从准确性到视觉依赖性：审计与过滤交通视频问答中的模态崩溃
From Accuracy to Visual Dependence: Auditing and Filtering Modality Collapse in Traffic VideoQA
AI总结：
针对交通视频问答中模型依赖文本捷径而非视觉证据的问题，提出盲差距和视觉增益两个数据集级诊断指标，以及实例级捷径分数实现无训练过滤，提升视觉基础。
摘要：
高基准准确率并不能保证真正使用视觉证据。我们在交通事故视频问答（VideoQA）中研究这一问题，其中正确答案应依赖于场景特定的视觉证据，但可能反而从文本捷径中推断。通过对四个公开基准的审计，我们发现几个最近的开源视觉语言模型（VLM）在没有视频输入的情况下表现具有竞争力，有时甚至更好。在MM-AU基准上，移除视频持续提高准确率，而添加更多帧进一步降低性能。为了量化视觉依赖性，我们引入了两个数据集级诊断指标：盲差距，衡量高于随机水平的纯文本性能；视觉增益，衡量添加视频的边际收益。我们进一步提出一个实例级捷径分数，结合纯文本置信度和视觉必要性信号，实现对易受捷径影响问题的连续、无训练过滤。得到的子集减少了捷径偏差并改善了视觉基础。我们的发现揭示了不同基准在基础质量上的巨大差异，并表明在安全关键的VideoQA中，视觉基础的评估（而不仅仅是高准确率）至关重要。
英文摘要：
High benchmark accuracy does not guarantee genuine use of visual evidence. We study this problem in traffic accident Video Question Answering (VideoQA), where correct answers should depend on scene-specific visual evidence but may instead be inferred from textual shortcuts. Through an audit of four public benchmarks, we find that several recent open-weight Vision-Language Models (VLMs) perform competitively, and sometimes better, without video input. On the MM-AU benchmark, removing video consistently improves accuracy, and adding more frames further degrades performance. To quantify visual dependence, we introduce two dataset-level diagnostics: Blind Gap, measuring above-chance text-only performance, and Visual Gain, measuring the marginal benefit of adding video. We further propose an instance-level Shortcut Score that combines text-only confidence with visual necessity signals, enabling continuous, training-free filtering of shortcut-prone questions. The resulting subsets reduce shortcut bias and improve visual grounding. Our findings reveal large differences in grounding quality across benchmarks and show that visually grounded evaluation, not just high accuracy, is essential in safety-critical VideoQA.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
