并展示了从3D CBCT生成DRR作为2D输入以实现自动化分析的可行性，为临床评估提供了高效工具。

Abstract: Manual cephalometric landmark annotation is important for craniofacial assessment but is labor-intensive and difficult to scale. We introduce CephViT, a Vision Transformer-based model for automated 2D lateral cephalometric landmark localization, and evaluate its use in downstream skeletal malocclusion classification. CephViT was trained and benchmarked on a public lateral cephalogram dataset, achieving a mean radial error of 1.28 +/- 1.42 mm and a successful detection rate of 92.0% at 3.0 mm. Because the private evaluation cohort consisted of 3D CBCT scans, lateral cephalogram-like digitally reconstructed radiographs (DRRs) were generated from each volume and used as 2D inputs to the landmark localization model. Landmark coordinates were normalized into a common coordinate frame, and skeletal malocclusion classification was performed using landmarks shared between the reference and DRR-based pipelines. Classification performance using DRR-localized landmarks was comparable to that obtained using manually annotated reference landmarks, with accuracies of 70.0% and 68.3%, respectively. These results support the feasibility of automated cephalometric analysis on CBCT-derived DRRs for skeletal malocclusion assessment.

[146] Supervising the Path to Fine Scales: GalerkinFlow for Scientific-Field and Image Super-Resolution cs.CV | cs.CE | cs.LGPDF

Zikang Zhan

TL;DR: 本文提出了GalerkinFlow，一种与方程无关的超分辨率框架，通过在重建路径的多个中间状态进行监督，而不仅仅是最终的高分辨率输出。该方法将每个粗-细图像对转化为沿整个重建路径的监督，利用粗锚点定义伪终点，并通过已知的时间相关权重将伪终点的重建损失与中间速度损失关联起来。此外，还监督用于一步推理的粗端点，并引入有限差分目标来约束局部空间变化。

Details

Motivation: 现有超分辨率模型通常仅通过监督最终高分辨率输出来从配对数据中学习，这导致对从下采样观测到精细目标之间的预测演化过程缺乏控制。

Result: 在Navier-Stokes和Darcy Flow基准测试中，GalerkinFlow在评估的与方程无关的基线方法中取得了最低的原始空间误差，同时在DIV2K数据集上保持竞争力。

Insight: 创新点在于将整个重建路径的中间状态纳入监督，通过伪终点损失与中间速度损失的关联实现更有效的学习，并结合了卷积特征与尺度条件化的Galerkin算子混合，无需控制方程或物理元数据。

Abstract: Most super-resolution models learn from paired data by supervising only the final high-resolution output. This provides little control over how the prediction should evolve between the downsampled observation and its fine target. We introduce GalerkinFlow, an equation-agnostic framework that turns each coarse–fine pair into supervision along an entire reconstruction path. At a random sample of intermediate states on the reconstruction path, the model predicts the coarse-to-fine residual velocity and uses coarse-anchor point to define a pseudo-endpoint. We show that the reconstruction loss of this pseudo-endpoint is exactly related to the intermediate velocity loss through a known time-dependent weight. Consequently, every intermediate state contributes supervision toward the same fine target, rather than serving only as an internal step toward an endpoint loss. Because intermediate states already reveal part of the missing fine-scale structure, we additionally supervise the coarse endpoint used during one-step inference. A finite-difference objective further constrains local spatial variation. GalerkinFlow combines convolutional features with scale-conditioned Galerkin operator mixing and requires no governing equation or physical metadata. It achieves the lowest raw-space errors among the evaluated equation-agnostic baselines on Navier–Stokes and Darcy Flow, while remaining competitive on DIV2K.

[147] SQuad: Sub-Quadratic Attention Distillation for Efficient Video Generation cs.CVPDF

Animesh Karnewar, Denis Korzhenkov, Amirhossein Habibian, Mohsen Ghafoorian

TL;DR: 本文提出SQuad（Sub-Quadratic Attention Distillation）框架，用于高效视频生成。该方法通过两阶段蒸馏（Flow-Matching SFT和DMD2）将预训练的全Softmax自注意力Video Diffusion Transformer（DiT）压缩到一个复杂度为O(n√n)的注意力模块中，在保持生成质量的同时大幅降低了计算开销和推理延迟。

Details

Motivation: 视频扩散变换器（Video DiTs）的自注意力操作计算成本随潜在令牌数量呈二次方增长，这严重限制了生成视频的分辨率和时长。现有的线性或低秩近似方法虽能提升效率，但往往牺牲了表达力，导致生成质量下降。

Result: 在Wan 2.2 5B文本到视频模型上，SQuad在VBench基准测试中与原始二次方复杂度的教师模型性能相当（83.20 vs 83.08），同时将每步每块的自注意力FLOPs降低约67倍，注意力延迟降低约11倍，端到端DiT延迟降低2倍，且仅需6次神经函数评估（NFEs）而非默认的100次即可生成视频。

Insight: 创新点在于提出了一种亚二次方复杂度（O(n√n)）的注意力蒸馏框架，通过两阶段蒸馏（监督微调与改进的分布匹配蒸馏）在效率与表达力之间取得平衡，实现了计算效率的显著提升而不损失生成质量，为大规模视频生成模型的部署提供了实用方案。

Abstract: Video Diffusion Transformers (DiTs) spend most of their compute inside the Self-Attention operation, whose cost grows quadratically, $\mathcal{O}(n^2)$, with the number of latent tokens $n$. For the task of video generation, the token count is large, so this term dominates runtime and memory, and thereby caps the resolution and duration we can generate. Linear $\mathcal{O}(n)$ and low-rank $\mathcal{O}(nk)$ surrogates of Self-Attention trade the full softmax $QK^T$ for cheaper kernels, but rarely recover the original’s expressivity, leaving a stubborn quality gap. Motivated by this, we propose SQuad, a Sub-Quadratic Attention Distillation framework that achieves a complexity of $\mathcal{O}(n\sqrt{n})$ in the resulting distilled Attention, naturally balancing the efficiency v/s expressivity trade-off. Instead of training our own Video DiT from scratch, which is prohibitively expensive, we fit a pretrained full softmax Self-Attention DiT into our proposed SQuad-Attention one by distilling the former in two stages: Flow-Matching Supervised Fine-Tuning (SFT), followed by improved Distribution Matching Distillation (DMD2) which additionally makes the sampling more efficient. On the Wan~2.2 5B text-to-video model, SQuAD matches the quadratic teacher on VBench ($83.20$ v/s $83.08$) while cutting the per-step per-block attention FLOPs by $\sim$$67\times$ and attention latency by $\sim$$11\times$, and end-to-end DiT latency by 2$\times$, all while also generating a video in only $6$ Neural Functional Evaluations (NFEs) instead of the default $100$.

[148] Ultra: Unsupervised Cross-Task Optimization for Reliable Restoration Segmentation Collaboration under Adverse Weather cs.CVPDF

Shiqin Wang, Zhiqian Li, Haoyuan Du, Junming Chen, Jiayuan Li

TL;DR: 本文提出了一种名为Ultra的无监督跨任务优化框架，用于在恶劣天气条件下实现可靠的图像恢复与语义分割协作学习。该框架通过将跨任务交互重新定义为不确定性下的方向选择和因果效应估计，解决了现有方法中因幻觉驱动的错误传播问题，从而在无目标域监督的情况下实现了更可靠的协作。

Details

Motivation: 现有无监督域适应恶劣天气语义分割方法隐含地假设图像恢复和语义分割任务能提供互利的指导，但在严重退化且无目标域监督的情况下，跨任务优化方向的有效性变得无法识别，导致幻觉驱动的错误传播。

Result: 在三个广泛使用的UDA-ASS基准测试上进行的广泛实验证明了该方法在分割性能上达到了最先进的水平。此外，该框架在无监督图像恢复方面也优于现有的UDA-ASS恢复方法，并能推广到无监督恢复与目标检测协作任务。

Insight: 创新点在于将跨任务交互重新定义为不确定性下的方向选择（通过提出的CTDN模块生成候选方向并进行协作选择）和因果效应估计（通过CMIL模块将基于相关性的传播转化为因果效应评估以抑制幻觉传播），从而实现了更可靠的跨任务协作学习框架。

Abstract: Unsupervised Domain Adaptation for Adverse Weather Semantic Segmentation (UDA-ASS) aims to transfer semantic knowledge from labeled normal-weather images to unlabeled adverse environments. Existing approaches implicitly assume that restoration and segmentation provide mutually beneficial guidance. However, under severe degradation and without target-domain supervision, the validity of cross-task optimization directions becomes fundamentally unidentifiable, leading to hallucination-driven error propagation. In this work, we propose a novel Unsupervised Restoration-Segmentation Collaborative Learning Framework (Ultra), which reframes cross-task interaction as direction selection under uncertainty and causal effect estimation, enabling reliable collaboration through candidate direction generation and intervention-based filtering. In detail, we propose CTDN and CMIL. The former exploits complementary visual structures and semantic information to generate candidate optimization directions and performs cooperative direction selection between restoration and segmentation. The latter reformulates cross-task information transfer from correlation-based propagation into causal effect assessment, suppressing hallucination propagation. Extensive experiments on three widely used UDA-ASS benchmarks demonstrate state-of-the-art segmentation performance. Beyond segmentation, our framework achieves better unsupervised restoration results than existing UDA-ASS restoration methods and generalizes to unsupervised restoration and object detection collaboration tasks. Code and models will be available at https://github.com/Wang-Shiqin/Ultra.

[149] Interactive Whole Slide Images for RL-based Tumour Segmentation cs.CVPDF

Mohamad Mohamad, Francesco Ponzio, Maxime Gassier, Nicolas Pote, Xavier Descombes

TL;DR: 该论文提出了一种基于强化学习（RL）的端到端框架，用于直接在整张全切片图像（WSI）上进行顺序肿瘤分割。该方法将WSI建模为一个分层多分辨率的环境，智能体通过移动、缩放和选择肿瘤等动作在其中导航，并利用近端策略优化（PPO）进行训练。在肺腺癌WSI上的实验表明，该方法能在几秒内完成整张切片的推理，达到与基于图像块（patch-based）方法相当的粗分割质量。

Details

Motivation: 全切片图像（WSI）分析面临计算挑战，因为其空间分辨率极高且肿瘤区域分布稀疏。现有方法通常将WSI视为预定义图像块的集合，本工作旨在探索一种更直接、更高效的交互式分割范式。

Result: 在肺腺癌WSI上的实验表明，该方法实现了直接在完整WSI上进行顺序肿瘤分割的可行性。其分割质量与在相似放大倍数下运行的基于图像块的方法相当，同时将每张切片的推理时间减少到几秒钟。

Insight: 核心创新点在于将WSI本身建模为一个分层的、可交互的强化学习环境，而不是静态的图像块集合。这为基于RL的计算病理学提供了一个新的方向，并探索了环境设计和动作空间粒度的影响。

Abstract: Whole-slide image (WSI) analysis remains computationally challenging due to the extremely large spatial resolution of slides and the sparse distribution of tumour regions. We propose an end-to-end reinforcement learning framework for sequential tumour segmentation directly on WSIs. Instead of treating the slide as a predefined collection of candidate patches, we formulate the WSI itself as a hierarchical multi-resolution environment through which an agent navigates using movement, zooming, and tumour selection actions. The agent jointly processes local observations and a global thumbnail representation within an actor-critic architecture trained using proximal policy optimization (PPO). Experiments on pulmonary adenocarcinoma WSIs demonstrate the feasibility of direct sequential tumour segmentation on full slides, achieving comparable coarse segmentation quality relative to patch-based approaches operating at similar magnification levels, while reducing inference time to a few seconds per slide. We further analyse the impact of environment design and action-space granularity. Our results suggest that modelling WSIs as interactive environments provides a promising direction for RL-based computational pathology

[150] HarmTrace: Anchor-Calibrated Decoupled Optimization for Fine-Grained Target Identification in Harmful Memes cs.CV | cs.AIPDF

Yujia Li, Yiqun Zhang, Zihan Cheng, Yijie Huang, Tenglong Ye

TL;DR: 本文提出HarmTrace框架，用于有害梗图的细粒度目标识别，不仅检测有害性，还识别攻击的目标类型、具体实体、文本提及和视觉区域。作者构建了Meme3W数据集并设计了联合记录准确率（JRA）指标，通过实体感知微调和条件目标识别策略优化（CTPO）提升模型性能。

Details

Motivation: 现有有害梗图检测模型仅关注有害性分类，可能误判攻击目标或证据，因此需要扩展为细粒度目标识别任务，以准确识别被攻击目标的类型、身份和位置。

Result: 在Qwen3-VL-8B骨干网络上，HarmTrace将JRA从17.58%提升至52.51%，并在多个评估骨干上同时提高了JRA和有害性准确率。

Insight: 创新点包括引入细粒度目标识别任务、构建统一标注数据集Meme3W、设计严格记录级指标JRA，以及提出锚点校准解耦优化框架HarmTrace，其中CTPO利用虚拟正锚点（VPA）进行目标识别优势归一化，有效解耦有害性和目标识别优化。

Abstract: Multimodal harmful meme detection is typically formulated as image–text harmfulness classification. A model may correctly predict harmfulness while misidentifying the attacked target or its supporting evidence. We therefore extend harmful meme detection with fine-grained target identification, asking what type of target is attacked, who is targeted, and where the target appears in the meme. The model predicts harmfulness for every meme and, for harmful memes, outputs the target category, target entity, textual mention, and visual region. To support this task, we introduce Meme3W, which unifies multiple public harmful meme datasets and provides human-verified annotations for harmful instances. We further introduce Joint Record Accuracy (JRA), a strict record-level metric requiring the harmfulness label and all target-identification fields to be jointly correct. Experiments with representative multimodal large language models reveal a substantial gap between harmfulness accuracy and JRA. To narrow this gap, we propose HarmTrace, an anchor-calibrated decoupled optimization framework. HarmTrace strengthens target-entity supervision through entity-aware supervised fine-tuning. It then applies Conditional Target-identification Policy Optimization (CTPO) to decouple harmfulness and target-identification advantages, restricting target-identification optimization to label-correct responses for harmful examples. CTPO uses a Virtual Positive Anchor (VPA) as a fully correct reference for target-identification advantage normalization. HarmTrace improves both JRA and harmfulness accuracy across the evaluated backbones, with JRA on the Qwen3-VL-8B backbone increasing from 17.58% to 52.51%. Our code is publicly available at https://github.com/llly1234/HarmTrace-for-Harmful-Memes.

[151] X$^2$Localizer: Cross-grained Alignment for Progressive Cross-view Video Geo-localization cs.CV | cs.AI | cs.ROPDF

Zichao Zeng, Weijia Fan, Yufan Chen, June Moh Goo, Junwei Zheng

TL;DR: 本文提出了渐进式跨视角视频地理定位（PCVG）任务及其评估协议，旨在解决传统方法依赖固定长度输入和事后精炼的问题，以支持在线、动态观测下的定位。为此，作者提出了X²Localizer框架，通过跨粒度对齐联合监督全局前缀-航拍图像检索和令牌聚合的帧-航拍瓦片匹配，并引入滑动窗口重定位策略以实现失败恢复和长距离部署。实验表明，该方法在保持全视频性能的同时，显著提升了早期定位能力，并在单帧设置和长距离场景下表现出色。

Details

Motivation: 传统跨视角视频地理定位方法依赖固定长度输入和事后精炼，难以适应在线部署中部分或动态观测的需求。因此，本文旨在提出一个面向部署的渐进式定位框架，支持在变化的时间预算下进行定位，包括基于前缀的推理、随机起始评估和带中断的长距离定位。

Result: 在传统全视频设置下，X²Localizer保持了性能，Recall@1和Recall@10分别有+0.1和+0.3的边际提升。在更具挑战性的单帧设置中，该方法相比之前的最先进方法，在粗检索上Recall@1提升了+4.7，Recall@10提升了+11.5。结合滑动窗口重定位策略后，该方法在随机起始和长距离场景下实现了鲁棒的渐进式定位。

Insight: 创新点包括：1) 提出PCVG任务和评估协议，将地理定位扩展到渐进式、在线部署场景；2) 设计跨粒度对齐框架，通过预算依赖的非对称目标联合优化全局检索和局部匹配；3) 引入滑动窗口重定位策略，动态刷新候选区域以实现失败恢复和长序列处理，无需全序列重处理，缩小了基准评估与实际部署间的差距。

Abstract: Cross-view Video Geo-localization (CVG) aims to localize ground-view videos by retrieving their corresponding geo-tagged aerial images. However, CVG approaches rely on fixed-length inputs and post-hoc refinement, hindering online-oriented localization under partial or dynamic observations. In this work, we formulate Progressive Cross-view Video Geo-localization (PCVG) as a deployment-oriented extension and evaluation protocol of CVG, enabling localization under varying temporal budgets, prefix-based inference, random-start evaluation, and long-range localization with interruptions. To explore PCVG, we introduce X$^2$Localizer, a cross-grained alignment framework that jointly supervises global prefix-to-aerial retrieval and token-aggregated frame–aerial-tile matching with a budget-dependent asymmetric objective. Furthermore, we introduce a Sliding-Window Re-Localization (SWRL) strategy that dynamically refreshes candidate regions for failure recovery and long-range deployment without full-sequence reprocessing. Extensive experiments show that X$^2$Localizer preserves conventional full-video performance, with marginal gains of +0.1 Recall@1 and +0.3 Recall@10, while substantially improving early localization. In the challenging single-frame setting, X$^2$Localizer improves coarse retrieval by +4.7 Recall@1 and +11.5 Recall@10 over the previous state-of-the-art method. With SWRL, our approach further enables robust progressive localization under random-start and long-distance scenarios, narrowing the gap between benchmark evaluation and real-world deployment.

[152] Concept-based explanation of gene expression prediction from H&E images cs.CVPDF

Amos Muench, Jonathan Thielmann, Reduan Achtibat, Maximilian Dreyer, Philip Bischoff

TL;DR: 本文提出了一种结合相关性传播和概念发现的可解释性框架，用于从H&E病理图像中预测空间转录组学（ST）数据，并揭示形态学概念如何影响ST预测。该框架在结直肠癌ST数据上进行了验证，能够准确预测临床相关的ST特征和分子表型，并建立了基于相关性的概念图谱，将分子表型与组织病理学表征联系起来。

Details

Motivation: 现有的基于视觉Transformer（ViT）的病理基础模型可解释性方法主要局限于局部热力图，无法揭示形态学概念如何贡献于空间转录组学预测。

Result: 在HEST-1k队列的结直肠癌ST数据上，该框架对iCMS亚型的空间分辨和聚合分类分别达到了0.872和0.819的加权F1分数（在TCGA COAD上为0.770），均能有效分层患者预后。

Insight: 创新点在于将ViT感知的层相关性传播与基于松弛原型TopK稀疏自编码器的概念发现相结合，提供了局部解释和全局洞察；研究表明，基于相关性的概念比基于激活的概念能更直接地连接组织形态与下游预测，为基于概念的空间预测解释提供了通用策略。

Abstract: Recent advances in pathology foundation models have enabled accurate prediction of spatial transcriptomics (ST) from routine H&E images. However, existing explainability methods for vision transformer (ViT)-based models are largely limited to local heatmaps and do not reveal how morphological concepts contribute to ST predictions. Here, we introduce an explainable framework that combines relevance propagation and concept discovery to link transcriptional programs to tissue morphology. We developed a ViT-based framework for virtual ST from H&E images that combines ViT-aware layer-wise relevance propagation with relaxed archetypal TopK sparse autoencoder-based concept discovery. This approach provides both local explanations and global insights into the morphological patterns associated with transcriptional programs. We applied the framework to colorectal cancer ST data from the HEST-1k cohort and evaluated its generalizability in TCGA COAD. Our architecture accurately predicts clinically relevant ST signatures and accompanying molecular phenotypes. Measured and predicted gene expression profiles reveal substantial spatial heterogeneity of the colorectal cancer subtypes iCMS2 and iCMS3 across a large number of samples. Spatially resolved and aggregated iCMS classification achieve weighted F1 scores of 0.872 and 0.819 (0.770 in TCGA COAD), respectively, and both stratify patient outcome. Beyond prediction, our framework establishes a relevance-based concept atlas linking molecular phenotypes to histopathological representations. Comparison of activation- with relevance-derived concepts demonstrates that relevances provide a more direct link between tissue morphology and downstream predictions. We establish a general strategy for concept-based explanation of spatial prediction, and our framework is readily applicable to a broad range of ViT-based pathology models.

[153] AnchorScore: A CLIP-Based Diagnostic of MLLM Annotation Difficulty cs.CVPDF

Yan Ma, Lizhuo Zhang

TL;DR: 本文提出了一种名为AnchorScore的低成本诊断方法，用于预测多模态大语言模型（MLLM）在自动标注任务中不同类别的标注难度。该方法基于CLIP的零样本准确率构建，能够高效地识别MLLM可能标注不可靠的类别，从而指导资源分配。

Details

Motivation: MLLM用于自动标注时，其在不同类别上的准确率差异巨大且评估成本高昂，目前缺乏一种低成本的事先方法来预测和排序各类别的标注难度。

Result: 在课堂行为数据集（SCB5）上，AnchorScore与MLLM的每类准确率显著相关（Spearman rho = 0.769），优于其他预测指标；在Stanford40 Actions数据集上的独立复现也得到类似强相关结果（rho = 0.817）。

Insight: 创新点在于利用冻结CLIP模型的零样本准确率（AnchorScore）作为低成本、高效的事先诊断信号，以识别MLLM标注的困难类别，并应用于混合路由、提示消歧和人工审核优先级预测等实际场景。

Abstract: Multimodal large language models (MLLMs) are widely used for automated annotation, yet their per-class accuracy varies widely (e.g., 12%-98% across the 13 classes of three classroom sub-datasets) and is expensive to measure: evaluating one 27B MLLM on 5,416 validation images takes roughly 14 hours, whereas a frozen-CLIP pass over the same images completes in about 3 minutes. A low-cost signal for ranking classes by expected MLLM annotation difficulty a priori remains underexplored. Building on the AnchorProxy construct (per-class zero-shot CLIP accuracy) introduced in the companion study, this paper systematically evaluates its full-frame formulation, termed AnchorScore here, as an a priori diagnostic that flags the classes MLLMs are least likely to annotate reliably. On classroom behavior data (SCB5, 13 classes, 6 MLLMs), AnchorScore correlates with per-class MLLM accuracy (Spearman rho = 0.769, p = 0.002, n = 13). None of the alternative difficulty predictors (DINOv2, ResNet-50, SigLIP, or MLLM self-verbalized uncertainty) showed a significant class-level correlation at n = 13. A cross-model consensus control suggests AnchorScore primarily captures a shared class-difficulty factor rather than a CLIP-specific signal. An independent replication on Stanford40 Actions yields a nearly identical effect (rho = 0.817, p < 0.001); the association is strongest on activity-recognition data and attenuates on medical and satellite imagery. Three practical applications follow: a deployable hybrid CLIP/MLLM routing strategy (predicted-class routing: up to +23 pp over CLIP-only at roughly 44% MLLM cost savings), prompt disambiguation on hard classes (exploratory), and review-priority prediction for human verification. AnchorScore does not estimate exact MLLM accuracy; it provides a low-cost ranking signal that directs expensive MLLM evaluation to the classes where it is most informative.

[154] How Sampling Strategy Affects Imbalance Mitigation in LiDAR Segmentation: A Study of Structured vs. Random Point-Based Architectures cs.CVPDF

Antonis Savva, Christos Kyrkou, Theocharis Theocharides

TL;DR: 本文研究了LiDAR点云语义分割中的类别不平衡问题，重点分析了采样策略（结构化采样与随机采样）如何影响不同不平衡缓解技术的效果。通过在三个数据集（DALES、S3DIS、STPLS3D）上对两种架构（KPConv、RandLA-Net）进行六种重加权方案和五种不平衡感知损失的基准测试，发现逆频率加权会显著降低性能，而均匀加权在结构化采样中表现接近复杂损失，但在随机采样中效果较差。损失景观分析揭示了采样策略、不平衡程度与数据采集特性之间的复杂相互作用决定了缓解方法的有效性。

Details

Motivation: LiDAR点云中的类别不平衡给自动驾驶导航和城市地图绘制的语义分割带来了挑战。尽管2D视觉领域已有多种缓解技术，但它们在3D点云中的有效性尚不明确，因此需要系统研究不同不平衡缓解方法在3D点云分割中的表现及其与采样策略的关联。

Result: 在三个数据集（DALES、S3DIS、STPLS3D）上使用两种架构（KPConv、RandLA-Net）进行实验。逆频率加权相比均匀加权性能下降高达12%，且在少数类上出现灾难性失败。对于结构化采样（KPConv），均匀加权与复杂损失的性能差距在2%以内；而对于随机采样（RandLA-Net），差距可达4.6%。损失景观分析表明，结构化采样的损失几何受真实LiDAR数据的不平衡比影响，但在合成数据上解耦；随机采样的损失景观则对数据集几何高度敏感，与不平衡比无关。

Insight: 论文的创新点在于首次系统揭示了采样策略（结构化vs.随机）是决定3D点云不平衡缓解方法有效性的关键因素，而非仅仅是不平衡程度本身。从客观角度看，该研究提供了重要的实践指导：对于结构化采样架构，简单的均匀加权可能已足够；而对于随机采样架构，则需要更精细的不平衡感知损失设计。损失景观分析为理解不同数据（真实vs.合成）和架构的交互机制提供了新视角。

Abstract: Class imbalance in LiDAR point clouds poses challenges for semantic segmentation in autonomous navigation and urban mapping. While 2D vision has numerous mitigation techniques, their effectiveness in 3D remains unclear. We benchmark six reweighting schemes and five imbalance-aware losses across three datasets (DALES, S3DIS, STPLS3D) using two architectures (KPConv, RandLA-Net). Inverse-frequency weighting degrades performance by up to 12% compared to uniform weighting, with catastrophic failures in minority classes. Uniform weighting performs within 2% of complex losses for structured sampling (KPConv) but benefits less for random sampling (RandLA-Net, up to 4.6% gap). Loss landscape analysis reveals a complex interplay: for structured sampling, imbalance ratio determines landscape geometry on real LiDAR data but decouples from it on synthetic data; for random sampling, landscapes show high sensitivity to dataset geometry regardless of imbalance ratio. For the two evaluated point-based architectures, these results suggest that the interaction between sampling strategy (structured vs. random), imbalance severity, and data acquisition characteristics shapes which mitigation approaches are effective.

[155] MIRROR: Multimodal Intelligent Radiology Reasoning and Observation Reporter cs.CV | cs.AI | cs.LGPDF

Vignesh Nagarajan, Sriram Venkatapathy

TL;DR: MIRROR是一个用于放射学报告生成的多模态智能系统原型，旨在解决模型输出缺乏可解释性以及后续文本生成可能添加虚假主张的问题。它通过串联一个多标签分类器、一个基于Grad-CAM的定位器和一个不接触图像的文本生成器，确保报告中的发现与分类器的概率输出严格对应，从而实现了可审计性。

Details

Motivation: 解决放射科医生在解读模型输出时面临的两个核心问题：模型仅返回数值而缺乏解释理由，以及将数值转换为可读文本的系统可能悄悄添加模型从未做出的诊断主张。

Result: 在ChestMNIST数据集上，分类器的宏观AUROC达到0.729，在所有14个标签上的排序均优于随机，精度是随机排序器的1.6至6.8倍。然而，在默认0.5阈值下，对其中11个标签未产生任何阳性预测，其出色的Brier分数（0.045）与忽略图像的预测器所得分数（0.047）相近，表明在放射学常见的类别不平衡下，聚合指标容易夸大无所作为的模型性能。

Insight: 系统架构的创新在于通过强制分离视觉分析与文本生成（语言层无法接触像素），确保了报告发现的严格可审计性，防止了生成文本的幻觉。同时，论文揭示了在类别不平衡的放射学任务中，仅依赖宏观聚合评估指标（如AUROC、Brier分数）具有误导性，必须报告相对于“不作为基线”的性能，这是一个重要的方法论洞见。

Abstract: A radiologist reading a model’s output faces two problems. The model returns a number and no reason, and any system that turns that number into readable prose can quietly add claims the model never made. MIRROR is a research prototype built to separate those failures. It chains a multi-label classifier, a Grad-CAM localizer that turns each positive finding into a named anatomical region, and a report writer that receives the labels, probabilities, and regions but never the image. Because the language layer cannot see pixels, it cannot assert a finding the classifier did not make. We are precise about what that buys: a MIRROR report’s findings are auditable against the probability vector, while the sentences framing them are ordinary generated text, and we show one stating a cardiothoracic ratio the system never measured. One registry holds the taxonomy, anatomy, and phrasing for chest X-ray, brain MRI, and head CT, so adding a modality is a data change; all three are routed and tested, one is trained. On ChestMNIST that classifier reaches macro AUROC 0.729 and ranks better than chance on all 14 labels, at 1.6 to 6.8 times the precision a random ranker would get. Yet at the default 0.5 threshold it emits no positive prediction at all for 11 of them, and its excellent-looking Brier score of 0.045 sits beside the 0.047 earned by a predictor that ignores the image. The discrimination is real; the decisions are not. Under the class imbalance normal in radiology, aggregate metrics flatter models that do nothing, and should be reported against that floor.

[156] GenRouter: Unified Workflow Routing for Agentic Image Generation cs.CVPDF

Harold Haodong Chen, Zhiyu Hou, Wen-Jie Shu, Weilin Ruan, Yingjie Xu

TL;DR: 本文提出了GenRouter，这是首个用于智能体图像生成的统一工作流路由框架。该框架通过标准化不同的智能体流水线为基本原语和可执行模板（GenCanvas），然后根据需求分析、经验匹配和帕累托过滤，将异构提示自适应地路由到最优工作流。实验表明，与静态重型流水线相比，GenRouter在保持卓越视觉对齐的同时，显著降低了计算成本和延迟。

Details

Motivation: 当前基于智能体的图像生成工作流虽然增强了静态推理能力（如外部知识检索和迭代推理），但它们大多是具有固定拓扑结构的孤立系统，导致计算不匹配问题，即简单查询被迫经过计算量大的流水线。本文旨在解决这种效率低下的问题。

Result: 在多个基准测试上的广泛实验表明，与静态重型流水线相比，GenRouter实现了卓越的视觉对齐，同时将执行成本降低了95%以上，延迟降低了65%。系统还能通过积累的经验持续自我进化，实现鲁棒的零样本泛化，从而提升性能并减半计算开销。

Insight: 主要创新点在于：1）提出了GenCanvas，将多样化的智能体流水线标准化为统一的基础原语和模板，为路由提供了统一的操作空间；2）设计了包含需求分析、经验匹配和帕累托过滤的三阶段自适应路由机制；3）系统具备通过经验积累实现持续自我进化和零样本泛化的能力，这是一个重要的系统级设计思路。

Abstract: The rapid evolution of text-to-image (T2I) generation models has effectively solved the foundational challenge of raw pixel synthesis, shifting the community’s focus toward fulfilling increasingly intricate user requests. While recent agentic image generation workflows enhance static inference with advanced capabilities like external knowledge retrieval and iterative reasoning, they mostly operate in isolated silos with fixed ``one-size-fits-all” topologies. This inevitably leads to severe compute-mismatch, where simple queries are forced through computationally heavy pipelines. To bridge this gap, we present GenRouter, the first unified workflow routing framework for agentic image generation. We first formulate GenCanvas, standardizing diverse agentic pipelines into a universal set of foundational primitives and executable templates. Operating over this unified space, GenRouter adaptively routes heterogeneous prompts to their optimal workflows via (i) demand profiling, (ii) experience matching, and (iii) Pareto filtering. Extensive experiments across diverse benchmarks demonstrate that GenRouter achieves superior visual alignment while reducing execution costs by over 95% and latency by 65% compared to heavyweight static pipelines. Furthermore, the system continuously self-evolves via accumulated experience, enabling robust zero-shot generalization that boosts performance and halves computational overhead.

[157] PersonaShot: Benchmarking Person-Centric Narrative Continuity in Multi-Shot Video Generation cs.CVPDF

Yuji Wang, Yuheng Chen, Teng Hu, Ran Yi, Yijia Hong

TL;DR: 本文提出了PersonaShot，首个以人物为中心的多镜头视频生成叙事连贯性基准。该基准包含约1000个多镜头片段和16个评估指标，涵盖物理连贯性、情感动态和电影语法三个维度。论文还开发了与人类专家判断对齐的轻量级专用评估器，用于系统性地评估现有先进模型，揭示了它们在跨镜头叙事连贯性上的不足。

Details

Motivation: 现有视频生成基准主要评估角色外观或单镜头质量，缺乏对跨镜头物理与情感状态连贯性的衡量，且缺少针对特定评估准则（如物理连贯性、面部动态、电影关系）的专用评估方法。

Result: 在PersonaShot基准上的系统评估揭示了最先进模型在跨镜头叙事连贯性方面存在明显能力差异和不足，即使视觉质量高的视频也常出现物理状态重置、情感突变和电影关系断裂。人类研究证实了所提评估器与专家判断具有高度一致性。

Insight: 创新点在于提出了首个专注于多镜头视频中人物叙事连贯性的综合基准，并设计了基于多模态大模型知识蒸馏的、针对不同评估准则的轻量级专用评估器，实现了与人类判断对齐的自动化评估。

Abstract: Video generation is rapidly evolving from single-shot clips to multi-shot narratives, where the human character serves as the core narrative anchor. However, existing benchmarks mainly assess character appearance or individual-shot quality, without measuring whether physical and emotional states remain coherent across cuts. They also rarely provide criterion-specific evaluation methods, although physical continuity, facial dynamics, and cinematic relations require different visual, temporal, and relational evidence. To address these limitations, we introduce PersonaShot, the first person-centric benchmark for narrative continuity in multi-shot video generation. PersonaShot contains approximately 1,000 multi-shot segments and 16 metrics spanning physical continuity, affective dynamics, and cinematic grammar. \textbf{\textit{1)} Narrative Continuity Benchmark:} We evaluate character coherence across three temporal levels: within-shot states, cross-shot transitions, and sequence-level trajectories. \textbf{\textit{2)} Human-Aligned Specialist Evaluators:} We distill reasoning from a large multimodal teacher into lightweight criterion-specific evaluators, each grounded in the visual, temporal, or relational evidence required by its metric, and align them with expert human judgments. \textbf{\textit{3)} Systematic Evaluation and Insights:} Our evaluation reveals distinct capability profiles across state-of-the-art models and a clear gap between perceptual quality and cross-shot narrative continuity. Even visually compelling videos frequently exhibit physical-state resets, abrupt affective shifts, and broken cinematic relations across shots. Human studies further demonstrate strong agreement between our evaluators and expert judgments.

[158] VicEdit: Learning to Edit Videos from Visual In-Context Examples cs.CVPDF

Yuji Wang, Teng Hu, Yuheng Chen, Ran Yi, Han Feng

TL;DR: 本文提出了一种名为VicEdit的视频编辑新范式，旨在通过多模态视觉上下文（如图像、图像对、视频对）来弥补纯文本指令在传达细粒度纹理和复杂动态方面的不足。为此，作者构建了首个大规模视觉上下文视频编辑数据集VicEdit-400K，并开发了一个统一的框架，通过模态自适应语义蒸馏和双上下文注入技术，将视觉和文本信号协同整合到生成过程中。

Details

Motivation: 基于文本指令的视频编辑方法存在固有局限，难以精确传达细粒度的视觉纹理和复杂的动态信息，这导致了感知上的差距。本文旨在通过引入多模态视觉指导来弥合这一差距，提升视频编辑的精细度和可控性。

Result: 在VicEditBench基准测试上的广泛评估表明，VicEdit在基本指令编辑和视觉上下文编辑任务上都达到了最先进的性能水平。

Insight: 论文的核心创新在于提出了视觉上下文编辑的新范式，并构建了首个大规模数据集VicEdit-400K来支持该范式。技术上，其设计的模态自适应语义蒸馏能够从异构视觉参考中自适应地提取编辑语义，而双上下文注入机制则实现了视觉与文本信号的协同融合，这为可控的多模态视频生成提供了可借鉴的思路。

Abstract: Despite progress in instruction-based video editing, unimodal textual instructions inherently struggle to convey fine-grained textures and complex dynamics. To bridge this perceptual gap, we propose Visual In-context Editing, a new paradigm elevating video editing from textual instructions to multi-modal visual guidance encompassing single image, image pair, and video pair. To facilitate this paradigm, we curate VicEdit-400K, the first large-scale dataset for visual in-context video editing. We develop an automated pipeline to generate 400K high-quality samples across ten task types, ensuring superior visual fidelity and semantic consistency through multi-dimensional filtering. Leveraging this foundation, we introduce VicEdit, a unified framework to bridge visual and textual contexts. To adaptively extract editing semantics from heterogeneous references, we design Modality-Adaptive Semantic Distillation, which produces modality-specific semantic tokens from visual references. These tokens are then synergistically integrated with textual instructions through Dual-Context Injection, enabling the generation process to benefit from both visual and textual signals. Extensive evaluations on VicEditBench demonstrate that VicEdit achieves state-of-the-art performance across both basic instruction editing and visual in-context editing tasks, establishing visual in-context learning as a powerful and controllable paradigm for video editing.

[159] Binarized High-Efficiency RAW Video Restoration and Beyond cs.CVPDF

Tianyu Zhu, Ying Fu, Hesong Li, Gengchen Zhang, Xin Yuan

TL;DR: 本文提出了BinRVR，一个用于RAW视频恢复的二值化神经网络框架，旨在实现高效轻量部署。该框架通过提出的二值化信息交互模块（BIIM）统一高效地建模时空信息，并利用分布感知二值化卷积（DAB-Conv）缓解量化误差，在计算量和参数量减少约96%的情况下，性能仅下降约4%。

Details

Motivation: 解决现有二值化神经网络（BNNs）在应用于视频恢复任务时，因难以建模时间连贯性和激活值分布而导致的性能不足问题，以实现高效的RAW视频恢复。

Result: 在RAW视频恢复任务（包括低光增强、去噪、去模糊和超分辨率）上，与最先进的二值化方法相比取得了有竞争力的性能。框架还支持多比特量化，能在不同硬件约束下实现灵活的精度-效率权衡。

Insight: 创新点在于提出了BIIM模块来联合建模时空信息，以及DAB-Conv利用全精度激活的统计信息来减轻量化误差。从客观角度看，其将高效二值化设计系统性地扩展到视频领域，并支持灵活的量化位宽，是具有实用价值的架构创新。

Abstract: RAW video restoration is fundamental to high-quality low-level perception and serves as the basis for a wide range of downstream vision applications. While binary neural networks (BNNs) enable efficient lightweight deployment for image enhancement, their deficiencies in modeling temporal coherence and activation value distributions hinder their effectiveness when applied to video scenarios. In this paper, we propose BinRVR, a binarized RAW video restoration framework that reduces computation and parameters by approximately 96% while incurring only about 4% performance degradation. Specifically, we present a Binarized Information Interaction Module (BIIM) to jointly model spatial and temporal information in an efficient and unified manner. Moreover, we develop a Distribution-Aware Binarized Convolution (DAB-Conv) that leverages the statistics of full-precision activations to mitigate quantization errors. The proposed framework further supports multi-bit quantization, enabling flexible accuracy-efficiency trade-offs across different hardware constraints. Extensive experiments demonstrate that our BinRVR achieves competitive performance compared with state-of-the-art binarized methods on RAW video restoration tasks, including low-light enhancement, denoising, deblurring, and super-resolution. We further explore the potential of our method on downstream video applications, including object detection and monocular depth estimation.

[160] Diagnosing Dense Same-Class Attribute Misbinding in Large Vision-Language Models cs.CV | cs.AIPDF

Yuanzhi Xu, Qian Gao, Jun Fan, Guohui Ding, Zhenyu Yang

TL;DR: 该论文提出了密集同类别属性误绑定（DSCAM）问题，即大型视觉语言模型在拥挤场景中会将属性错误地分配给同类的其他实例。研究创建了InstaBind-Lite基准测试来量化这一问题，包含524张图像和9580个可确定性评估的问题，并评估了多个开源和商业模型，揭示了传统准确率指标无法发现的系统性错误。

Details

Motivation: 现有视觉问答准确率和物体幻觉指标无法揭示模型将属性错误绑定到同类其他实例的问题，这构成了评估盲点。

Result: 在InstaBind-Lite基准上，开源模型平均误绑定率为19.84%，API系统为7.55%；其中约80%的错误转移发生在相邻实例之间。

Insight: 创新点在于形式化了DSCAM问题并构建了专门的评测基准，通过源实例标注和绑定特异性指标（如转移频率、邻接性）实现了对模型属性绑定可靠性的细粒度诊断，超越了传统聚合指标。

Abstract: Large vision-language models can recognize the objects and attributes in a crowded scene yet assign an attribute to the wrong same-class instance. Generic visual-question-answering accuracy marks the response as wrong, while object-hallucination metrics may regard both the object and attribute as image-supported; neither reveals the transfer. This study formalizes this blind spot as Dense Same-Class Attribute Misbinding (DSCAM) and presents InstaBind-Lite, a controlled benchmark that makes it directly measurable. Its 524 images contain 529 curated groups of 3-6 same-class entities, 1773 boxed instances, ordered neighbors, distinguishable color-like attributes, and four complementary question levels, yielding 9580 deterministically evaluated questions. Unlike existing protocols, source-instance annotations separate unsupported generation and recognition failure from an attribute copied from another visible entity. Binding-specific metrics further quantify transfer frequency, adjacency, ordinal distance, and intervention effects. Across five open-source and two commercial/API models, the open-source systems average 19.84% Misbinding Rate and the API systems 7.55%; these errors are hidden by aggregate accuracy. Among identifiable transfers, 80.70% and 81.51%, respectively, originate from adjacent instances. Localization and instance-first interventions help selected models but are not universal remedies. InstaBind-Lite therefore turns previously undifferentiated wrong answers into source-identifiable failure categories and tests a reliability dimension that conventional benchmarks cannot determine: whether a model knows not only what is visible, but which instance owns each attribute.

[161] TRACE-Bench: Decomposing and Diagnosing Multi-Reference Image Generation cs.CV | cs.AIPDF

Haoran Wang, Chaofan Ma, Ran Yi, Lizhuang Ma

TL;DR: 本文提出了TRACE-Bench，一个用于多参考图像生成任务的新型诊断性基准测试。该基准通过将复杂的多参考提示分解为四个原子操作（锚定、解缠、应用和组合），并基于操作符槽位数量构建了约1600个评估案例，从而能够对模型能力进行细粒度评估和故障定位。

Details

Motivation: 现有针对多参考图像生成的基准测试通常围绕预定义的任务类型（如“主体组合”）组织，这导致评估覆盖碎片化、复杂度不可控且缺乏诊断价值。本文旨在解决这一问题，通过能力导向的视角，为统一的多模态模型提供一个系统性的评估框架。

Result: 在TRACE-Bench上评估了9个领先模型，揭示了整体评分无法发现的洞见：主要瓶颈在于解缠（g）和属性绑定（⊕）操作，而非场景级组合（C）。即使在最佳模型上，属性保真度得分也仅为0.74。

Insight: 创新点在于采用能力分解的视角，将多参考任务形式化为四个原子操作的组合公式，并据此构建了一个结构化的、可诊断的基准。这为模型能力的细粒度评估和故障根源的递归定位提供了系统方法，超越了传统基于任务类型的评估范式。

Abstract: Despite recent advances in unified multimodal models for multi-reference image generation, existing benchmarks remain organized around predefined task types (e.g., “subject composition”), which are ill-suited to this combinatorial setting and lead to fragmented coverage, uncontrolled complexity, and little diagnostic value. Recognizing that diverse multi-reference tasks share a common set of atomic operations, we adopt a capability-oriented perspective and formalize four operators: Anchor ($f$), Disentangle ($g$), Apply ($\oplus$), and Compose ($C$). Any multi-reference prompt can then be represented as a compositional formula over these operators, whose structural complexity is quantified by the number of operator slots. Building on this formulation, we construct TRACE-Bench, comprising approximately 1,600 evaluation cases across slot counts 1–8, built from 631 formula templates and around 4,000 reference images spanning diverse artistic styles and real-world subjects. The formula structure directly drives an operator-aligned evaluation protocol for per-capability scoring and a diagnostic tree analysis for recursive failure localization. Evaluating 9 leading models reveals insights invisible to holistic scoring: the primary bottleneck lies in disentanglement ($g$) and attribute binding ($\oplus$) rather than scene-level composition ($C$), with even the best model scoring only 0.74 on attribute fidelity. Project page: https://amuseum-whr.github.io/TraceBench

[162] PixRestore: Unified Image Restoration via Pixel Diffusion Transformer cs.CVPDF

Lingchen Sun, Rongyuan Wu, Xiangtao Kong, Jixin Zhao, Qiaosi Yi

TL;DR: PixRestore 是一种用于统一图像恢复（UIR）的、无需变分自编码器（VAE）的像素空间扩散Transformer模型。它通过从零开始训练扩散主干，直接在像素块上进行流匹配来保留细节，并利用特征可靠性预测来适应不同退化。该模型经过大规模数据训练，并可微调为单步生成器，在保持高效推理的同时，在多个基准测试中实现了最佳的保真度、感知质量和鲁棒性。

Details

Motivation: 现有UIR方法通常依赖预训练的文本到图像（T2I）潜在扩散模型，但其VAE可能丢失对恢复敏感的细节，且其开放式合成先验可能引入内容不一致的伪影。PixRestore旨在解决这些问题，通过设计一个不依赖T2I预训练、直接在像素空间操作的模型来更好地保留细节并避免伪影。

Result: 在公共基准测试和真实世界测试集上的实验表明，PixRestore（约5000万参数，单步推理）在竞争的UIR模型中实现了最佳的整体保真度、感知质量和退化鲁棒性，且效率更高。更大的模型变体可以进一步提升性能。

Insight: 创新点在于：1）提出了一种无需VAE、从零开始训练的像素空间扩散Transformer（DiT）架构，避免了潜在空间方法可能丢失细节的问题；2）引入基于LQ-HQ DINO特征相似度的层特征可靠性预测机制，自适应地融合可靠特征并监督不可靠层以去除退化；3）展示了通过基于DINO的对抗目标将模型微调为单步生成器以实现高效推理的可行性，证明了像素空间设计的可扩展性。

Abstract: Unified image restoration (UIR) aims to recover high-quality (HQ) content from low-quality (LQ) images with different degradations using a single model. Most recent methods adapt large pretrained text-to-image (T2I) latent diffusion models for their strong capacity and generative priors. However, the variational autoencoder (VAE) in latent T2I models may discard restoration-sensitive details, while the open-ended synthesis prior can introduce content-inconsistent artifacts. We present PixRestore, a VAE-free pixel-space Diffusion Transformer (DiT) for UIR, where the diffusion backbone is trained entirely from scratch, without relying on T2I pretraining. PixRestore performs flow matching directly on patchified pixels, preserving fine-grained details while keeping the token sequence tractable. To adapt to different degradations, PixRestore learns to predict the reliability of layer features using LQ–HQ DINO feature similarity. Features from more reliable layers are fused as dense conditioning, while less reliable layers receive stronger HQ-feature supervision to encourage degradation removal. We train PixRestore on a large-scale corpus of diverse scenes and degradations, and further finetune it into a one-step generator using DINO-based adversarial objectives for efficient inference. Experiments on public benchmarks and real-world test sets show that, with only about 50M parameters and single-step inference, PixRestore achieves the best overall fidelity, perceptual quality, and robustness to degradations among competing UIR models while being far more efficient. Larger PixRestore variants can further boost performance, demonstrating the scalability of our pixel-space design. Code and the curated benchmark can be found at https://github.com/csslc/PixRestore.

[163] Calibration-Free Vehicle Speed Estimation: A Monocular Keypoint-Template Approach cs.CVPDF

Gaofeng Su, Keya Li, Raja Sengupta, Kara M. Kockelman

TL;DR: 本文提出了一种无需标定的单目视频车辆速度估计框架，该方法不依赖道路特征、相机标定或基于道路特征的参考物。该框架使用一个36关键点的车辆模板和逐帧更新的单应矩阵来估计速度，并比较了仅关键点跟踪与带密集空间聚合的扭曲光流两种策略。在包含400多个视频片段的数据集上验证，速度覆盖30至100英里/小时。

Details

Motivation: 解决现有基于视觉的速度估计方法依赖道路特征、相机标定或特定参考物的局限性，旨在实现低成本、高效的便携设备（如行车记录仪和智能手机）速度执法，支持基于公民的交通安全执法项目。

Result: 在VS13和BrnoCompSpeed数据集上，扭曲光流方法分别实现了15.0%和9.7%的平均绝对误差（MAE），77.9%和93.1%的估计误差在±20%以内。经过10%的修剪去除边缘异常值后，MAE分别提升至11.7%和7.6%，±20%以内的准确率分别提升至85.3%和95.4%。

Insight: 创新点在于提出了一种完全无需标定的单目速度估计框架，结合了预定义的车辆关键点模板和动态更新的单应矩阵，将像素位移投影到度量空间；客观分析认为，该方法通过关键点模板和光流增强，减少了对环境先验知识的依赖，提高了在多变场景下的鲁棒性和实用性。

Abstract: This paper proposes a calibration-free framework for reliably and effectively estimating vehicle speeds from monocular videos, without relying on roadway features, camera calibration, or roadway-feature-based reference objects. The proposed framework estimates vehicle speeds using a 36-keypoint vehicle template and a homography matrix updated at each frame. A YOLO-based keypoint detection module is trained on diverse datasets, and two estimation strategies are compared: keypoint-only tracking and warped optical flow with dense spatial aggregation. Speed is estimated by projecting displacements into metric space using the homography, with validation conducted on over 400 video clips from roadside and overhead datasets, covering speeds from 30 to 100 mph. The method achieves reliable speed estimation on the VS13 and BrnoCompSpeed datasets, with the warped optical flow method delivering MAEs of 15.0% and 9.7%, respectively, and 77.9% and 93.1% of estimates falling within +/-20% error. After applying a 10% trim to remove edge-of-frame outliers, performance improves to MAEs of 11.7% and 7.6%, with within-+/-20% accuracy increasing to 85.3% and 95.4%. This work addresses key limitations of existing vision-based approaches and enables low-cost and efficient speed enforcement using portable devices such as dashcams and smartphones, thereby supporting citizen-based enforcement programs for traffic safety.

[164] Unlocking the Potential of Image Editing via Concept Scaling and Dense Supervision cs.CVPDF

Long Cui, Xiaoqian Liu, Qi Qin, Yi Xin, Tao Lin

TL;DR: 该论文针对现有图像编辑框架在概念粒度和监督信号稀疏性方面的不足，提出了一个包含1000多个细粒度编辑概念的层次化分类体系，并构建了包含1200万高质量编辑对的ConceptEdit-12M数据集。通过改进的数据合成框架和密集监督训练策略，有效提升了模型的训练效率和性能，并提出了一个全面的评估基准ConceptEdit-Bench。

Details

Motivation: 现有基于文本到图像扩散模型的图像编辑范式存在两个固有缺陷：对编辑概念粒度关注不足，以及稀疏监督信号导致的训练效率低下。

Result: 所提方法在训练结果上显著优于先前工作，并通过ConceptEdit-Bench评估套件在大量真实场景中诊断模型能力。

Insight: 创新点包括建立细粒度编辑概念分类体系、构建大规模高质量数据集以纠正生成数据分布坍塌、提出密集监督训练策略以合成多个非干扰概念来提供更丰富的学习信号，从而提升训练效率和模型性能。

Abstract: Existing image editing frameworks predominantly follow the training paradigm of text-to-image diffusion models. However, extending this paradigm to image editing highlights two inherent discrepancies, specifically, the insufficient attention to edit concept granularity and the training inefficiency caused by sparse supervision signals. To address these issues, we establish a comprehensive hierarchical taxonomy featuring over 1,000 fine-grained edit concepts and build ConceptEdit-12M, a massive dataset of 12 million high-quality editing pairs via an improved synthesis framework. This library-driven approach effectively rectifies the distribution collapse of generated data while ensuring high data fidelity. Furthermore, we propose a dense supervision training strategy that synthesizes multiple non-interfering concepts into single image pairs. By providing richer learning signals, this strategy significantly enhances both training efficiency and overall model performance. Training results validate our strategy, significantly outperforming prior works. Finally, we present ConceptEdit-Bench, a granular evaluation suite designed to diagnose model capabilities across a vast array of real-world scenarios.

[165] HarnessEval-W: Agentifying the Evaluation of Visual Worlds cs.CVPDF

Weiliang Chen, Haowen Sun, Jun Gao, Jiawei Chi, Hanyang Wang

TL;DR: 本文提出了HarnessEval-W，一个用于评估视觉世界模型的代理化评估框架。它通过将评估问题分解为可测量的子问题，并生成专门的子代理进行推理，构建透明的证据树来为评分提供可验证的推理链，从而超越了仅提供标量分数的传统基准。

Details

Motivation: 现有世界模型基准的评估指标是暴力计算的，缺乏可被检查或验证的推理链，无法判断物理、因果和世界状态是否正确演化。论文旨在解决评估的可信度和透明度问题。

Result: 该方法在330个评估案例上应用于18个代表性世界模型，其判断与人类偏好高度一致，并为每个生成的推演提供了可验证的、细粒度的诊断。

Insight: 核心创新点是将LLM领域的评估范式（harness paradigm）引入世界模型基准测试，通过分层代理工作流实现透明、可解释的评估，将每次评估转化为一个可追溯推理过程的证据树。

Abstract: A benchmark should deliver more than a scalar score: what makes an evaluation trustworthy is the reasoning that justifies the score. This is especially critical for world models, where judging a rollout requires understanding whether physics, causality, and world state evolve correctly. Humans spot such violations naturally, yet no existing benchmark automates this capability: metrics are computed brute-force, leaving no reasoning chain that can be examined or verified. We introduce HarnessEval-W, an agentified evaluation pipeline that brings the harness paradigm from the LLM ecosystem to world model benchmarking. Rather than applying a fixed rubric, HarnessEval-W interprets the context of each evaluation case, decomposes the evaluation question into measurable subproblems, and spawns specialized sub-agents, each equipped with tailored context and diagnostic tools to reason over its own subproblem. The parent agent then validates the gathered evidence and summarizes it into the final verdict. This hierarchical workflow turns every evaluation into a transparent evidence tree whose complete reasoning chain justifies the result. We apply HarnessEval-W to 18 representative world models over 330 evaluation cases. Its judgments closely align with human preferences while providing verifiable, fine-grained diagnoses of every generated rollout. We open-source the full pipeline as a live benchmark and invite the broad community to contribute to grow new skills and evaluation cases as world models evolve.

cs.RO [Back]

Haolin Yang, Yuxing Long, Zihan Yang, Hao Dong

TL;DR: VTInstructor是首个用于连续环境的视觉语言导航指令生成框架。它通过将密集的RGB视频流中的隐含轨迹几何信息转换为显式的视觉轨迹提示，解决了在连续环境中因缺乏离散视角图而难以恢复轨迹线索的挑战。该方法在R2R-CE和RxR-CE基准测试中取得了新的SOTA性能，并显著提升了下游导航代理的成功率。

Details

Motivation: 解决在连续环境（仅接收密集RGB视频流）中生成导航指令的挑战。传统方法依赖离散的视角图和全景观测，而连续环境缺乏显式的轨迹结构，使得轨迹线索难以恢复。

Result: 在R2R-CE和RxR-CE Val Unseen基准测试上，VTInstructor在所有标准自然语言生成指标上均达到新的SOTA，分别超过最强基线+0.357 CIDEr和+0.109 CIDEr。生成的指令使冻结导航代理的成功率提升至63.3%（相对最佳对比源提升14.7个百分点），并为下游导航任务带来+3 SR点的数据增强增益。

Insight: 核心创新是将隐含轨迹几何显式化为视觉轨迹提示，通过关键帧提取、视觉提示叠加、视觉编码器注入和训练校准四个模块实现，无需导航图、预建地图或场景重建。这种将空间轨迹信号直接注入视觉表示的方法，为连续环境下的指令生成提供了新范式。

Abstract: Navigation instruction generation from ego-centric RGB video in continuous environments is an important yet challenging task for human-robot interaction and scalable dataset construction. Prior instruction generators assume discrete viewpoint graphs with panoramic observations, where trajectory structure is explicit; in continuous environments, however, the agent receives only a dense RGB stream, making trajectory cues difficult to recover. We propose VTInstructor, the first VLN instruction generation framework for continuous environments. Our key idea is to convert implicit trajectory geometry into explicit visual trajectory prompts: EDTC condenses long RGB trajectories into navigation-critical keyframes, VTP overlays path, turn, and goal cues onto these anchors, VTMod injects the resulting trajectory signals into the visual encoder, and VT-GRPO further calibrates this spatial injection during training, all without requiring a navigation graph, pre-built map, or scene reconstruction. On the challenging R2R-CE and RxR-CE Val Unseen benchmarks, VTInstructor sets a new state of the art across all standard NLG metrics, surpassing the strongest baseline by +0.357 CIDEr and +0.109 CIDEr, respectively. Beyond automatic metrics, VTInstructor-generated instructions raise a frozen follower’s success rate to 63.3%, a +14.7 percentage-point gain over the best competing instruction source, and provide consistent data augmentation gains of +3 SR points on downstream navigation tasks.

[167] Neurosymbolic Embodied Agents cs.RO | cs.AI | cs.CLPDF

Mohammad Albinhassan, Yuming Feng, Alessandra Russo, Pranava Madhyastha

TL;DR: 本文提出了一种神经符号化智能体，通过将长时域家庭任务分解为任务导向的视觉探索和受限符号规划，以解决语言和视觉语言模型生成的可执行性差的问题。该方法利用视觉语言模型获取符号初始状态，并通过PDDL转移模型和蒙特卡洛树搜索生成可执行计划，在VirtualHome和ALFWorld环境中实现了超过90%的成功率。

Details

Motivation: 解决语言和视觉语言模型生成的具身计划可能违反环境动态或基于错误实体的问题，确保生成计划的可靠执行性。

Result: 在VirtualHome和ALFWorld基准测试中，4B-27B开放模型在两个环境中均超过90%的成功率，最小智能体显著优于27B直接视觉策略；约束和搜索结合解决超过95%的任务，而单独使用任一方法仅解决不到三分之一。

Insight: 创新点在于将神经符号化方法结合，通过视觉探索获取符号状态和受限规划确保可执行性，无需专门训练即可定位失败原因，且生成令牌和可见图像数量大幅减少，提高了效率和可靠性。

Abstract: Language and vision-language models generate plausible embodied plans but do not guarantee executability, as their outputs can violate environment dynamics or act on incorrectly grounded entities. We present a neurosymbolic agent that factors long-horizon household tasks into task-directed visual exploration and constrained symbolic planning. In the first phase, a vision-language model and exploration harness acquire goal-relevant predicates and instance bindings from egocentric observations and grounded interactions, producing a symbolic initial state. In the second, a PDDL transition model restricts decoding to tokens that extend applicable actions. Monte Carlo tree search then evaluates executable continuations using a domain-independent planning heuristic. The resulting plans are executable by construction under the transition model, with transfer to the environment conditioned on correct visual grounding. On VirtualHome and ALFWorld, open 4B-27B models exceed 90% success in both environments, and our smallest agent substantially outperforms a 27B direct visual policy in each. Constraints and search prove complementary rather than interchangeable: in ALFWorld either alone solves under a third of tasks, whereas their combination solves over 95%. The method also uses several times fewer generated tokens than extended thinking and far fewer model-visible images than direct interaction, and residual failures localize to state acquisition rather than plan generation without any specialized training.

[168] SpotlessGS: Relightable 3D Gaussian Splatting under Dynamic Illumination for Robotic Perception cs.RO | cs.CVPDF

Liang Hong, Jiaxin Wei, Simon Schaefer, Stefan Leutenegger, Jaehyung Jung

TL;DR: 本文提出SpotlessGS，一种可重照明的3D高斯溅射框架，旨在解决机器人在暗光环境下使用机载光源时因光照不均导致感知性能下降的问题。该方法通过联合优化光照参数、引入球谐函数低频光照模型以及基于MLP的BRDF模型，实现了对动态光照下非朗伯反射表面的高保真3D重建。

Details

Motivation: 现有基于2D图像增强的方法缺乏可靠监督且难以保持多视角几何一致性，因此需要一种能够处理动态光照、保持几何一致性的3D重建方法以提升机器人感知任务性能。

Result: 在合成和真实数据集上的实验表明，该方法有效减少了光照伪影，在渲染质量和定量性能上优于先前方法，并通过下游任务验证了其对机器人感知的益处。

Insight: 创新点包括在GS框架内联合优化光照参数以消除显式标定需求、使用球谐函数建模空间变化的残差和环境光照、以及引入MLP-BRDF处理非朗伯反射，从而实现了更灵活准确的可重照明3D重建。

Abstract: Robots operating in dark or poorly lit environments rely on onboard lights, which often produce uneven illumination that degrades downstream perception tasks. Prior approaches based on 2D image enhancement lack reliable supervision and fail to preserve multi-view geometric consistency. To address these limitations, we extend Dark Gaussian Splatting (DarkGS) toward a more accurate and flexible relightable 3D reconstruction framework. First, we eliminate the need for explicit light parameter calibration by jointly optimizing lighting parameters within the Gaussian Splatting framework. Second, we introduce a low-frequency illumination model based on spherical harmonics (SH) to capture spatially varying residual and ambient lighting effects. Third, we incorporate an MLP-based Bidirectional Reflectance Distribution Function (BRDF) to model non-Lambertian reflectance. Experiments on synthetic and real-world datasets demonstrate that our method effectively mitigates illumination artifacts while improving rendering quality and quantitative performance over prior approaches. We further validate its benefits for robotic perception through a downstream task.

[169] Imagining Recovery: Inference-Time Counterfactual Realignment for Vision-Language-Action Models cs.RO | cs.CVPDF

Yanyan Zhang, Disheng Liu, Kai Ye, Chaoda Song, Xinpeng Li

TL;DR: 本文提出了Counterfactual Realignment (CoRe)框架，用于在推理时无需训练即可恢复视觉-语言-动作（VLA）模型在机器人操作中的在线中断（如任务目标变化、场景配置改变或机器人状态扰动）。该方法通过想象从最近可行状态到当前目标的策略延续，并最小化对齐机器人和场景，以统一方式处理指令变更和物理扰动，避免物理试错并保留已完成任务进度。

Details

Motivation: 解决VLA模型在机器人操作中对在线中断（如任务目标、场景配置或机器人状态的变化）的脆弱性问题，现有恢复方法通常需要失败数据、策略重新训练或外部纠正代理，这引入了额外数据需求和执行风险。

Result: 在多个模拟器、VLA骨干和真实世界设置中的广泛实验表明，CoRe将成功率提高了高达85.0个百分点至接近名义水平，同时减少了42.2%的物理恢复，且无需策略微调或针对失败的恢复训练。

Insight: 创新点在于提出了一种无需训练的推理时反事实对齐框架，通过合成观察替代物理执行来想象策略延续，并最小化对齐以恢复控制，这统一处理了多种中断类型，避免了数据依赖和试错风险。

Abstract: Vision-language-action (VLA) models have improved the flexibility and generality of robotic manipulation, yet they remain fragile to online disruptions, such as changes in task goal, scene configuration, or robot state. Existing recovery methods often require failure data, policy retraining, or external corrective agents, introducing additional data requirements and execution risks. We propose Counterfactual Realignment (CoRe), a training-free framework that recovers a frozen VLA at inference time without failure data. Upon detecting a deviation, CoRe imagines how the policy would continue toward the current goal from a recent viable state, using synthesized observations in place of physical execution, and then minimally realigns the robot and scene to rejoin this imagined continuation before returning control to the policy. Recovery is therefore planned without physical trial-and-error, preserves completed task progress, and handles both mid-episode instruction changes and physical perturbations in a unified manner. Extensive experiments across multiple simulators, VLA backbones, and real-world settings show that CoRe improves success rates by up to 85.0 percentage points to near-nominal levels while reducing physical restorations by 42.2%, without policy fine-tuning or failure-specific recovery training.

Cong Xu, Ravi Sankar

TL;DR: 该论文提出了一种跨模态溯因风险感知框架，旨在当视觉感知失效时，通过处理互补的听觉模态信息来维持世界模型。该方法将预期共现证据的缺失视为隐藏原因的证据，利用麦克风阵列前端估计声源方位和接近率，在视觉证据缺失时触发溯因推理，从而发出校准后的风险预警。

Details

Motivation: 当主要视觉模态被遮挡或退化时，依赖视觉观测来填充结构化世界状态的方法会失效。论文旨在解决如何从互补模态（如听觉）中维持世界模型，通过将预期共现证据的缺失视为隐藏原因（如被遮挡的道路使用者）的证据，来持续感知风险。

Result: 在真实盲区路口被遮挡接近的录音测试中，该方法在视线进入前平均1.7秒发出预警，与已发表的声学基线方法的持续窗口变体检测率相当，但误报率减少了42%，一旦进入视野后中位定位精度为3.4度，校准良好（预期校准误差0.034），在视觉退化至0.03时仍能将危险感知保持在0.87以上。

Insight: 创新点在于提出了一种模态无关的溯因框架，将证据缺失本身作为推理隐藏状态的依据，并应用于听觉模态实现风险感知。该方法将隐藏状态的可恢复性分析为可识别性问题，并将提示建模为在明确误报预算下的Neyman-Pearson检测，实现了在视觉失效时通过跨模态信息维持世界模型的鲁棒性。

Abstract: A structured world-state (entities, relations, context, and predictive cues) is designed to preserve prediction-critical content when perception degrades, but it presumes observations to populate it; when the primary visual modality is occluded or degraded, those observations may be missing. We address how to sustain the world model from a complementary modality by treating the absence of expected co-evidence as evidence of a hidden cause. The abductive framework is modality-agnostic; this article instantiates it acoustically. A microphone-array front-end estimates the bearing of engine and tire sources and extracts approach-rate evidence (Doppler when a stable tone exists, a broadband looming readout otherwise); the event “signature present, visual co-evidence absent” then triggers abductive inference of a hidden road user, emitting a calibrated risk advisory rather than a control command. Recoverability of the hidden state is analyzed as an identifiability question separating shared from modality-unique information, and cueing is cast as Neyman-Pearson detection under an explicit false-alarm budget. On real occluded-approach recordings at blind junctions, the method warns a mean 1.7 seconds before line-of-sight entry, matches the sustained-window variant of the published acoustic baseline’s detection rate with 42% fewer false alarms, localizes to 3.4 degrees median once in view, is well calibrated (expected calibration error 0.034), and keeps hazard awareness above 0.87 under staged vision degradation that collapses a vision-only channel to 0.03. We also measure the method’s limits: calibration transfers to an unseen junction almost losslessly, the signature classifier does not, and moving-ego noise is the binding deployment constraint.

[171] ForceU-VLA: A Force-Aware Vision-Language-Action Model for Embodied Ultrasound Scanning cs.RO | cs.CVPDF

Xingzheng Wu, Cheng Zhang, Guihao Yan, Xifeng Hu, Zhi Liu

TL;DR: 本文提出了ForceU-VLA，一种用于自主实体超声扫描的力感知视觉-语言-动作模型。该模型通过融合超声视觉反馈和力信号，并引入阶段自适应调制机制，旨在解决现有方法中力与超声模态耦合松散、缺乏扫描阶段感知的问题，以提升扫描过程中探针-组织动态交互的捕获能力。

Details

Motivation: 现有实体智能超声扫描方法在力与超声模态的建模上耦合松散，且缺乏对扫描阶段的感知，这限制了其捕捉动态探针-组织交互的能力。本文旨在解决这些问题，以实现更准确、高质量的超声图像采集。

Result: 广泛的实验结果表明，ForceU-VLA在实体超声扫描中显著改善了接触稳定性和探针压力调节，从而有效提升了任务执行质量和整体系统可靠性。

Insight: 论文的创新点在于提出了力-超声协同融合模块（FUSFM）和阶段自适应调制机制（SAMM），前者实现了视觉与力反馈信息的协同融合，后者通过自适应调制多模态特征以适应不同扫描阶段的需求。此外，还贡献了一个真实世界的力感知实体超声数据集ForceU-VLA-Data。

Abstract: Embodied intelligent ultrasound scanning enables the automation and standardization of the ultrasound examination process by integrating perception, decision-making, and execution capabilities. However, existing methods suffer from loosely coupled modeling between force and ultrasound modalities and lack awareness of scanning stages, which limits their ability to capture dynamic probe-tissue interactions. To address these issues, we propose ForceU-VLA, a force-aware Vision-Language-Action model for autonomous embodied ultrasound scanning, which leverages force signals and ultrasound image feedback throughout the scanning process to enable accurate and high-quality ultrasound acquisition. Firstly, we propose a Force-Ultrasound Synergistic Fusion Module (FUSFM) that synergistically fuses ultrasound visual and force-feedback information to provide stable, reliable guidance for probe motion. Secondly, a Stage-Adaptive Modulation Mechanism (SAMM) is proposed to accommodate the task requirements across different scanning stages by adaptively modulating multimodal features to enhance their representation quality. Additionally, we introduce ForceU-VLA-Data, a real-world, force-aware embodied ultrasound dataset that integrates visual, force, and action signals, including data from two organs across five representative clinical scanning views, and comprising 450 expert-collected trajectories with approximately 100,000 synchronized multimodal frames. Extensive experimental results demonstrate that ForceU-VLA significantly improves contact stability and probe pressure regulation in embodied ultrasound scanning, thereby effectively enhancing task execution quality and overall system reliability. The source code is available at https://github.com/VMVLab/ForceU-VLA.

[172] MotionGS-SLAM: Event-Modulated Gaussian Splatting for Motion-Blur Robust SLAM cs.RO | cs.AI | cs.CVPDF

Zhiqiang Hu, Shouren Huang, Masatoshi Ishikawa

TL;DR: MotionGS-SLAM提出了一种新颖的SLAM方法，通过事件相机和3D高斯泼溅技术，将运动模糊处理重新定义为正向建模问题，而非传统的去模糊逆问题。该方法引入事件调制的高斯核，在渲染管线中生成性地建模模糊形成过程，从而在剧烈运动条件下实现鲁棒的姿态估计和地图构建。

Details

Motivation: 传统视觉SLAM系统在运动模糊严重破坏视觉输入时会完全失效，因为它们试图从退化的观测中恢复清晰内容，这是一个病态的逆问题。本文旨在从根本上重新构想运动模糊处理，将其转化为一个约束良好的正向问题。

Result: 大量实验表明，在严重的高运动条件下，该方法在轨迹精度和地图质量方面相比最先进（SOTA）方法有显著提升。

Insight: 核心创新在于范式转变：从“去除模糊”转向“正向建模模糊”。具体技术包括利用事件相机的高时间分辨率，提出事件调制的高斯核，通过空间（将各向同性点变为运动对齐的椭圆笔触）和时间（基于局部速度自适应采样密度）双重调制机制，实现相机轨迹与场景几何的联合优化。这为处理动态模糊场景的SLAM提供了新的物理启发式解决方案。

Abstract: Current Vision-based SLAM systems fail catastrophically when motion blur corrupts the visual input, as they attempt the ill-posed inverse problem of recovering sharp content from degraded observations. We present MotionGS-SLAM, which fundamentally reimagines motion blur handling through a paradigm shift: rather than removing blur artifacts, we reformulate the challenge as a well-constrained forward problem that generatively models blur formation within the rendering pipeline. By leveraging event cameras’ microsecond temporal resolution and immunity to motion blur, we introduce a novel event-modulated Gaussian kernel that dynamically adapts each Gaussian’s rasterization based on precise motion cues. Our dual-modulation mechanism transforms 2D Gaussian projections from isotropic dots into anisotropic, motion-aligned elliptical brush strokes (spatial modulation) while adaptively varying exposure integral sampling density based on local velocity (temporal modulation). This physics-based approach enables joint optimization of intra-exposure camera trajectories and 3D scene geometry through blur-aware photometric and event-based constraints. Extensive experiments demonstrate significant improvements over state-of-the-art methods in trajectory accuracy and map quality under severe high-motion conditions.

[173] Pre-training Visual Dexterity in Simulation cs.RO | cs.AI | cs.CVPDF

Sarthak Kamat, Adam Rashid, Satvik Sharma, Aseem Doriwala, Chelsea Finn

TL;DR: 本文提出了SPD（Simulation Pre-training for Dexterity）框架，通过VR头显在仿真环境中收集人类操作虚拟物体的轨迹数据，用于预训练灵巧手操作策略，并在真实世界的双臂灵巧机器人上通过少量演示进行微调，验证了仿真预训练对真实灵巧操作的有效性。

Details

Motivation: 解决灵巧手（多指手）操作数据稀缺的问题，因为真实遥操作成本高，而人类手部视频数据存在本体不匹配和姿态估计误差，因此探索利用仿真环境进行低成本、大规模数据收集。

Result: 在真实世界56自由度的双臂灵巧机器人设置上，仅用1-2小时的物理演示进行微调，SPD方法优于从零开始训练的行为克隆策略，证明了仿真遥操作作为预训练数据源的可行性。

Insight: 创新点在于利用VR仿真环境进行本体匹配的轨迹数据收集，避免了真实数据采集的高成本和视频数据的重定向损失；同时研究了历史条件化和短动作块对反应式控制的益处。

Abstract: Large-scale pre-training has made robot policy fine-tuning increasingly data-efficient, but this progress has largely been driven by datasets and embodiments built around simple parallel-jaw grippers. Dexterous, multi-fingered hands remain comparatively data-starved because real teleoperation is costly to scale, while human hand video is off-embodiment and requires lossy pose estimation and retargeting. We introduce Simulation Pre-training for Dexterity (SPD), a pre-training framework for dexterous manipulation that uses data entirely collected in simulation. In SPD, humans manipulate virtual objects inside a VR headset, enabling on-embodiment trajectories and robot-free collection. With the help of five operators, we collect 75 hours of multi-task dexterous manipulation over one week, and use it to pre-train a causal transformer on a sequence modeling objective. We study the benefits of simulation pre-training on real-world tasks by fine-tuning on 1-2 hours of physical demonstrations on a 56-DoF bimanual dexterous setup. We find that our approach outperforms training behavior cloning policies from scratch, showing that simulation teleoperation is a viable pre-training source for real-world dexterous manipulation. We perform ablation studies, measuring the benefits of history conditioning and short action chunks for reactive control.

[174] MM-BEV: Enhancing Timeliness by Computing Where and When it Matters cs.RO | cs.CV | cs.DC | eess.SYPDF

Liangkai Liu, Kang G. Shin

TL;DR: MM-BEV是一个实时多模态鸟瞰图感知系统，其核心设计原则是只在关键的时间和空间位置进行计算。该系统通过区分安全关键对象（如制动距离内、碰撞时间短的物体）的强制处理任务和非紧急区域的可选处理任务，并优先保证强制任务，从而在计算资源紧张时显著降低延迟。

Details

Motivation: 现有BEV感知方法计算成本高且难以实时部署，同时忽略了输入数据的结构化稀疏性、模态间的时间错位以及并非所有检测对象都影响规划器即时决策的事实。论文旨在解决这些挑战，实现高效实时的多模态BEV感知。

Result: 在nuScenes数据集上，MM-BEV将推理延迟降低了1.96倍，端到端延迟降低了2.93倍，同时几何关键召回率无损失，安全关键召回率仅下降0.2个百分点。在搭载Ouster-128激光雷达和Jetson AGX Orin的Clearpath Husky A300平台上，平均延迟进一步降低了2.11倍，验证了其在实际自动驾驶系统中的潜力。

Insight: 论文的创新点在于将感知任务按紧急程度（基于制动距离和碰撞时间）划分为强制任务和可选任务，并围绕此原则集成了四个机制：基于运动外推的关键性排序时间ROI选择器、稀疏且ROI感知的特征提取、根据场景动态和TTC调整的延迟感知协调器以及解耦感知与推理的异步调度器。这种以任务关键性为导向的自适应计算策略是提升实时系统效率的有效途径。

Abstract: Multimodal bird’s-eye-view (BEV) perception combines LiDAR depth accuracy with dense camera semantics, but its high computational cost and imperfect sensing conditions make real-time deployment challenging. Existing methods largely compress individual detectors and overlook three opportunities: structured sparsity within camera and LiDAR inputs, timing misalignment between modalities, and the fact that many detected objects do not affect the planner’s immediate action. We present MM-BEV, a real-time multimodal BEV system guided by a simple principle: compute where and when it matters. MM-BEV divides perception into mandatory work for safety-critical objects within braking distance of the ego vehicle and with short time-to-collision (TTC), and optional work for less urgent regions. It prioritizes mandatory work and reduces or sheds optional work under tight compute budgets. MM-BEV integrates four mechanisms: (1) a criticality-ranked temporal ROI selector based on motion-extrapolated detections from prior frames; (2) sparse, ROI-aware feature extraction using shared-shape camera crops at context-adaptive resolution and ROI-aware LiDAR voxelization; (3) a latency-aware coordinator that adapts LiDAR sweeps, image resolution, and keyframes according to scene dynamics and TTC; and (4) an asynchronous scheduler that decouples sensing from inference and skips stale frames. On nuScenes, MM-BEV reduces inference latency by 1.96x and end-to-end latency by 2.93x, with no loss in geometry-critical recall and only a 0.2 percentage-point drop in safety-critical recall. On a Clearpath Husky A300 equipped with an Ouster-128 LiDAR, BEV cameras, and a Jetson AGX Orin, MM-BEV further reduces mean latency by 2.11x, demonstrating its potential for real-world autonomous systems.

[175] US-VLA: An Ultrasound Vision-Language-Action Model for Embodied Abdomina cs.RO | cs.CVPDF

Cheng Zhang, Xingzheng Wu, Guihao Yan, Xifeng Hu, Zhi Liu

TL;DR: 本文提出了一种用于自动化超声扫描的超声视觉-语言-动作模型（US-VLA），该模型通过编码临床语义目标并基于实时超声反馈生成连续的探头操作动作，旨在解决现有强化学习和学习辅助超声扫描方法依赖精心设计的奖励函数或大量交互数据、泛化能力受限的问题。

Details

Motivation: 现有AI辅助超声扫描方法通常依赖于精心设计的奖励函数或大量交互数据，这限制了其在不同设备、患者群体和复杂临床场景中的泛化能力和稳定性。

Result: 在覆盖肝脏和肾脏检查的真实世界数据集US-VLA-Data上进行的大量实验表明，US-VLA在超声探头操作任务中取得了有竞争力的性能，显示了其在评估的腹部超声场景中的有效性和良好的泛化潜力。

Insight: 创新点包括设计了一个超声感知的专家融合模块，用于联合整合超声观测与辅助上下文信息，使语义超声反馈能有效指导扫描过程；并构建了一个包含五个临床定义标准平面、约8万个同步时间步的真实世界超声扫描轨迹数据集US-VLA-Data。

Abstract: Artificial intelligence-assisted ultrasound scanning enhances diagnostic reliability and efficiency by providing real-time guidance for standardized image acquisition and reducing operator dependence. However, existing reinforcement learning and learning-assisted ultrasound scanning methods typically rely on carefully designed reward functions or extensive interaction data, which limits their generalization ability and stability across different devices, patient populations, and complex clinical scenarios. To address these challenges, we propose an ultrasound vision-language-action model (US-VLA) for automated ultrasound scanning that explicitly encodes clinical semantic goals and generates sequential probe manipulation actions under real-time ultrasound feedback. In particular, we first design an ultrasound-aware expert fusion module to jointly integrate ultrasound observations with auxiliary contextual information, enabling semantic ultrasound feedback to effectively guide the scanning process. Then, we construct US-VLA-Data, a real-world dataset covering liver and kidney examinations, which includes five clinically defined standard planes and comprises 320 expert scanning trajectories with approximately 80,000 synchronized timesteps. Extensive experiments demonstrate that US-VLA achieves competitive performance in ultrasound probe manipulation tasks, indicating its effectiveness and promising generalization within the evaluated abdominal ultrasound setting. The source code is available at https://github.com/VMVLab/US-VLA.

[176] Don’t Drop the BATON: Long-Horizon Robot Manipulation via Agentic Subtask Exploration and Transition-aware Memory cs.RO | cs.AI | cs.CVPDF

Bingxin Xu, Yuzhang Shang, Emilio Ferrara

TL;DR: 本文提出了BATON方法，用于解决长视野机器人操作任务中存在的两个关键问题：多阶段探索成本指数级增长和子任务间状态转换缺乏表示。BATON将子任务作为探索单元，通过过渡感知记忆来管理子任务间的调用、交接和前瞻转换，从而在无需更新模型参数的情况下，显著提升了长视野任务的成功率。

Details

Motivation: 现有基于视觉-语言-动作模型和LLM智能体的方法在处理多阶段接触式操作任务时存在两大缺陷：一是全任务探索成本随阶段数指数增长且失败归因困难；二是缺乏对子任务间状态转换的显式表示，导致前序子任务的成功状态可能无法被后续子任务有效利用。

Result: 在长视野机器人操作基准测试RoboMemArena上，BATON方法将任务成功率提升了11.6%，累计成功率提升了14.9%，超越了当前最先进水平。

Insight: 核心创新在于将探索单元从整个长视野任务分解为独立的子任务，并通过引入过渡感知记忆来显式管理子任务间的状态转换，包括调用验证、状态恢复和策略前瞻，从而实现了探索成本的线性化（而非指数级）和失败的精确定位。

Abstract: Long-horizon robot manipulation chains many contact-rich skills into one multi-stage task. Vision-language-action (VLA) models increasingly master the individual skills, yet the chain still fails: errors compound beyond the policy’s ability to correct, and one subtask silently constrains the next. A promising recipe freezes the VLA and puts an LLM agent in charge: it plans in language, moves in free space with analytic primitives, invokes the VLA only for contact-rich segments, and writes adaptation into language memory. Applied to long horizons, it breaks twice. (1) Competence comes from whole-task exploration at test time, whose cost is multiplicative in stages: if one stage needs T episodes, a K-stage task needs about T^K, and a failure does not reveal which stage caused it. (2) It has no representation of transitions: the VLA primitive carries an exit but no entry condition, so a subtask can succeed in a form its successor cannot use. We present BATON. Against (1), BATON makes the subtask the unit of exploration: each is explored in the cheap short-horizon regime and its solution stored in memory; a long-horizon trajectory is then composed from these solutions rather than discovered whole. Cost becomes additive (T*K) and every failure is attributed to a single stage. Against (2), BATON equips exploration with a transition-aware memory. Within a subtask, a verifier agent governs the invocation transition: the VLA is called only after the wrist view confirms the scene is ready. Across subtasks, a handoff transition restores an entry state disturbed by the predecessor’s residue, and a lookahead transition selects the strategy whose outcome the successor can inherit. No parameters are updated. On the long-horizon benchmark RoboMemArena, BATON improves task success by 11.6% and cumulative success by 14.9% over the SoTA.

cs.DC [Back]

[177] FloodReasonBench: Benchmarking VLM Reasoning Segmentation for Embodied Flood Response at the Edge cs.DC | cs.AI | cs.CV | cs.RO | eess.SYPDF

Rajat Bhattacharjya, Yoomee Jung, Minwoo Kim, Sing-Yao Wu, Eli Bozorgzadeh

TL;DR: 本文提出了FloodReasonBench基准，用于评估面向边缘端具身洪水响应的视觉语言模型（VLM）推理分割能力。该基准包含一个基于真实洪水场景构建的专用数据集FloodResponseSeg，并从任务精度和系统层面（如轻量视觉编码、分层推理和压缩表示）对推理分割流程进行了全面评估。

Details

Motivation: 现有基准主要关注通用视觉场景，忽视了洪水响应平台面临的领域特定性和资源约束问题，因此需要构建一个针对边缘端具身洪水响应的专用基准。

Result: 评估显示，在通用预适应设置下，模型精度存在强烈的分区依赖性；而经过洪水适应优化的目标-工作负载设计空间，其精度范围在不同分区上更为紧凑。在NVIDIA Jetson AGX Xavier上的进一步评估揭示了推理分割精度、边缘端延迟、能耗和通信开销之间的权衡关系。

Insight: 创新点在于构建了首个面向边缘端具身洪水响应的VLM推理分割基准，并系统性地从任务精度和系统资源约束（如轻量编码、分层推理）两个维度对模型进行评估，为资源受限场景下的模型部署提供了质量约束下的边缘操作点选择依据。

Abstract: Reasoning segmentation enables vision-language models (VLMs) to translate mission-relevant language requests into pixel-level visual grounding, offering a natural perception interface for embodied agents. However, existing benchmarks largely focus on generic visual scenes and overlook the domain and resource constraints encountered in flood-response platforms. We present FloodReasonBench, a benchmark for VLM reasoning segmentation for embodied flood response at the edge. At its core, FloodReasonBench introduces FloodResponseSeg, a flood-specific reasoning-segmentation dataset constructed from real-world scenes and response-relevant targets. Beyond task accuracy, the benchmark characterizes reasoning-segmentation pipelines under lightweight visual encoding, hierarchical split inference, and compressed intermediate representations. We observe strong partition-dependent accuracy variation in the generic pre-adaptation setting, while the flood-adapted target-workload design space exhibits a substantially more compact accuracy range across partitions. Evaluation on an NVIDIA Jetson AGX Xavier further exposes the tradeoffs among reasoning-segmentation accuracy, edge-side latency, energy, and communication footprint, enabling quality-constrained selection of edge operating points. Together, these results provide a task- and system-level characterization of reasoning segmentation for resource-constrained embodied flood response at the edge.

cs.LG [Back]

[178] DUET: Dual-Teacher On-Policy Distillation via Same-Weight Disagreement for Prohibition Compliance cs.LG | cs.CLPDF

Zihan Li, Feifei Li, Wenhui Que

TL;DR: 本文提出了一种名为DUET的双教师策略内蒸馏方法，用于提升大型语言模型在运行时遵守动态注入的禁令（如企业政策、个人身份信息红线等）的能力。该方法利用一个能看到禁令的教师模型和一个相同权重但看不到禁令的教师模型，通过它们在token级别上的分歧来生成纯净的监督信号，从而指导学生模型学习。

Details

Motivation: 现实世界中的LLM部署需要遵守随请求和租户变化的运行时禁令，而传统的后训练方法（如SFT和DPO）在结构上不适合处理这种token级别的违规问题。

Result: 在涵盖五个任务系列的工业级禁令遵守基准测试中，DUET在1.5B到8B参数的Qwen模型变体上实现了72.3%到85.2%的违规遵守率，同时保持了88%到93%的正常效用，显著优于教师模型和其他蒸馏基线。在SysBench的外部评估中也证实了安全对齐的提升，且在GSM8K和MATH-500上的性能下降最小。

Insight: 创新点在于提出了基于相同权重教师模型分歧的token选择性策略内蒸馏框架，通过信号清洗和偏好导向学习两个互补机制，将DPO风格的优化直接嵌入到策略内蒸馏中，无需离线偏好数据，从而实现了对禁令因果效应的精准隔离和高效学习。

Abstract: Real-world LLM deployments increasingly rely on runtime-injected prohibitions–enterprise policies, PII redlines, tool boundaries–that vary per request and per tenant. Conventional post-training is structurally ill-suited: SFT hides the violation signal in compliant labels, and DPO’s sequence-level preferences mismatch token-localized violations. We propose DUET, a token-selective on-policy distillation method for prohibition compliance. DUET pairs a teacher that sees the prohibition (positive) with an identical-weight teacher that does not (negative). Because the two teachers differ only in prohibition visibility, their per-token disagreement isolates the prohibition’s causal effect–yielding a clean supervision signal uncontaminated by model capacity or mismatch. This disagreement drives two complementary mechanisms: signal cleaning, which discards agreement tokens as redundant or prefix-corrupted, and preference-directed learning, which pushes the student away from the negative teacher and toward the positive one at token granularity, embedding DPO-style optimization directly into OPD without offline preference data. We construct an industrial Prohibition-Compliance benchmark spanning five task families covering explicit-refusal, paraphrase robustness, and over-refusal. Across 1.5B-8B Qwen variants, DUET achieves 72.3-85.2% violation compliance while preserving 88-93% normal utility, dramatically outperforming teacher model and other distillation baselines. External evaluation on SysBench confirms improved safety alignment with minimal degradation on GSM8K and MATH-500.

[179] Proteus: Incremental Memory Activation for Long-Context Sequence Modeling cs.LG | cs.AI | cs.CLPDF

Reza Bayat, Ali Behrouz, Vahab Mirrokni, Aaron Courville

TL;DR: 论文提出了一种名为Proteus的增量内存激活机制，用于解决长上下文序列建模中静态内存模型因早期token占用过多自由度而污染内存状态、导致后期上下文容量不足和干扰增加的问题。该机制通过逐步扩展内存有效容量，强制模型更有效地压缩历史，并随时间释放新容量以减少干扰。

Details

Motivation: 动机在于基于注意力的序列模型在处理长上下文时存在二次成本，而现有内存模型通常使用静态内存，导致早期token因缺乏压缩压力而过度占用内存，污染状态并损害后期上下文的保留能力。

Result: 将Proteus应用于SWLA、Comba、Titans和Hope-Attention等最先进模型后，在标准语言建模、推理以及长上下文检索和理解任务上均观察到一致改进，且增益随上下文长度增加而增长，表明其能提升性能。

Insight: 创新点在于引入了增量内存激活范式，通过动态调度内存有效容量来优化压缩和减少干扰，这是一种简单且广泛适用于神经内存架构的机制，无需额外成本，揭示了静态内存的次优性。

Abstract: The quadratic cost of attention-based sequence models for long contexts has motivated a growing line of research on memory-based models that can compress context into a compact state. However, most existing memory models expose a static memory throughout the entire sequence. Because early tokens face no compression pressure, they occupy too many degrees of freedom and “pollute” the memory state, leaving little capacity for later context and increasing interference between what is stored and what arrives next. We study a new paradigm of incremental memory activation, where the effective capacity of memory is progressively expanded as the context grows. Imposing an early bottleneck forces the model to compress history more effectively, while unlocking fresh capacity over time reduces interference and improves retention of later context. We instantiate this paradigm in Proteus, a straightforward mechanism that can be incorporated into a broad class of neural memory architectures at no additional cost. We apply Proteus to state-of-the-art models, including SWLA, Comba, Titans, and Hope-Attention, and observe consistent improvements on standard language modeling and reasoning, as well as on long-context retrieval and understanding, with gains that grow at longer context lengths. Overall, our results show that static memory is suboptimal and that scheduling effective capacity is a simple and broadly applicable tool for sequence modeling.

[180] Diagnosing and Mitigating Perception-Decision Misalignment in Omni-LLMs via Modality Subspace Activation cs.LG | cs.CVPDF

Hongbo Jiang, Jie Li, Yunhang Shen, Tianyu Xie, Pingyang Dai

TL;DR: 本文针对全模态大语言模型（Omni-LLMs）中存在的感知-决策错位问题，提出了一个诊断框架和一个无需训练的推理时缓解方法。作者首先形式化了因果模态敏感性（CMS）概念，并通过宏观行为层面的答案保留率（ARR）和微观分布层面的对数概率角差异（LAD）进行量化诊断。基于此，他们构建了诊断数据集CausalMSBench，并发现主流Omni-LLMs的CMS极低。为纠正此问题，作者提出了模态子空间激活（MSA）方法，利用奇异值分解（SVD）估计模态激活强度，动态平衡最终隐藏状态中的模态投影，从而有效恢复模型的因果模态敏感性。

Details

Motivation: 全模态大语言模型在多模态推理任务中表现出色，但其决策往往与多模态感知不一致，存在深刻的感知-决策错位问题。论文旨在诊断并缓解这一模型内部的对齐缺陷。

Result: 在构建的诊断数据集CausalMSBench上的评测表明，流行的全模态大语言模型表现出极低的因果模态敏感性（CMS），即使关键模态被移除，其分布变化也微乎其微。提出的模态子空间激活（MSA）方法在无需训练的情况下，有效恢复了模型在多个基准测试上的CMS。

Insight: 论文的核心创新点在于形式化了因果模态敏感性（CMS）这一诊断指标，并提出了一个双视角（宏观行为与微观分布）的量化框架。此外，提出的模态子空间激活（MSA）是一种新颖的、无需训练的推理时干预方法，通过奇异值分解动态调整隐藏状态，为解决模型内部模态利用不均衡问题提供了新思路。

Abstract: Omni-Large Language Models (Omni-LLMs) power complex multi-modal reasoning in applications like World Action Models and autonomous agents. However, their strong performance often masks a profound Perceptual-Decision Misalignment (PDM), where decisions remain unfaithful to multi-modal perceptions. To diagnose this, we formalize Causal Modality Sensitivity (CMS), operationalized via a dual-lens framework: Answer Retention Rate (ARR) at the macro behavioral level, and Logit Angular Discrepancy (LAD) to track microscopic distribution shifts. We also curate CausalMSBench, a diagnostic dataset isolating language priors. Benchmarking reveals that popular Omni-LLMs exhibit critically low CMS, showing negligible distribution shifts even when key modalities are removed. To rectify this, we propose Modality Subspace Activation (MSA), a training-free inference-time framework that uses Singular Value Decomposition (SVD) to estimate modal activation strengths. MSA dynamically balances modal projections in the last hidden state, effectively restoring CMS across benchmarks.

[181] LUNG-KGMM: Knowledge-Guided Multimodal Learning for Lung Cancer Incidence Prediction cs.LG | cs.CVPDF

Chunlei Yang, Shuyan Li, Zhong Cao

TL;DR: 本文提出了LUNG-KGMM，一个知识引导的多模态学习框架，用于预测1至6年内的肺癌发病率。该框架整合了纵向电子健康记录、放射学报告、胸片表征以及指南衍生的知识图谱。为了解决模态异质性和潜在的数据泄露问题，作者开发了防泄露的报告处理流程和面向不完整随访的掩码累积训练目标。

Details

Motivation: 现有肺癌风险预测模型主要依赖单一数据模态，且无法有效利用结构化的临床知识，限制了早期识别的准确性。

Result: 在公开的MIMIC数据库开发队列上的大量实验表明，LUNG-KGMM的性能优于现有最先进方法。在厦门医疗大数据平台真实世界验证队列上的测试进一步验证了其跨队列的可移植性，但也指出了本地适配的必要性。

Insight: 创新点包括：1）构建了编码报告-发现-属性-行动关系的可审计临床指南知识图谱作为知识流；2）设计了防泄露报告处理流程和掩码累积训练目标以处理模态异质性和不完整随访数据；3）实现了多模态临床数据的深度融合与知识引导。

Abstract: Early identification of lung cancer risk is critical for timely intervention, yet existing prediction models are limited by their reliance on single data modalities and their inability to leverage structured clinical knowledge. We propose LUNG-KGMM, a knowledge-guided multimodal framework that integrates longitudinal electronic health records, radiology reports, chest radiograph representations, and guideline-derived knowledge for 1-to-6-year incident lung cancer prediction. To address modality heterogeneity and potential data leakage, we develop a leakage-sanitized report processing pipeline and a horizon-masked cumulative training objective that handles incomplete follow-up. We further introduce a knowledge-graph representation of clinical guidance that encodes report-triggered finding-attribute-action relations as an auditable knowledge stream. We build a multimodal development cohort from the publicly available MIMIC databases and construct a real-world validation cohort from the Xiamen Medical Big Data Platform. Extensive experiments on the MIMIC cohort demonstrate that LUNG-KGMM achieves superior performance over state-of-the-art methods, and validation on the Xiamen cohort further characterizes its cross-cohort portability and the need for local adaptation. The MIMIC development cohort is publicly accessible; the Xiamen cohort is governed by local data privacy regulations.

[182] OceanDepths: A Global Dataset of Paired Subsurface and Surface Ocean Observations cs.LG | cs.AI | cs.CVPDF

Simon Donike, Ruben Cartuyvels, Antonino Ian Ferola, Elisa Carli, Diego Fernandez Prieto

TL;DR: 本文介绍了OceanDepths数据集，这是首个开放的、全球性的、经过重网格处理的AI就绪数据集，它将卫星反演的海表温度、盐度和高度L4产品与同位的EN4次表层温盐剖面配对，并辅以匹配的GLORYS12海洋再分析数据。该数据集覆盖2000-2024年，空间分辨率为0.1°×0.1°，时间分辨率为周，包含超过950万个配对剖面，并插值到50个标准深度层，旨在为AI方法提供一个具有挑战性的测试平台。

Details

Motivation: 全球海洋观测严重不足，且缺乏一个标准化的、高分辨率的、将卫星海表场与同位现场深度剖面耦合的AI就绪数据集，现有资源多为模型重建的网格产品、仅覆盖单一变量或海盆，或分辨率过低，无法满足中尺度动力学研究需求。

Result: 论文构建了OceanDepths数据集，其具有4D多变量结构、高分辨率、长时间跨度和次表层观测极端稀疏（约每深度层0.01%）的特点，并展示了使用简单基线模型进行次表层状态重建的示例任务。

Insight: 主要创新在于创建了首个全球、开放、配对的表层与次表层海洋观测AI数据集，填补了标准化高分辨率数据集的空白；其设计的空间分块配置系统和极端稀疏的次表层观测特性，为开发新型AI方法（如基于观测的预报方法）提供了独特且具有挑战性的测试场景。

Abstract: Despite comprising over 70% of its surface, the world’s oceans are critically underobserved compared to the land surface or the atmosphere.Understanding the global ocean requires jointly observing its surface and subsurface structure, yet no standardized, high-resolution dataset couples satellite surface fields to co-located \emph{in situ} depth profiles in an AI-ready format.Existing resources either consist of model-reconstructed gridded products rather than observations, cover only a single variable or basin, or operate at resolutions too coarse for mesoscale dynamics.We introduce \textsc{OceanDepths}, the first open, global, regridded AI-ready dataset that pairs satellite-derived sea surface temperature (SST), sea surface salinity (SSS), and sea surface height (SSH) L4 products with co-located EN4 subsurface temperature and salinity profiles, complemented by matched GLORYS12 ocean reanalysis data to support comparisons or multi-stage learning.The dataset spans 2000–2024 at \SI{0.1}{\degree}$\times$\SI{0.1}{\degree} spatial resolution and at weekly temporal resolution, covering the entire globe’s sea surface and with over 9.5 million paired profiles interpolated to 50 standardized depth levels.We provide a configurable system to split the globe in equally sized spatial patches.The 4D multivariate structure, high resolution, long temporal extent, and extreme sparsity of subsurface observations (${\sim}$0.01% per depth level) make \textsc{OceanDepths}a challenging testbed for novel AI methods.We demonstrate subsurface state reconstruction as an example task with simple baseline models, but also envision \textsc{OceanDepths}to support the development of observation-based forecast methods and other related tasks.\added{Available at: https://huggingface.co/datasets/ESA-philab/OceanDepths.}

cs.HC [Back]

[183] FollowUpBot: An LLM-Based Conversational Robot for Automatic Postoperative Follow-up cs.HC | cs.CL | cs.ROPDF

Chen Chen, Jianing Yin, Jiannong Cao, Zhiyuan Wen, Mingjin Zhang

TL;DR: 本文提出了FollowUpBot，一种基于大语言模型（LLM）的边缘部署对话机器人，用于自动化术后随访。它通过动态规划最优路径、利用边缘部署的LLM进行自适应面对面多模式交互，并自动生成结构化随访报告，旨在解决传统随访方式耗时耗力、现有数字方案交互僵化或存在隐私泄露的问题。

Details

Motivation: 传统术后随访方式（如床边访谈和手动记录）效率低下，而现有的数字化解决方案（如网络问卷和智能自动呼叫）存在交互不灵活或隐私泄露的局限性。本文旨在开发一个既能减轻医护人员负担，又能提供灵活、安全交互的自动化随访系统。

Result: 实验结果表明，该机器人在随访互动中实现了高覆盖率和满意度，并且在多种现场类型下都表现出高精度的报告生成能力。

Insight: 创新点在于将LLM与边缘计算结合，部署于移动机器人平台，实现了动态路径规划、自适应面对面多模式对话以及自动结构化报告生成，在保障数据隐私的同时提升了术后随访的自动化和智能化水平。

Abstract: Postoperative follow-up plays a crucial role in monitoring recovery and identifying complications. However, traditional approaches, typically involving bedside interviews and manual documentation, are time-consuming and labor-intensive. Although existing digital solutions, such as web questionnaires and intelligent automated calls, can alleviate the workload of nurses to a certain extent, they either deliver an inflexible scripted interaction or face private information leakage issues. To address these limitations, this paper introduces FollowUpBot, an LLM-powered edge-deployed robot for postoperative care and monitoring. It allows dynamic planning of optimal routes and uses edge-deployed LLMs to conduct adaptive and face-to-face conversations with patients through multiple interaction modes, ensuring data privacy. Moreover, FollowUpBot is capable of automatically generating structured postoperative follow-up reports for healthcare institutions by analyzing patient interactions during follow-up. Experimental results demonstrate that our robot achieves high coverage and satisfaction in follow-up interactions, as well as high report generation accuracy across diverse field types. The demonstration video is available at https://www.youtube.com/watch?v=_uFgDO7NoK0.

[184] Do Assessment Instruments Measure the Same Thing for Humans and LLMs? A Latent Structure Analysis cs.HC | cs.AI | cs.CLPDF

Alona Strugatski, Licol Zeinfeld, Giora Alexandron

TL;DR: 本研究通过潜在结构分析，探讨了人类与大型语言模型（LLMs）在教育评估工具（如高中化学和大学入学考试定量推理部分）中是否测量相同的潜在能力。研究发现，人类与LLMs的因子结构存在系统性差异，表明这些评估工具可能无法在两者间测量相同的构念。

Details

Motivation: 随着LLMs的广泛应用，常使用为人类设计的评估工具（如标准化考试）来评估LLMs的能力，但需验证这些工具在人类和LLMs中是否测量相同的潜在构念，以确保评估结果的有效性。

Result: 在高中化学和大学入学考试定量推理两个案例中，通过探索性因子分析、因子一致性和重采样方法，发现人类与六个多模态LLMs的响应数据在潜在结构上存在显著差异，表明评估工具可能不适用于跨物种的能力比较。

Insight: 论文创新点在于从心理测量学角度质疑了使用人类评估工具直接评估LLMs的有效性，强调了跨物种评估中潜在结构相似性的必要性，为AI评估实践提供了重要的方法论警示。

Abstract: The rapid development and growing deployment of large language models (LLMs) have made it increasingly important to understand their capabilities. A common approach is to evaluate LLMs using assessment instruments originally designed to measure skills and competencies in humans, such as standardized exams, and to use performance on these instruments as evidence for generalizable claims about LLMs’ underlying abilities on the same skills the assessments are intended to measure in humans. However, from a validity perspective, such inferences require that the relationship between observed performance and underlying constructs established for humans also holds for LLMs. In particular, a necessary condition for transferring score interpretations is similarity in the latent structure of responses to the assessment. In this study, we examine whether this condition holds in two educational contexts: high-school chemistry and a quantitative reasoning section of a university entrance exam. Using a case study design, we compare human response data with responses generated by six multimodal LLMs. Our analytical approach combines exploratory factor analysis, factor congruence, and resampling to assess latent structure similarity across human learners and LLMs. Across both instruments, we find systematic differences between human and LLM factor structures, showing evidence that the analyzed assessments may not measure the same constructs for humans and LLMs. These findings call into question the validity of evaluation practices that use educational assessments to make claims about AI capabilities.

[185] PolyDebate: A Game-Orchestrated Multimodal System for Debate Skills Practice and Evaluation cs.HC | cs.CLPDF

Jianing Yin, Weng Pan Kuan, Xiaoyun Liu, Zhiyuan Wen, Yuxuan Li

TL;DR: PolyDebate是一个游戏化编排的多模态系统，用于英语辩论技能练习与评估。它通过引导学习者与AI对手进行分阶段的一对一辩论，并利用技能卡、道具和硬币等游戏化元素使说服策略显性化，从而提供完整的多模态实践体验。系统在辩论过程中捕捉学习者的语音和视觉表达证据，生成情境感知的对手回应，并提供基于量规的阶段性和整体反馈。

Details

Motivation: 辩论是一种结构化的说服性交流，能训练论证构建、反驳、口头表达和听众意识等技能，这些技能在教育、语言学习和专业交流中备受重视。然而，现有的AI辩论系统和基于LLM的评判大多以文本为中心，很少能为学习者提供完整的多模态实践体验。

Result: 通过四项研究（涵盖AI对手质量、评估覆盖度、AI评委反馈和用户感知）表明，PolyDebate成功地将辩论互动、游戏化支架、多模态评估和结构化反馈整合到一个实用的辩论技能练习工作流程中。

Insight: 论文的创新点在于将游戏化机制（如技能卡、道具和硬币）与多模态辩论实践深度融合，使抽象的辩论策略变得具体可操作。系统同时提供沉浸式3D游戏和Web平台两种版本，共享相同的工作流和评估服务，实现了灵活且一致的练习体验。从客观角度看，其将语音、视觉证据捕捉与情境感知的AI响应相结合，为辩论技能评估提供了更全面的多模态框架。

Abstract: Debate is a structured form of persuasive communication that trains argument construction, rebuttal, oral delivery, and audience awareness. These skills are valued in education, language learning, and professional communication. Recent AI debate systems and LLM-based judges have advanced argument generation and debate evaluation, but most remain text-centered and rarely support learners through a complete multimodal practice experience. We introduce PolyDebate, a game-orchestrated multimodal system for English debate practice and evaluation. PolyDebate guides learners through staged one-on-one (1v1) debates with an AI opponent, while skill cards, props, and coins make persuasive strategies explicit and turn practice into a game-like interaction. During each session, the system captures learner speech and visual delivery evidence, generates context-aware opponent responses, and produces rubric-informed stage-level and overall feedback. PolyDebate is available as both an immersive Unity 3D game version and a web platform version that share the same workflow and evaluation services. Four studies covering AI opponent quality, evaluation coverage, AI judge feedback, and user perception show that PolyDebate brings debate interaction, gamified scaffolding, multimodal assessment, and structured feedback together in a practical workflow for debate skills practice. The demonstration video is available at https://youtu.be/mHwBG1_8Ebk.

Zi Haur Pang, Casey Kennington, Tatsuya Kawahara

TL;DR: 本文提出了AffectLoop系统，一个部署在Misty II机器人上的多模态说话者-听者情感动态感知的语音对话系统。该系统通过追踪说话者的言语和面部情感动态，并估计机器人听者自身的言语和行为情感状态，来指导基于LLM的回应生成，从而形成一个闭合的说话者-听者情感循环。

Details

Motivation: 现有共情对话系统通常以文本为中心，并将共情建模为用户情感到系统回应的单向映射，这限制了其捕捉具身化的说话者-听者情感交换的能力。

Result: 在一项包含五名参与者的初步受试者内研究中，与一个省略了说话者和听者情感状态输入的基线系统相比，所提出的系统获得了更高的总体印象评分，尤其是在共情回应和用户满意度方面。事后日志分析进一步显示了更高的说话者-听者情感对齐和更强的基于情感效价的痛苦恢复能力。

Insight: 核心创新在于明确建模了说话者的情感动态和听者的情感状态，形成了一个闭合的、多模态的情感循环，从而提升了具身化共情交互的效果。这为构建更自然、更具情感交互能力的社交机器人提供了新思路。

Abstract: Empathetic social robots should respond not only to what users say, but also to how their emotions dynamically evolve during interaction. However, existing empathetic dialogue systems are often text-centered and primarily model empathy as a one-way mapping from the user’s emotion to the system response, limiting their ability to capture embodied speaker–listener affective exchange. We present AffectLoop, a multimodal speaker-listener emotion-dynamics-aware spoken dialogue system implemented on the Misty II robot. The system tracks the speaker’s verbal and facial affective dynamics, estimates the robot listener’s own verbal and behavioral affective state, and conditions LLM-based response generation on both affective streams. The robot then generates a short spoken empathetic response together with emotionally congruent embodied behavior, forming a closed speaker–listener affective loop. We evaluate the system in a pilot within-subject study with five participants, comparing it with an otherwise identical utterance-conditioned baseline that omits the speaker- and listener-affective-state inputs. The proposed system received higher overall impression ratings, especially for empathetic response and user satisfaction. Post-hoc log analysis further showed higher speaker-listener affective alignment and stronger valence-based distress recovery. These preliminary results suggest that explicitly modeling both speaker emotional dynamics and listener affective state can improve embodied empathetic interaction.

cs.CY [Back]

[187] Plausible but Not Valid: A Psychometric Audit of LLMs as Synthetic Survey Respondents cs.CY | cs.AI | cs.CL | stat.APPDF

Mantas Lukauskas, Viktorija Šarkauskaitė

TL;DR: 该论文对大型语言模型作为合成调查受访者的有效性进行了心理测量学审计。研究发现，尽管LLMs能复现人类心理测量关系的定性方向，但其生成的样本在联合分布、潜在结构、可靠性等方面与真实人类调查数据存在显著差异，不能直接替代人类数据。

Details

Motivation: 解决现有评估仅关注LLMs生成答案在个体层面的合理性，而缺乏从心理测量学角度评估其是否保留真实人类调查数据的联合分布、潜在结构、可靠性等关键属性的问题。

Result: 在立陶宛组织心理学数据集上评估了37个模型。高斯copula基线在样本驱动的PSS组件上优于所有LLM；LLM“群体”内部相似性高于与人类的相似性；反事实交换显示教育驱动的效应远大于性别和角色；所有LLM都表现出强烈的默许偏移，且用合成数据训练的回归器在预测真实人类数据时有效性下降。

Insight: 提出了心理测量相似性评分框架，系统评估LLMs作为合成受访者的心理测量特性；揭示了LLM生成数据与人类数据在统计结构上的根本差异，警示了直接替代的风险；通过反事实分析量化了不同人口统计属性的影响差异。

Abstract: Large language models (LLMs) are increasingly used as synthetic survey respondents, but existing evaluations ask whether answers look plausible at the individual level. We argue the right question is psychometric: do LLMs preserve the joint distribution, latent structure, reliability, mediation pathways, and demographic effects of real human survey data? We introduce a Lithuanian organisational-psychology dataset (n=263 employees; Dunham Attitudes Toward Change, UWES-17, Koopmans IWPQ; 68 items, 12 subscales) and condition a 37-model lineup spanning OpenAI, Anthropic, Google, and twelve open-weight families on real respondent profiles under a five-level persona-disclosure ladder, presentation and reasoning-effort ablations, counterfactual demographic swaps (gender, role, education), a cross-language check, and a verbatim-recall memorization probe. The resulting Psychometric Similarity Score (PSS) is anchored against five non-LLM statistical baselines and a held-out human-vs-human ceiling, with respondent-bootstrap confidence intervals and an item-permutation null for Tucker’s phi. LLMs reproduce the qualitative direction of human psychometric relationships, but a Gaussian-copula baseline beats every LLM on the sample-driven PSS components; the LLM “crowd” is more similar to itself (mean inter-LLM PSS 0.73) than to humans; and memorization does not drive the leaderboard (recall-PSS rank correlation 0.00). Counterfactual swaps reveal education-driven effects (mean |d|=0.56) that dwarf gender (0.12) and role (0.18); Tucker’s phi on UWES falls inside the permutation null for 8 of 37 models. Downstream, every LLM shows a strong acquiescence shift (+0.84 SD), synthetic-trained regressors lose predictive validity on held-out humans (mean R^2 -0.18 vs 0.28), and models fabricate indirect effects on 3 of 10 placebo mediation paths. LLM samples are not a drop-in replacement for human survey data.

cs.GR [Back]

[188] AnyTalk: Speech Animation for Arbitrary Characters Leveraging a Video Generation Model cs.GR | cs.CV | cs.MM | cs.SDPDF

Kwan Yun, Serin Yoon, Sunjin Jung, Jung Eun Yoo, Inyup Lee

TL;DR: AnyTalk是一种无需动画数据即可为任意角色生成3D语音动画的新方法。它通过微调预训练的视频扩散模型来适应目标角色，然后通过优化过程将生成的说话头部视频提升为3D语音动画，并进一步蒸馏为实时网络AnyTalk_RT。

Details

Motivation: 现有音频驱动的3D语音动画方法依赖于角色特定的训练数据或繁琐的绑定/重网格化，AnyTalk旨在克服这些限制，利用大规模视频数据集训练的视频扩散模型，为任意角色生成语音动画，减少手动工作和数据需求。

Result: 论文未在摘要中提供具体的定量结果或基准测试比较，但声称方法能够为多样化的面部网格和混合形状配置生成口型同步的动画，并实现实时性能。

Insight: 创新点包括：1) 通过角色特定微调（CsF）技术，利用零音频嵌入（代表“无运动”）在渲染图像上微调视频扩散模型，无需动画数据即可保留大规模视频扩散模型的运动先验；2) 通过优化过程估计混合形状参数，将说话头部视频提升为3D动画；3) 将方法蒸馏为实时网络，提升实用性。这为任意角色的音频驱动语音动画技术提供了更广泛的访问途径。

Abstract: We present AnyTalk, a novel method for generating 3D speech animations for arbitrary characters without requiring any animation data. While existing audio-driven 3D speech animation methods rely on character-specific training data or laborious rigging/re-meshing, AnyTalk circumvents these limitations by leveraging recent video diffusion models trained on extensive video datasets. We first adapt a pre-trained video diffusion model to a target character through our Character-specific Fine-tuning (\textit{CsF}) technique. By fine-tuning on rendered images of the 3D character paired with zeroed-out audio embeddings (representing “no motion”), we eliminate the need for animation data while preserving the motion prior of large-scale video diffusion model. We then uplift the resulting talking-head video into a 3D speech animation by estimating blendshape parameters through a proposed optimization process. AnyTalk enables lip-synced animations across diverse face meshes and blendshape configurations, significantly reducing manual effort and data requirements. We further enhance usability by distilling AnyTalk into a streamlined network, $\text{AnyTalk}_{RT}$, thereby enabling real-time performance. By leveraging talking-head video generation, our method broadens access to audio-driven speech animation technology for arbitrary characters. The code is publicly available at https://serin-yoon.github.io/projects/anytalk/.

cs.AI [Back]

[189] Trust Is Not Enough: Influence Calibration for On-Policy Self-Distillation in Agentic RL cs.AI | cs.CLPDF

Qizhen Lan, Xi Xiao, Xiangchen Guan, Mengchen Fan, Moule Lin

TL;DR: 本文提出了一种名为影响校准自蒸馏（ICSD）的新方法，用于解决策略自蒸馏（OPSD）中教师信任分配与策略目标不匹配的问题。ICSD通过测量每个监督标记对策略目标的一阶响应，动态调整蒸馏损失的权重，从而更有效地利用教师提供的密集监督。

Details

Motivation: 现有基于策略的自蒸馏方法主要依赖教师信任来分配监督，但信任并不能反映强调某个标记是否支持当前策略目标，存在信任-效用不匹配的问题。

Result: 在ALFWorld、WebShop和Search-QA三个基准测试中，ICSD在GRPO和GiGPO优化器下，对1.5B到7B参数规模的模型均提升了所有匹配的聚合指标。在7B模型上，ALFWorld成功率达到了96.1%，WebShop得分达到93.1。分析表明，ICSD将分配给与目标相反标记的教师支持质量从60.1%降低到37.8%，并将与RL梯度的余弦兼容性提高了0.192。

Insight: 核心创新在于引入了影响校准机制，通过一阶响应分析将非平稳信号转化为有界的分配权重，从而更精确地将教师监督与强化学习目标对齐。该方法仅影响蒸馏损失，无需额外的模型前向传播，计算高效。

Abstract: On-policy self-distillation (OPSD) gives language agents dense token-level supervision from a privileged self-teacher on the policy’s own trajectories. Existing methods allocate this supervision mainly by teacher trust, but trust does not reveal whether emphasizing a token supports the current policy objective. We call this the trust-utility mismatch and introduce Influence Calibration for Self-Distillation (ICSD). For each supervised token, ICSD measures the first-order response of its importance-weighted RL surrogate contribution to a teacher-directed output perturbation. Batch-adaptive calibration converts this non-stationary signal into a bounded allocation weight while preserving the original auxiliary-loss mass within each action turn. These detached weights affect only the distillation loss and require no additional model pass. Across ALFWorld, WebShop, and Search-QA, ICSD improves all matched aggregate metrics over trust-only allocation under Group Relative Policy Optimization (GRPO) and Group-in-Group Policy Optimization (GiGPO), across two model families spanning 1.5B to 7B. At 7B, it reaches 96.1% ALFWorld success and a WebShop score of 93.1. Frozen-batch analyses show that ICSD reduces teacher-supported mass assigned to objective-opposed tokens from 60.1% to 37.8% and raises cosine compatibility with the RL gradient by 0.192. A companion repository is avail- able at https://github.com/lanqz7766/Influence-Calibration-for-On-Policy-Self-Distillation-in-Agentic-RL.

[190] LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning Tasks cs.AI | cs.CL | cs.LGPDF

Chih-Hsuan Yang, Jingyan Jiang, Cheng-Hau Yang, Vikram Vasudevan, Huihuo Zheng

TL;DR: 该论文研究了在多智能体大语言模型系统中，如何根据计算成本决定是否采用额外协作协议来提升推理性能。作者在多个基准测试上比较了四种协议（直接求解、迭代自校正、规划-执行-评审协作、多智能体审议）的性能，并评估了固定策略、训练路由器和冻结LLM路由器的效果。研究发现，保守策略倾向于协作不足，而高求解率的路由器则容易过度协作；基于置信度的预协作门控机制在效率和性能上取得了平衡，但针对特定协议的成本感知路由问题仍未解决。

Details

Motivation: 部署多智能体LLM系统时，需要权衡额外协作带来的性能提升与其计算成本，但目前缺乏有效的方法来决定何时采用何种协作协议是值得的。

Result: 在包含4,181个竞赛级数学问题的主要基准上，预协作的gpt-oss-120b探针预测基线失败取得了0.8847 AUROC，但对预测特定协议（如PER或Broadcast）的价值较弱（AUPRC分别为0.1674和0.1041）。基于置信度的门控机制在45K tokens下达到78.0%的求解率，优于冻结路由器的73.8%（71.3K tokens），但低于回顾性固定顺序预言机的92.4%。预言机在10个配对模型-条件设置中比基线提升了23.2-58.3个百分点的回顾性覆盖率。

Insight: 论文的创新点在于系统性地量化了不同协作协议的成本-收益权衡，并揭示了基于置信度的门控机制可作为有效的初始协作决策工具。然而，研究也指出，针对特定协议进行精细化的成本感知路由仍然是一个未解决的挑战，这为未来研究指明了方向。

Abstract: Multi-agent large language model (LLM) systems can improve reasoning by spending more computation, but deployment requires deciding when extra collaboration is worth its cost. We isolate this decision by running every problem under four protocols while holding the solver fixed within each setting: direct solving (Baseline), iterative self-correction (Single), planner-executor-reviewer collaboration (PER), and multi-agent deliberation (Broadcast). The primary benchmark comprises 4,181 competition-level math problems; paired robustness checks cover four benchmarks spanning competition math, biology, and broader science with two solver families. Across fixed policies, trained routers, and frozen LLM routers, conservative policies under-escalate, whereas higher-solve frozen routers often over-escalate. A post-answer, pre-collaboration gpt-oss-120b probe ranks Baseline failures with 0.8847 AUROC (4,151 parseable cases; 95% CI [0.8732, 0.8955]). The same score remains informative for predicting whether any collaboration helps (0.7683 AUPRC), but is much weaker for identifying PER- or Broadcast-specific value (0.1674 and 0.1041 AUPRC). Separately, the pre-answer self-confidence gate reaches 78.0% solve at 45K tokens, compared with 73.8% at 71.3K for a frozen gpt-oss-120b router and 92.4% for a retrospective fixed-order oracle. Across 10 paired model-condition settings, the oracle adds 23.2-58.3 points of retrospective coverage over Baseline, but protocol profiles vary by task. In the six settings with held-out router evaluations, oracle gaps remain 18.5-28.9 points. Confidence can therefore support initial escalation, while protocol-specific cost-aware routing remains unresolved.

[191] KV-Rescue: Recovering Reasoning Language Model KV Eviction Loss via Stepwise Interleaving cs.AI | cs.CLPDF

Minsoo Cheong, Woosang Lim, Vincent-Daniel Yun, Sungjoo Yoo

TL;DR: 本文提出KV-Rescue，一种无需训练、基于交错推理的推理框架，旨在解决大语言模型在长推理任务中因KV缓存驱逐导致的信息丢失和性能退化问题。该方法通过结合一个轻量级全上下文助手模型，在线检测并终止基础模型生成的不连贯或重复内容，从而恢复因KV驱逐损失的模型精度。

Details

Motivation: KV缓存驱逐虽然能控制长推理轨迹的内存开销，但会导致信息丢失，在激进的内存预算下不仅降低准确性，还可能引发模型生成不连贯或重复令牌的失控退化现象。作者认为这种损失主要是由缺失上下文造成的信息差距，而非模型能力不足。

Result: 在五个数学基准测试（使用Qwen2.5-Math 7B和72B模型）上，KV-Rescue在驱逐预算B=64时平均恢复了因驱逐损失的87%的准确率。解码成本分析进一步表明，防止失控退化使基础模型的令牌生成平均减少了43%。

Insight: 核心创新在于将KV驱逐损失归因于信息差距而非能力差距，并利用轻量级全上下文助手模型与基础模型进行交错推理来弥补这一差距。通过在线检测（基于熵和可压缩性）提前终止低质量生成，有效防止了退化，实现了无需训练的高效推理恢复。

Abstract: KV-cache eviction caps the memory cost of long reasoning traces but is inherently lossy because the model decodes from a partial view of its history. Under aggressive budgets, this not only lowers accuracy but can also cause runaway degeneration, where the model produces incoherent or repetitive tokens until reaching the length limit. We characterize much of this loss as an information gapf caused by missing context, rather than a capability gap caused by limited model capacity. An evicted 7B model and a full-context 1.5B model make complementary errors, and an oracle choice between their answers recovers 79% of the accuracy gap to the full-KV 7B model. Based on this observation, we propose KV-Rescue, a training-free inference framework that bridges the information gap introduced by KV eviction using a lightweight full-context helper. KV-Rescue interleaves reasoning steps from the two models into a shared trajectory. An online detector uses entropy and compressibility to terminate the generation of incoherent or repetitive base-model candidates early. Across five math benchmarks with Qwen2.5-Math 7B and 72B, KV-Rescue recovers an average of 87% of the accuracy lost to eviction at eviction budget B=64. A decode-cost analysis further shows that preventing runaway degeneration cuts base-model token generation by 43% on average.

[192] Schema-Agnostic Graph Reasoning Agent for Hybrid Knowledge Graphs cs.AI | cs.CL | cs.DBPDF

Marius Dragic, Ruben Ifrah, Alexandre Rio

TL;DR: 本文提出了一种图推理智能体（GRA），它通过七种通用工具探索混合知识图谱（包含文本概念和关系表节点），在运行时动态发现领域特定信息。在工业基准UFK-M上，GRA以更少的输入令牌（少于三分之一）实现了比全上下文智能体更高的准确率（88.4% vs. 83.3%）。研究表明，性能提升主要源于智能体对结构化底层的选择性访问，而非图谱拓扑本身。

Details

Motivation: 受代码库探索工具（如ls、cat、grep）的启发，作者认为知识图谱可以支持类似的通用操作（如列出邻居、读取节点内容、搜索描述），从而构建一个能够灵活探索混合知识图谱的智能体，解决在复杂结构化数据中高效获取答案的问题。

Result: 在UFK-M工业基准（包含258个分析性问题，其标准答案通过执行已验证的SQL程序生成）上，GRA的准确率达到88.4%，比全上下文智能体高出5.1个百分点，同时输入令牌数不到后者的三分之一。消融实验表明，性能增益主要来自智能体的选择性访问能力，且依赖于模型可靠驱动工具的能力。

Insight: 创新点在于将知识图谱视为与代码库类似的可操作接口，通过少量通用工具实现模式无关的图探索；核心洞见是‘所见即所得’——在结构化底层上进行选择性导航比提供详尽上下文更有效，这为构建高效知识推理智能体提供了新思路。

Abstract: Tool-calling LLM agents navigate unfamiliar codebases with a handful of generic primitives for listing, reading and searching files (ls, cat, grep). A knowledge graph admits the same interface: listing neighbours, reading node content and searching descriptions are the same operations on a different substrate. Building on this correspondence, we present GRA, a Graph Reasoning Agent that explores hybrid knowledge graphs, whose nodes are either textual concepts or relational tables, with seven generic tools, discovering everything domain-specific at run time. On UFK-M (Unified Factory Knowledge Model), an industrial benchmark of 258 analytical questions whose gold answers are produced by executing validated SQL programs, GRA beats a full-context agent by 5.1 pp (88.4% vs. 83.3%), while reading under a third of its input tokens. A graph-free control shows the gain comes chiefly from selective agentic access rather than graph topology, and that the effect depends on a model able to drive tools reliably. Seeing less, the agent answers better: selective navigation over a structured substrate beats exhaustive context.

[193] Prior Audit-Repair Context Shifts LLM Verifier Thresholds Toward Leniency cs.AI | cs.CLPDF

Parsa Mazaheri, Kasra Mazaheri

TL;DR: 该论文研究了在自动化检查流程中，当语言模型（LLM）作为检查器时，其上下文是否包含先前的审计-修复记录会影响其检查报告的严格程度。研究发现，存在审计-修复记录会显著降低检查器的误报率，并且这种影响主要源于检查器判断阈值（criterion）的调整，而非其辨别能力（d’）的变化。

Details

Motivation: 动机是探究在自动化检查与修复的流水线中，将语言模型同时或分别用作检查器和修复器时，检查器是否因其上下文包含审计和修复记录而改变其报告行为，即解决这种系统架构是否引入检查偏差的问题。

Result: 在ProcessBench人工验证正确的数据上进行测试，发现存在审计-修复记录使所有15个模型与措辞组合的误报率均降低（降低2.8至11.5个百分点，相对减少9%至25%）。信号检测分析表明，这种变化源于判断阈值的调整（15/15组合），而非辨别能力的变化。

Insight: 创新点在于揭示了LLM检查器的判断阈值会因上下文中的审计-修复记录而向更宽松（lenient）的方向系统性偏移，且这种偏移主要由审计结论和修复内容共同驱动，而非模型辨别能力的改变。这为理解和设计基于LLM的自动化检查系统提供了重要洞察。

Abstract: Automated checking pipelines increasingly place one language model as the checker and another (or the same one) as the fixer. We ask whether that wiring changes what the checker reports. Measuring false alarms on human-verified-correct ProcessBench traces with the present task held byte-identical, we find that a completed audit -> repair episode already in the model’s context lowers false alarms in 15 of 15 model x wording combinations, by 2.8 to 11.5 percentage points against a length-matched non-audit control, a 9 to 25% reduction relative to that control. The direction contradicts what the accumulated-message literature predicts: an episode whose audit reported an error lowers false alarms further still, at all five wordings on the model where that manipulation lands cleanly, though a negativity asymmetry predicts more flagging. Decomposing the episode finds repair content and audit verdict complementary: different components carry the effect on different model families. Signal-detection analysis locates the change in the threshold rather than in discrimination – the criterion moves in 15 of 15 combinations and survives correction in 13 while d’ survives in none, though the d’ test is half as sensitive by construction – and a hand audit of 50 false alarms finds 82% simply wrong, so at this operating point the shift need not be harmful. With reasoning enabled the effect keeps its relative size on both models tested, and the threshold reading holds there too.

[194] Policy Iteration with Human Feedback: Bringing Post-Training RL to In-context Learning cs.AI | cs.CLPDF

Minh-Ha Nguyen, Cathy Shyr

TL;DR: 这篇论文提出了基于人类反馈的策略迭代方法，将后训练强化学习引入上下文学习。该方法利用预训练语言模型作为执行基础，通过语言模型批评家和临床专家审查来定位错误并形成策略修订，最终在罕见疾病诊断任务中提升了多个执行器的召回率。

Details

Motivation: 动机是利用预训练语言模型的可重用任务表示能力，结合上下文学习和人类反馈，解决在罕见疾病诊断等复杂任务中，固定模型如何通过专家指导持续改进策略的问题。

Result: 在累积消融实验和超罕见疾病基准测试中，该方法显著提升了多个执行器的Recall@1指标，例如GPT-5.4提升了32.7个百分点，Qwen3.6-35B提升了31.1个百分点，验证了其在罕见疾病诊断中的有效性。

Insight: 创新点在于将广义策略迭代的评估-改进循环与人类反馈相结合，将策略修订过程版本化并置于自然语言层面，同时保持专家对修订的最终权威，为在固定权重语言模型上进行专家引导的策略开发提供了可行框架。

Abstract: Generative pretraining established reusable task representations; later work on language-based task conditioning and in-context learning showed that a fixed model could adapt its behavior from instructions and demonstrations. Policy Iteration with Human Feedback (PIHF) builds on this development and the recurrent evaluate-and-improve structure of generalized policy iteration. PIHF uses a pretrained language model as its execution substrate and moves persistent revision to a versioned natural-language policy and tool set. A language-model critic and clinical expert review complete-panel reasoning and tool-use trajectories to localize recurrent failures and form candidate revisions; the expert may reinterpret the evidence and retains authority over admission and rollback, while Recall@1 and Recall@5 validate outcomes after candidate execution. Across cumulative ablations and ultra-rare-disease benchmarks, a PIHF-derived policy improved Recall@1 in one proprietary executor and three open-weight executors spanning 3 to 49 billion active parameters. Gains were 32.7 percentage points for GPT-5.4 and 31.1 points for Qwen3.6-35B, a difference of 1.7 points. These results support the feasibility of using pretrained language models as fixed-weight execution substrates for expert-guided policy development in rare-disease diagnosis.

[195] The Unwritten Benchmark: A New Challenge for Multimodal Machine Learning in Abstract Perceptual Reasoning cs.AI | cs.CVPDF

Garima Arya Yadav, Nilay Yilmaz, Yezhou Yang

TL;DR: 本文提出了一个名为’The Unwritten Benchmark’的新基准测试，旨在评估多模态模型在抽象感知推理方面的能力。核心任务是通过观察书写时的手部运动视频和笔尖摩擦音频，推断出正在书写的单词，而没有任何可见的墨迹。实验发现，人类在此任务上表现出色，而当前领先的多模态模型（如GPT-4o和Gemini 2.5-Pro）表现极差，且存在多模态信息融合反而损害性能的悖论。

Details

Motivation: 当前多模态模型在静态视听内容识别上表现出色，但其从动态生成过程中推断未见信息的抽象感知推理能力仍是一个关键且未被充分探索的前沿领域。本文旨在填补这一空白，探究模型在跨模态因果推理和对微观运动学理解方面的根本局限性。

Result: 在提出的基准测试上，人类参与者获得了超过80%的有序字母准确率，而领先的多模态模型（GPT-4o和Gemini 2.5-Pro）的准确率均未超过10%。研究还发现了一个悖论性的融合效应：为模型同时提供音频和视频模态信息，其性能反而比使用单一模态更差。

Insight: 论文的创新点在于提出了一个新颖的、专注于抽象感知推理的基准测试任务（声学-运动学单词推断），该任务挑战了模型整合互补感知线索进行认知推理的能力。客观来看，该研究揭示了当前多模态模型在跨模态因果推理和微观运动学理解方面存在根本性缺陷，其多模态融合机制在复杂认知任务上可能失效，这为未来模型设计提供了关键的诊断方向。

Abstract: Current multimodal models have demonstrated remarkable proficiency in recognizing static visual and auditory content. However, their capacity for abstract perceptual reasoning, inferring unseen information from dynamic, generative processes, remains a critical and underexplored frontier. In this paper, we introduce The Unwritten Benchmark, a new challenge designed to probe this abstract perceptual and cognitive ability. We define the core task as acousto-kinematic word inference: models must decipher words, across 3 different writing styles, being written solely from the audio of pen scratches and the video of hand movements, without any visible ink trace. Our evaluation results reveal a profound gap between human and machine performance: while human participants achieve high ordered letter accuracy (over 80%), leading Multimodal Machine Learning Models, including GPT-4o and Gemini 2.5-Pro, struggle significantly, failing to surpass 10%. Furthermore, we identify a paradoxical fusion effect in the models, where providing both modalities often degrades performance rather than improving it. This finding indicates a fundamental breakdown in their ability to synthesize complementary perceptual cues for this cognitive task. These findings highlight significant limitations in both cross-modal causal reasoning and the understanding of the micro-kinematics essential for such cognitive and intuitive perceptual reasoning.

[196] From Generalist to Specialist: A Context-Fusion Framework for Endoscopic Polyp Reporting with a Frozen VLM cs.AI | cs.CVPDF

Ruijie Yang, Yan Zhu, Peiyao Fu, Siyuan Li, Te Luo

TL;DR: 本文提出了一种名为’上下文融合框架’的方法，用于将通用的视觉语言模型（VLM）专业化，以生成包含病灶尺寸、巴黎分类和形态描述在内的内窥镜息肉报告。该方法通过融合隐式指令上下文和显式转导上下文来实现，无需修改预训练模型的权重。

Details

Motivation: 现有方法在将通用VLM专业化用于内窥镜息肉报告时，通常需要训练特定任务模型或调整模型权重，这破坏了模型的统一接口并可能损害其预训练能力。本文旨在解决如何在保持VLM统一接口和预训练能力的同时，为其注入可靠的专家知识。

Result: 在包含2056张专家标注内窥镜图像的公开数据集上进行实验。所提框架在数值、分类和报告生成指标上均显著优于直接使用冻结VLM的推理结果，并在所有评估方法中取得了最强的综合性能。其新增可训练参数量仅为冻结VLM总参数的0.006%。当检索到的top-1案例包含正确目标类别时，该框架能纠正基线方法70.5%的错误。

Insight: 核心创新点在于通过’上下文融合’策略实现模型专业化：利用自监督息肉编码器检索相关图像-报告对作为显式证据，同时学习连续的专家令牌作为跨病例共享的隐式指令上下文。这是一种轻量级且有效的冻结VLM专业化策略，平衡了知识注入与能力保留。

Abstract: Reliable endoscopic polyp reporting requires integrating quantitative lesion sizing, standardized Paris classification, and clinically meaningful morphological description within a single record. General-purpose vision-language models (VLMs) offer a unified interface for image understanding and report generation. Existing specialization strategies, however, typically rely on task-specific models or model-weight adaptation, leaving unresolved how to introduce reliable specialist knowledge while preserving both this unified interface and the VLM’s pretrained capabilities. We introduce a context-fusion framework that specializes a frozen general-purpose VLM through both implicit instruction context and explicit transduction context without modifying its pretrained weights. Specifically, a self-supervised polyp encoder retrieves related image-report pairs as explicit, query-specific evidence, while learned continuous specialist tokens provide implicit instruction context shared across cases. Experiments were conducted on 2,056 expert-annotated public endoscopic images. We compared the framework with general-purpose VLMs, task-specific predictors, and weight-adaptation methods to assess specialist performance, unified reporting, and adaptation efficiency. Across numerical, categorical, and report-generation metrics, the proposed framework substantially improved direct frozen-VLM inference and achieved the strongest overall performance among the evaluated methods. It added trainable parameters equal to only 0.006% of the frozen VLM’s parameter count. When the top-1 retrieved case carried the correct target category, our framework corrected 70.5% of the errors made by a weight-adaptation baseline. These findings support the context-fusion framework as a lightweight and effective strategy for specialist adaptation of a frozen VLM.

[197] UI-Mate: Advancing Open-Weight Foundation GUI Agents with In-Context Demonstrations cs.AI | cs.CVPDF

Zihan Ding, Longxu Dou, Qi Gao, Xiangwu Guo, Shengchao Hu

TL;DR: 本文提出了UI-Mate，一个开源的GUI基础智能体，旨在通过环境驱动的训练栈和上下文演示学习来解决GUI自动化任务中数据稀缺、提示模糊和执行不可靠的问题。

Details

Motivation: 现有GUI基础智能体的部署受到训练数据稀缺且有偏、用户指令模糊以及执行过程不可靠的阻碍，特别是涉及用户特定工具和隐性惯例的常规工作流，未明确说明的指令会导致执行结果出现任意变化。

Result: UI-Mate-27B在通用计算机使用基准测试中取得了新的开源权重SOTA，在OSWorld-Verified上得分为77.0%，在WindowsAgentArena上为66.2%。在OSWorkerBench基准测试（包含100个长视野办公任务）上，其严格成功率为41.0%，进度为76.9%，显著优于其基础模型Qwen3.6-27B。在33个任务的自我演示子集上，一个演示将严格成功率从17.2%提升至35.4%。

Insight: 主要创新点包括：1）一个可扩展的环境驱动训练栈，通过闭环数据引擎自动化任务生成、环境构建、策略执行、过滤、能力平衡、监督微调和在线强化学习；2）上下文演示学习机制，将多模态演示转化为灵活的子任务级工作流，并能够从实时界面重新规划；3）引入了OSWorkerBench基准测试，支持仅指令和演示引导的评估，并区分了自我演示和变体演示两种设置，为评估长视野任务可靠性提供了新视角。

Abstract: Foundation GUI agents can automate complex digital tasks, but deployment is hindered by scarce and biased training data, ambiguous prompts, and unreliable execution. Routine workflows rely on user-specific tools and tacit conventions, so unstated instructions can produce arbitrary variations across runs. We present UI-Mate, a foundation GUI agent that integrates an environment-grounded training stack with in-context demonstration learning. UI-Mate makes three contributions: A Scalable Environment-Grounded Training Stack: A closed-loop data engine automates task generation, environment construction, rollout, filtering, capability balancing, SFT, and online RL across massively parallel environments via unified task-verifier bundles. In-Context Demonstration Learning: A mechanism that transforms multimodal demonstrations into flexible subtask-level workflows, follows relevant demonstrated steps, and re-plans from the live interface. OSWorkerBench Benchmark and Insights: A benchmark of 100 long-horizon office tasks across 41 applications that supports instruction-only and demonstration-guided evaluation. Its demonstration resources separate a 33-task self-demo setting, built from successful strong-agent rollouts of the same targets, from a 45-task variant-demo setting, built from human recordings of related but non-identical tasks. Experiments show that UI-Mate-27B sets a new open-weight state of the art on general computer-use benchmarks, scoring 77.0% on OSWorld-Verified and 66.2% on WindowsAgentArena. On OSWorkerBench, it reaches 41.0% strict success and 76.9% progress, outperforming its Qwen3.6-27B base by 17.7 and 24.5 points. On the 33-task self-demo subset, one demonstration raises strict success from 17.2% to 35.4% and progress from 67.9% to 81.1%, substantially improving long-horizon reliability. Project page: https://ui-mate.github.io.

[198] DriveCache: Action-Aware Caching for Driving World Model Inference cs.AI | cs.CVPDF

Jianchun Yang, Jian Liang, Xianda Guo, Pinhan Fu, Yanlun Peng

TL;DR: DriveCache是一种面向驾驶世界模型推理的动作感知缓存方法，旨在加速基于扩散模型的驾驶视频生成。它利用生成前可用的驾驶信号（如自车速度和规划轨迹），通过动态规划在去噪步骤间分配特征重用，并在生成偏离校准时进行特征刷新和重新规划。

Details

Motivation: 基于扩散的驾驶视频生成模型在去噪步骤中反复评估大型骨干网络，限制了生成吞吐量。现有扩散加速方法未充分利用生成前可用的驾驶信号，如自车运动和规划轨迹。

Result: 在三种生成器配置上，DriveCache相比评估的缓存方法，在整体保真度-效率权衡方面表现更优。实验表明缓存容忍度随自车平移/旋转、去噪进度和连续重用长度而变化。

Insight: 创新点在于提出无训练的动作感知控制器，利用规划运动跨场景分配重用，并通过动态规划在去噪步骤间放置缓存，同时引入因果漂移检查机制来刷新特征和重新规划剩余调度。

Abstract: Driving video generation models support autonomous-driving development by predicting controllable future scenes for simulation, planning evaluation, and offline data generation. Diffusion-based driving generators repeatedly evaluate large backbones across denoising steps, which limits generation throughput. Existing diffusion acceleration methods reduce this cost, but general-purpose designs omit driving signals available before generation, such as ego speed and planned trajectories. Experiments across driving motions show that cache tolerance varies with ego translation and rotation, denoising progress, and consecutive reuse length. We propose DriveCache, a training-free, action-aware controller that uses planned motion to allocate reuse across scenes and dynamic programming to place it across denoising steps under a calibrated response budget. A causal drift check refreshes features and replans the remaining schedule when generation departs from calibration. Across three generator configurations, DriveCache improves the overall fidelity-efficiency trade-off over evaluated cache methods. Our code will be publicly available.

cs.IR [Back]

[199] Ask to Be Sure: Informative Interactions for Confident Multi-Turn LLM Recommendation cs.IR | cs.AI | cs.CL | cs.LGPDF

Cedar Site Bai, Duanshun Li, Zhenyu Liao, Sheikh Sarwar, Huiyuan Chen

TL;DR: 本文提出了一种基于熵减奖励的对话式推荐系统优化方法，通过量化每次交互对推荐不确定性的降低程度来指导大语言模型生成更具信息量的多轮对话，从而提升推荐准确性和对话效率。

Details

Motivation: 现有对话推荐系统在引导多轮交互以有效获取用户偏好方面存在挑战，要么依赖模板化交互的强化学习代理，要么以另一个LLM的交互性判断为优化目标，未能衡量实际获取的有用信息量。

Result: 在INSPIRED和ReDial数据集上，通过监督微调（SFT）和直接偏好优化（DPO）的实验表明，该方法同时提高了推荐质量和对话效率。

Insight: 创新点在于提出以推荐熵减作为奖励信号来量化交互信息增益，无需依赖真实推荐数据即可优化模型生成策略性交互，为基于不确定性的对话策略学习提供了新思路。

Abstract: Recent advances in large language models (LLMs) have enabled their use as conversational recommender systems (CRS), demonstrating strong recommendation accuracy and natural dialogue. However, guiding multi-turn interactions to elicit user preferences effectively remains challenging. Existing approaches either use separate reinforcement learning agents with templated interactions or optimize for interactivity judged by another LLM, without measuring how much useful information is actually gained. We propose a new approach that quantifies the effectiveness of each interaction by the reduction in the assistant’s uncertainty, measured via entropy over recommendations. We apply this entropy reduction as a reward—without relying on ground-truth recommendations, which are often unavailable in real-world scenarios—to fine-tune the LLM, enabling strategic interaction generation. Empirical results with supervised fine-tuning (SFT) and direct preference optimization (DPO) on the INSPIRED and ReDial datasets show that our method improves both recommendation quality and conversational efficiency.

cs.SD [Back]

[200] Listen, Reason, and Segment: Aligning LALMs with Editorial Judgment for Media Chapterization cs.SD | cs.AI | cs.CL | eess.ASPDF

Tony Alex, Wish Suharitdamrong, Sara Atito, Armin Mustafa, Muhammad Awais

TL;DR: 本文提出了AudioChaps框架，通过思维链推理引导的组相对策略优化（GRPO）来对齐端到端的大型音频语言模型（LALMs），以解决音频章节化任务——将连续音频流分割成主题连贯的章节。该方法在无需监督微调冷启动的情况下显著提升了性能，并发布了三个配套数据集用于训练和评估。

Details

Motivation: 尽管大型音频语言模型在标准化基准上进展迅速，但在实际媒体工作流（如内容管理、归档索引和分发）中的应用仍不足。音频章节化任务因边界定义依赖于主观编辑判断而非客观声学事件，对模型的长上下文顺序推理能力提出了高要求，这暴露了现有模型的差距。

Result: 在无需监督微调冷启动下，AudioChaps-R1-Zero相比最先进的LALM Audio-Flamingo-3-Think将平均F1分数提升了33个百分点；最终模型AudioChaps-R1提升了49个百分点。评估在自建的AudioChaps-Eval基准上进行，达到了新的SOTA水平。

Insight: 创新点包括：1）将GRPO直接应用于LALMs对齐，无需SFT冷启动，简化了训练流程；2）利用思维链推理提供结构化监督，提升模型对主观编辑判断的近似能力；3）构建了专门针对音频章节化的高质量数据集，支持模型训练与评估。从客观角度看，该工作将强化学习对齐方法有效应用于音频理解任务，推动了LALMs在实际媒体处理中的落地。

Abstract: Large Audio Language Models (LALMs) have made rapid progress on standardized benchmarks, yet their deployment in practical media workflows, curation, archival indexing, and content distribution remains largely unrealized. We identify automated audio chapterization, the task of segmenting continuous audio streams into thematically coherent chapters, as a demanding and commercially consequential setting that exposes this gap. Chapterization is challenging because boundaries are defined less by objective acoustic events than by subjective editorial judgment, requiring models to reason sequentially over long acoustic contexts and approximate creator-authored boundary decisions. We present AudioChaps, a post-training framework for aligning end-to-end LALMs for this task via Group Relative Policy Optimization (GRPO) guided by Chain-of-Thought (CoT) reasoning. To support training and evaluation, we curate three datasets: AudioChaps-Alignment, derived from creator-annotated chapter boundaries on YouTube; AudioChaps-CoT, which provides structured supervision for well-formatted, high-quality, and evidence-grounded boundary reasoning; and AudioChaps-Eval, a held-out benchmark for audio chapterization. Applying GRPO directly without a Supervised Fine-Tuning (SFT) cold start, AudioChaps-R1-Zero already improves average F1 by 33 points over the state-of-the-art LALM Audio-Flamingo-3-Think. The AudioChaps framework produces our final aligned LALM, AudioChaps-R1, which improves average F1 by 49 points. These results demonstrate that GRPO-trained LALMs can reliably transform unstructured auditory streams into navigable, structured media. Our code, models, and dataset resources will be released upon acceptance at https://github.com/ta012/AudioChaps.

[201] SingDance: Compositional Zero-Shot Singing-and-Dancing Video Generation with Role-Aware Audio Conditioning cs.SD | cs.CVPDF

Tao Feng, Xu Li, Xiangyang Luo, Ming Wen, Huadai Liu

TL;DR: SingDance是一个统一的视频扩散框架，用于生成角色感知的歌唱与舞蹈视频。它通过将可控的声乐表达（如口型同步）定义为语义角色（源角色或听者角色），并结合硬紧凑路由与帧级联合音频注入技术，实现了从未在训练中观察过的歌曲/源角色配置下的组合式零样本歌唱舞蹈视频生成。

Details

Motivation: 现有方法主要关注音乐驱动的编舞或假设可见主体产生输入语音的语音驱动模型，而歌唱舞蹈视频生成需要同时满足音乐驱动的身体运动和可见主体的声乐表达（口型同步），这一组合场景尚未得到充分探索。

Result: 实验表明，该方法在运动-节拍对齐和视觉保真度方面表现强劲，能够在保持音乐对齐的身体运动的同时可靠地切换声乐表达角色，并且在口型同步方面与最强的语音驱动基线相比具有高度竞争力，同时生成时参数量显著减少。

Insight: 核心创新在于将声乐表达建模为语义角色（源/听者），并通过角色感知的音频条件化与组合式训练策略（使用非对称监督数据）实现了歌唱与舞蹈能力的解耦与组合，从而支持零样本生成。这为多模态条件视频生成中的角色与行为解耦提供了新思路。

Abstract: Generating personalized dance videos from a reference image, text prompt, and audio track requires music-conditioned body motion. Singing-and-dancing adds a second requirement: the visible subject must also articulate the vocals. Existing music-conditioned methods focus primarily on choreography, while speech-driven models generally assume that the visible subject produces the input voice, leaving this combined setting largely underexplored. We introduce SingDance, a unified video diffusion framework that formulates controllable vocal articulation as a semantic role: the visible subject is either the source, who produces the vocal signal, or the listener, who receives it from an off-screen performer. Hard-compact routing selects task-relevant speech, music, and role conditions, which are composed through frame-wise joint audio injection; source and listener retain the same speech pathway. Training uses asymmetric supervision: on-screen speaking and curated off-screen conversational-response videos establish role control, while instrumental and song-based dancing-only videos establish music-conditioned body motion. The target Song/Source configuration is never observed during training. At inference, assigning the source role to a song composes separately learned articulation and song-conditioned dance capabilities, enabling compositional zero-shot singing-and-dancing. Experiments demonstrate strong motion–beat alignment and visual fidelity, reliable paired switching of vocal articulation while preserving music-aligned body motion, and highly competitive lip synchronization with substantially fewer generation-time parameters than the strongest speech-driven baseline evaluated.

eess.IV [Back]

[202] KHiM-Mamba: Injecting Pathology Knowledge into Mamba via Hidden-State Modulation for Whole Slide Image Analysis eess.IV | cs.CV | q-bio.QMPDF

Qixiang Zhang, Yi Li, Tianqi Xiang, Haonan Wang, Mengjiao Wei

TL;DR: 本文提出KHiM-Mamba，一种用于全切片图像分析的知识感知隐藏状态调制架构。它通过将病理学知识先验注入Mamba模型的核心选择性状态空间机制，引导切片编码动态过程向诊断相关的证据积累方向发展，从而解决大规模WSI中视觉驱动选择机制可能误判关键区域的问题。

Details

Motivation: 现有基于选择性状态空间模型的MIL方法仅依赖视觉特征，在大规模全切片图像中，稀疏的诊断关键区域被大量无关信息包围，这种纯视觉驱动的选择动态可能导致状态更新和读取分配错误，使演化中的SSM状态积累任务无关证据并稀释关键诊断线索。

Result: 在涵盖4个任务的11个公共基准测试上的实验表明，KHiM-Mamba始终实现了最先进的性能。

Insight: 核心创新点在于重新设计了原始SSM层，在隐藏状态演化过程中执行知识调制操作，引导每个编码步骤从隐藏状态积累和检索何种视觉证据；此外，引入了一个局部自适应词汇检索模块，利用大语言模型为每个图像块分配细粒度、组织特异性的语义描述，实现跨不同任务的精确调制。

Abstract: Whole slide image analysis is commonly formulated as multiple instance learning (MIL), where instance features are contextually updated and aggregated into a slide representation, a process we term slide encoding dynamics. Recently, selective state-space models (SSM) have emerged as promising MIL architectures due to their long-sequence modeling capability and linear complexity. However, existing SSM-based MIL methods rely solely on visual features during MIL. Meanwhile, in large-scale WSIs, where sparse diagnostically decisive regions are surrounded by abundant irrelevant information, such purely vision-driven selective dynamics can misallocate state updates and readouts, causing the evolving SSM state to accumulate task-irrelevant evidence and dilute critical diagnostic cues over long scan trajectories. In this work, we propose the Knowledge-Aware Hidden-State Modulation architecture (KHiM-Mamba), which innovatively regulates Mamba’s core selective state-space mechanism with explicit knowledge priors, steering slide encoding dynamics toward diagnostically meaningful evidence accumulation. Specifically, we redesign the original SSM layer to perform knowledge modulation operations during the evolution of hidden states, thereby guiding what visual evidence is accumulated and retrieved from the hidden state at each encoding step. Furthermore, we additionally introduce a local-adaptive vocabulary retrieval module that uses large language models to assign each patch fine-grained, tissue-specific semantic descriptions, enabling precise modulation across diverse tasks. Experiments on 11 public benchmarks across 4 tasks show that KHiM-Mamba consistently achieves state-of-the-art performance.

Yuhao Huang, Yuanji Zhang, Yuhuan Lu, Dong Ni, P. Ellen Grant

TL;DR: 本文提出了一种名为VIFBA的超声视频框架，用于胎儿大脑评估。该框架能够从常规超声视频中预测MRI衍生的侧脑室体积、分类脑室扩大严重程度并识别潜在的非脑室扩大异常，旨在解决传统超声测量依赖操作者且无法全面反映脑室整体状况，以及MRI成本高、可及性差的问题。

Details

Motivation: 传统胎儿脑室扩大的超声评估主要依赖标准切面上侧脑室房宽度的测量，这种方法具有操作者依赖性且可能无法完全反映整体脑室扩大情况。胎儿脑MRI能提供更可靠的体积信息，但成本高昂且难以常规使用。因此，需要一种基于常规超声的、更准确且可负担的产前大脑筛查方法。

Result: 在一个包含857例（3,196个视频）配对胎儿脑超声和MRI检查的大型数据集上进行验证。在留出测试集上，VIFBA在脑室体积回归任务上取得了0.5909 mL的MAE和0.9907的皮尔逊相关系数，在脑室扩大严重程度分类上达到0.9400的准确率，在多异常分类上获得0.7764的F1分数，显著优于单任务基线、基于视频的强竞争模型以及最先进的基础模型。

Insight: 创新点包括：1）受联合嵌入预测架构启发的管状潜在预测目标，利用超声视频的时空连贯性增强表征学习；2）对比跨模态对齐策略，在训练阶段将MRI的结构信息迁移到超声，而推理时仅需超声；3）结合免训练的视觉语言模型和检索增强来验证不确定预测并识别潜在非脑室扩大异常。这为从单一模态（超声）获取多模态（MRI）信息提供了有效途径。

Abstract: Assessment of ventriculomegaly (VM) on fetal brain ultrasound relies primarily on measuring lateral ventricular atrial width on standard planes, which is operator-dependent and may not fully reflect the overall ventricular enlargement. Fetal brain MRI provides more reliable volumetric information but is costly and less accessible for routine use. To address these limitations, we propose VIFBA, an ultrasound video-based framework for fetal brain assessment that predicts MRI-derived lateral ventricular volume, classifies VM severity, and identifies potential non-VM fetal brain abnormalities. Our contribution is three-fold. First, we introduce a joint-embedding predictive architecture (JEPA)-inspired tube latent prediction objective that leverages spatio-temporal coherence in ultrasound videos to enhance representation learning. Second, we develop a contrastive cross-modal alignment strategy that transfers structural information from MRI to ultrasound during training, while requiring ultrasound alone at inference. Third, we augment VIFBA with a training-free vision-language model and retrieval augmentation to verify uncertain predictions and identify potential non-VM fetal brain abnormalities. We validated VIFBA on a large dataset comprising 857 cases (3,196 videos) with paired fetal brain ultrasound and MRI examinations. On held-out test data, VIFBA achieved an MAE of 0.5909 mL and Pearson correlation coefficient of 0.9907 for ventricular volume regression, 0.9400 accuracy for VM severity classification, and an F1 score of 0.7764 for multi-abnormality classification, substantially outperforming single-task baselines, video-based strong competitors, and state-of-the-art foundation models. By enabling MRI-informed volumetric assessment from routine ultrasound alone, VIFBA offers a practical and potentially broadly deployable pathway toward accurate and affordable prenatal brain screening.

[204] A Unified DINOv2-Based Framework for LVEF Estimation, GLS Dysfunction Classification, and Early Cardiotoxicity Prediction eess.IV | cs.CVPDF

Xiaotong Zhang, Mingyue Cui, Qing Cao, Jingming Xia

TL;DR: 该论文提出了一个基于DINOv2的统一框架，用于解决心脏肿瘤学评估中的三个互补任务：左心室射血分数（LVEF）估计、基于整体纵向应变（GLS）的功能障碍分类以及早期心脏毒性预测。该框架利用冻结的基础编码器，结合参数高效的LoRA适配和时序聚合来学习任务专用表示，并在推理时无需心脏周期检测或ED/ES标注。此外，论文还引入了一个ED/ES引导的2D/3D混合多视图回归模型来专门优化LVEF估计任务。

Details

Motivation: 动机在于LVEF、GLS功能障碍分类和早期心脏毒性预测为心脏肿瘤学评估提供了互补信息，但现有方法可能依赖复杂的预处理（如周期检测）。论文旨在开发一个统一的、无需周期检测的框架，利用基础模型表征来同时解决这三个任务，并探索生理学引导建模对LVEF估计的额外益处。

Result: 在包含237名患者的1,203个训练视频和59名独立患者的300个验证视频的患者级数据集上，基于DINOv2的框架在Task 1（LVEF估计）上取得了5.03%的平均绝对误差（MAE），在Task 2（GLS功能障碍分类）上取得了76.48%的AUC-ROC，在Task 3（早期心脏毒性预测）上取得了70.26%的AUC-ROC。专门为Task 1设计的ED/ES引导模型进一步将MAE提升至4.64%。

Insight: 创新点在于提出了一个统一、无需周期检测（cycle-detection-free）和相位标注（phase-free）的DINOv2框架，通过LoRA适配高效处理多个任务。核心洞察是基础模型表征（DINOv2）在多样化的心脏肿瘤学任务中具有有效性和泛化能力，并且针对特定任务（如LVEF估计）引入生理学引导（ED/ES信息）的专门化建模可以进一步提升性能，这展示了通用表征与领域知识引导的专用模型相结合的优势。

Abstract: Left ventricular ejection fraction (LVEF) estimation (Task 1), global longitu-dinal strain (GLS)-based dysfunction classification (Task 2), and early cardi-otoxicity prediction (Task 3) provide complementary information for cardio-oncology assessment. LVEF reflects macroscopic ventricular volume chang-es as the clinical standard, whereas GLS captures subtle myocardial defor-mation, indicating subclinical cardiotoxicity before overt LVEF decline. Fur-thermore, predicting cardiotoxicity from baseline echocardiography prior to treatment enables preventive interventions at an early stage. To address these three tasks, we employ a DINOv2-based framework with task-specific adap-tation and prediction heads. Built upon a frozen foundation encoder, the framework incorporates parameter-efficient Low-Rank Adaptation (LoRA) and temporal aggregation to learn task-specialized representations, ensuring robust generalization. Crucially, during inference, it operates in a fully cycle-detection-free and phase-free manner, requiring neither cardiac cycle seg-mentation nor explicit End-Diastolic/End-Systolic (ED/ES) annotations. Ad-ditionally, we introduce an ED/ES-guided 2D/3D hybrid multi-view regres-sion model specifically to optimize Task 1. On a patient-level split containing 1,203 training videos from 237 patients and 300 validation videos from 59 independent patients, the DINOv2-based framework achieved a mean abso-lute error (MAE) of 5.03% for Task 1, an AUC-ROC of 76.48% for Task 2, and an AUC-ROC of 70.26% for Task 3. For Task 1, the specialized ED/ES-guided model further improves performance, achieving an MAE of 4.64%. This framework demonstrates the effectiveness of foundation model repre-sentations across diverse cardio-oncology tasks and the additional benefit of physiology-guided modeling for accurate LVEF estimation.

[205] Modality-Invariant Coarse-to-Fine Retinal Image Registration eess.IV | cs.CVPDF

Bo Wen, Nehal Nailesh Mehta, Melanie Tran, Dirk-Uwe Bartsch, William Freeman

TL;DR: 本文提出了一种通用的两阶段模态不变视网膜图像配准框架，旨在解决现有方法对特定成像模态依赖性强、灵活性不足的问题。该框架首先通过通用视网膜血管分割驱动的稀疏特征匹配实现跨模态的鲁棒粗全局对齐，然后利用模态不变光流估计网络MI-RAFT进行密集局部配准以细化对齐效果。

Details

Motivation: 现有视网膜图像配准方法通常依赖于特定成像模态（单模态或固定模态对），限制了其在涉及多种视网膜成像模态及不同组合的实际场景中的灵活性和适用性。

Result: 大量实验表明，所提方法能够处理常用视网膜成像模态的多种组合，展现出强大的模态不变性，并在性能上超越了当前最先进的模态依赖配准方法。

Insight: 创新点在于将配准过程解耦为稀疏特征匹配与密集光流估计两阶段，并引入通用血管分割作为跨模态对齐的桥梁，实现了对任意模态组合的泛化能力；从客观角度看，该方法通过分离模态特定特征与几何结构，有效提升了跨模态配准的鲁棒性和通用性。

Abstract: Retinal image registration is essential for ophthalmic diagnosis, longitudinal disease monitoring, and multimodal retinal image analysis. Existing retinal registration methods are typically modality-dependent: they are designed or optimized either for a single imaging modality in mono-modal registration or for a fixed pair of modalities in cross-modal registration. This limits their flexibility and applicability in practical scenarios involving diverse retinal imaging modalities and different combinations of them. In this work, we propose a generalizable two-stage, modality-invariant framework for retinal image registration. First, we introduce a sparse feature-matching model driven by a universal retinal vessel segmentation to achieve robust coarse global alignment across modalities. Second, we develop a modality-invariant optical flow estimation network, termed MI-RAFT, to refine the alignment through dense local registration. Extensive experiments demonstrate that the proposed method can handle diverse combinations of commonly used retinal imaging modalities, exhibiting strong modality invariance while outperforming state-of-the-art modality-dependent registration methods.

cs.CR [Back]

[206] DSPrompt: Dynamic Soft Prompt Defense Against M-RAG Corruption cs.CR | cs.CLPDF

Chang Liu, Yuni Lai, Mingyue Cui, Cong Tian, Yunyan Zhang

TL;DR: 本文提出了DSPrompt，一种针对多模态检索增强生成（M-RAG）对抗性攻击的动态软提示防御框架。该方法通过在冻结的检索器编码器中插入少量可学习的软提示，直接重塑嵌入语义，无需修改检索流程，从而有效抵御恶意数据攻击。

Details

Motivation: 现有的M-RAG防御方法主要在查询时操作，存在推理开销大、对未见攻击策略泛化能力差、且常假设特定攻击分布的问题。本文旨在设计一种高效、通用的防御机制来应对这些挑战。

Result: 在四个基准测试和三种代表性投毒攻击上的广泛实验表明，DSPrompt显著降低了攻击成功率和投毒检索率，同时保持了近乎无损的检索效用和生成保真度，以远低于现有基线的计算成本持续优于它们。

Insight: 创新点在于提出了一种动态极小极大训练方案，其中在线多模态攻击者持续生成对抗性文档，而防御者被更新以将这些文档推出前k个结果，同时保持良性证据的排序和多样性。该方法通过浅到深的长度调度自适应模型层容量，引入少于1%的额外参数，且支持预计算索引，无额外查询开销。

Abstract: Multimodal Retrieval Augmented Generation (M-RAG) is increasingly vulnerable to adversarial attacks where malicious data are crafted to produce embeddings that align with benign entries in the vector space, deceiving retrieval and inducing harmful outputs. Existing defenses primarily operate at query time, relying on auxiliary detectors, similarity re-ranking, or feature-consistency checks. However, these approaches suffer from non-trivial inference overhead, generalize poorly to unseen attack strategies, and often assume specific attack distributions. To address this, we propose DSPrompt, a Dynamic Soft Prompt defense framework that directly reshapes the retriever’s embedding semantics, without modifying the retrieval pipeline. It inserts few learnable soft prompts into each layer of the visual and textual encoders of a frozen retriever, utilizing a shallow-to-deep length schedule that is adaptive to the capacity in the model layers. These prompts are trained under a dynamic min-max scheme: an online multimodal attacker continually crafts hard adversarial documents against the current retriever, while the defender is updated to push such documents out of the top-k while preserving the ranking and diversity of benign evidence. Because the defended encoder can be pre-computed and indexed exactly as in standard dense retrieval, DSPrompt incurs no additional per-query optimization and introduces fewer than 1% additional parameters. Extensive experiments across four benchmarks and three representative poisoning attacks show that DSPrompt substantially reduces the attack success rate and poison retrieval rate while maintaining near-lossless retrieval utility and generation fidelity, consistently outperforming existing defense baselines at a fraction of their computational cost.

Table of Contents

cs.CL [Back]

[1] Inference-Time Mitigation of Adversarial Political Bias in Large Language Models cs.CL | cs.AIPDF

[2] Multi-Modal Generative Fuzzy System: Fuzzy Inference Guided Large Model Interactive Question Answering Framework cs.CL | cs.AIPDF

[3] Characterizing Rhetorical Misalignment in Decision-Making with Language Models cs.CL | cs.AIPDF

[4] Beyond Tokens: A Survey on Decoding Methods for Large Language and Vision-Language Models cs.CLPDF

[5] RamseyGadgets: A Graph Construction Dataset for LLMs cs.CL | cs.AIPDF

[6] Interpretable Cross-Lingual Alignment in Small Language Models: Probing Cultural and Pragmatic Reasoning in Japanese-English Bilingual LLMs cs.CL | cs.LGPDF

[7] Handoff-H1: An Orchestrated Vision-Agent System for Material Quantity Takeoff from Construction Blueprints cs.CL | cs.AI | cs.CVPDF

[8] TRACE-BN: Transferring Bangla-English Tutoring Behavior to a Sub-1B Offline Language Model cs.CLPDF

[9] Why Vision Fails as a Universal Bridge: Rectifying Modality Asynchrony in Multilingual MLLMs cs.CLPDF

[10] The Machine’s Internal Clock: Do LLMs Share Human Temporal Illusions? cs.CLPDF

[11] Do Language Models Consistently Encode the Current Year? cs.CL | cs.LGPDF

[12] Why Summaries Turn Neutral: Policy Attribution for Sentiment Drift in Reinforcement Learning from Human Feedback cs.CLPDF

[13] TaoLive Digital Avatar Agent Technical Report: Training Agents to Evolve with Their Harness cs.CLPDF

[14] MicroVerse: An Instrument for Measuring Self-Authored Identity Drift in Long-Horizon Multi-Agent Language-Model Simulations cs.CLPDF

[15] SEER: Long-Context Reasoning via Selective Visual-Text Compression cs.CL | cs.CVPDF

[16] LLMs Get Smarter from Targeted Synthetic Multilingual Data cs.CL | cs.AIPDF

[17] From Sequence to Structure: Relational Uncertainty Propagation for LLM Agents cs.CL | cs.AIPDF

[18] $R^3$-Bench: LLMs Struggle with Resource-Rational Reasoning under Shared Budgets cs.CLPDF

[19] Skill2Query: Exploiting Skill Structure to Generate Pseudo-Queries for Agent Skill Retrieval cs.CL | cs.IRPDF

[20] STAIR: Semantic-Temporal Automaton for Interpretable Reasoning in Temporal Question Answering cs.CL | cs.AIPDF

[21] Step-Level On-Policy Distillation: Interpolating Between On-Policy Distillation and Supervised Fine-Tuning cs.CL | cs.AIPDF

[22] Mint-Agent: Introducing Finance-Native Agentic Foundation Models cs.CL | cs.LGPDF

[23] D2-ScaleAgent: Dual-Dimensional Scaling for Long Document Understanding cs.CLPDF

[24] BabelSteering: Multilingual Safety Alignment via English Steering Vectors cs.CLPDF

[25] Ask, Condition or Abstain: Reinforcement Learning for Missing-Premise Reasoning cs.CLPDF

[26] When Do Explanations Help In-Context Learning? A Comparative Study of Natural Language Explanation Types and Faithfulness cs.CL | cs.AIPDF

[27] Every Coin Has Two Sides: On the Dual Nature of Generalization in On-Policy Distillation of Large Language Models cs.CLPDF

[28] ClawGym II: Exploring Black-Box RL on Agent Harness cs.CL | cs.AI | cs.LGPDF

[29] PCA-guided Activation Scaling for Monotonic Bidirectional Control over LLM Sycophancy cs.CLPDF

[30] Model Hypnosis: Strong control of AI via additive subliminal effects cs.CL | cs.AIPDF

cs.CV [Back]

[31] Deep Analog: Open-Set Film Emulation with Reference-Conditioned 3D LUTs cs.CV | cs.GR | cs.LG | cs.MMPDF

[32] Xemo-Talker: Unlock Emotions Explicitly for Audio-Driven Talking Portrait Synthesis cs.CV | cs.SDPDF

[33] On Cross-Validation for Hyperparameter Optimization of Deep Learning Image Classifiers cs.CV | cs.LGPDF

[34] Equilibrium Forcing: Adaptive Video Generation Without Noise Conditioning cs.CV | cs.AI | cs.LGPDF

[35] VideoGAIA: A Benchmark for General AI Assistants on Agentic Video Understanding cs.CV | cs.CLPDF

[36] DeCo-MIL: Debiased Counterfactual Reasoning for Long-Tailed Whole Slide Image Analysis cs.CV | cs.AIPDF

[37] AeroGround: A Comprehensive Benchmark for Aerial-Ground Collaborative Reasoning cs.CVPDF

[38] Braided Vision Transformer for Stroke Detection in Multi-view Retinal Fundus Imaging cs.CVPDF

[39] A Vision Transformer for ECG-Based Detection of Left Ventricular Systolic Dysfunction Across Multiple Clinical Sites cs.CV | cs.LGPDF

[40] Privacy-Preserving Dataset Curation for Kuala Lumpur Urban Traffic: Grounded Vision-Language Detection with Spatial Vehicle-Context Filtering cs.CV | cs.AI | cs.LGPDF

[41] Spatial Attention Noise Masking for Causally Sufficient Interpretability cs.CVPDF

[42] Low Cost Two-Stage Fabric Defect Detection at the Edge cs.CVPDF

[43] From Dense Prediction to Visual Editing: Structured Supervision for Unified Image and Video Creation cs.CVPDF

[44] PolyComp: A Polycube-based Benchmark for Compositional 3D Spatial Reasoning in Multimodal Models cs.CV | cs.AIPDF

[45] Beyond Boundary Noise: Aggregated Aleatoric Uncertainty Fails to Capture Presence Ambiguity in 3D Lung Nodule Segmentation cs.CV | cs.LGPDF

[46] NARRATE: A Multimodal Real-World Australian Driving Dataset for Human-Centred Explanations in Automated Driving cs.CV | cs.AI | cs.CL | cs.ROPDF

[47] AMPLIFAI: A Multiphase CT Dataset for Benchmarking Clinical Reasoning in LI-RADS Assessment of Liver Lesions cs.CV | cs.LGPDF

[48] Qwen-Video-Edit: Instruction-Based Video Editing by Repurposing an Image Editing Model cs.CVPDF

[49] Where the Cost Falls: A Deployment-Aware Adoption Order for Stability Enhancements to Cycle-Consistent Adversarial Networks cs.CVPDF

[50] Zero-Shot Adaptation of Medical Vision Foundation Models for High-Frequency Micro-Ultrasound Prostate Segmentation cs.CV | cs.LGPDF

[51] Artificial Intelligence as a Tool for Combating Child Labour: A Real-Time Edge Vision Pipeline for Child Detection and Age Estimation cs.CV | cs.AIPDF

[52] OvDSGG: End-to-End Open-Vocabulary Dynamic Scene Graph Generation cs.CVPDF

[53] Zero-MELO: Test-Time Evidence Calibration with Multimodal LLMs for Zero-Shot Micro-Gesture Recognition cs.CVPDF

[54] SpIn-ViT: Designing a Sparsity-Induced Vision Transformer That Is Mechanistically Interpretable cs.CV | cs.AIPDF

[55] PaSTel: Anchoring Histology in Spatial Transcriptomics via Multi-Scale Hierarchical Bio-Prior Contrastive Pretraining cs.CV | cs.AIPDF

[56] Beam-Wise Statistical Background Subtraction for Static Roadside LiDAR: A Cross-Sensor Benchmark Study cs.CV | cs.ROPDF

[57] Risk-Adaptive Edge–Cloud Visual Reasoning for Communication-Efficient Autonomous Driving cs.CVPDF

[58] Registration-Free Hyperspectral Reconstruction from RGB via a Permutation-Invariant Gram-Matrix Principle cs.CVPDF

[59] MetaReason: Precise Interleaved Multimodal Reasoning via Editing Meta Information for Solving Geometry Problems cs.CV | cs.AI | cs.MMPDF

[60] FZ-VLM: A Two Stage Florence-Zephyr Vision Language Model Framework for Pulmonary Nodule Characterization and Clinical Decision Making cs.CV | cs.AIPDF

[61] Frequency and Edge-Guided Segment Anything Model for Remote Sensing Image Semantic Segmentation cs.CVPDF

[62] MOSS-VL Technical Report cs.CVPDF

[63] MEDR: Query-Independent Frame Selection via Multi-Signal Event Modeling and Dynamic Rescoring cs.CVPDF

[64] Do Visual Grounding Decoders Need Feed-Forward Networks? A Controlled Study over Frozen Vision-Language Features cs.CVPDF

[65] EgoTac: In-the-wild Tactile Prediction from Egocentric Vision cs.CV | cs.ROPDF

[66] SA-GEM: Scale-Adaptive and Geospatial Evidence-Modulated Token Pruning for Efficient Remote Sensing Large Vision-Language Models cs.CVPDF

[67] MODAL: Multi-Modal Object Re-ID via Model-Driven Sparse Decoupling and Text-Image Differential Filtering cs.CV | eess.IVPDF

[68] Distribution-free false-alarm calibration and chance-corrected spatial evaluation for industrial anomaly detection cs.CV | cs.LGPDF

[69] Perspective-Invariant Attack with Enhanced Transferability of Adversarial Examples cs.CVPDF

[70] From “What-If” to “What-Is”: Counterfactual Thinking-Inspired Semantic Alignment for Visual Brain Decoding cs.CV | cs.HCPDF

[71] Beyond Natural-Image Foundation Models: Benchmarking Satellite Pretraining for Ophthalmic Image Analysis cs.CVPDF

[72] Anchor-Regularized Adaptation for Generalizable AI-Generated Image Detection with DINOv3 cs.CVPDF

[73] PersonaDrive: Controllable Trajectory Prediction with Multi-Dimensional Driving Personas cs.CVPDF

[74] DCA-MoE: Spatially Adaptive Cross-Layer Fusion and Density-Routed Experts for Crowd Counting cs.CV | cs.IRPDF

[75] UC-VLM: Consistency-Driven Learning for AI-Generated Image Detection with Vision-Language Large Models cs.CVPDF

[76] UAV Video Deblurring via Motion-Aware Diffusion: A Path to Robust Target Detection cs.CV | cs.AI | cs.ROPDF

[77] Boundary-Aligned Contribution Routing for Robust Optical–SAR Object Detection cs.CVPDF