arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2511.11206cs.CVcs.LG
收藏

质疑视觉问答的稳定性

Questioning the Stability of Visual Question Answering

Bar-Ilan University(巴伊兰大学)

Amir Rosenfeld, Neta Glazer, Ethan Fetaya

AI总结:

本文首次大规模系统性研究视觉语言模型对不改变语义的微小视觉、文本扰动的鲁棒性,发现当前先进模型对此类扰动高度敏感,且样本稳定性可作为答案正确性的有效指标。

摘要:

视觉语言模型(VLMs)已取得显著进展,但学界对其在不改变语义的微小输入变动下的可靠性仍知之甚少。本文首次针对视觉语言模型对良性视觉与文本扰动的鲁棒性开展大规模系统性研究,涉及的扰动包括像素级偏移、轻度几何变换、填充式重缩放、释义改写以及多语言重写,这些扰动均不会改变图像-问题对的底层语义。在覆盖多款模型与多个数据集的实验中,我们发现现代视觉语言模型对这类微小扰动高度敏感:占比相当大的样本在至少一种视觉或文本修改下会改变预测答案。我们刻画了这种不稳定性在不同扰动类型、问题类别、模型之间的差异,揭示出即使是当前最先进的系统(如GPT-4o、Gemini 2.0 Flash),在仅几个像素的偏移或无恶意的改写下也会频繁出错。我们进一步证明,样本级稳定性是答案正确性的强指标:稳定样本被正确回答的概率始终远高于不稳定样本。利用这一规律,我们证实小型、易获取的开源模型的稳定性模式可用于高精度预测规模大得多的闭源模型的答案正确性。本研究的发现暴露了当前视觉语言模型存在的根本性脆弱性,强调鲁棒性评估需要跳出对抗性扰动范畴,转而聚焦模型本应可靠具备的不变性。

英文摘要:

Visual Language Models (VLMs) have achieved remarkable progress, yet their reliability under small, meaning-preserving input changes remains poorly understood. We present the first large-scale, systematic study of VLM robustness to benign visual and textual perturbations: pixel-level shifts, light geometric transformations, padded rescaling, paraphrasing, and multilingual rewrites that do not alter the underlying semantics of an image-question pair. Across a broad set of models and datasets, we find that modern VLMs are highly sensitive to such minor perturbations: a substantial fraction of samples change their predicted answer under at least one visual or textual modification. We characterize how this instability varies across perturbation types, question categories, and models, revealing that even state-of-the-art systems (e.g., GPT-4o, Gemini 2.0 Flash) frequently fail under shifts as small as a few pixels or harmless rephrasings. We further show that sample-level stability serves as a strong indicator of correctness: stable samples are consistently far more likely to be answered correctly. Leveraging this, we demonstrate that the stability patterns of small, accessible open-source models can be used to predict the correctness of much larger closed-source models with high precision. Our findings expose a fundamental fragility in current VLMs and highlight the need for robustness evaluations that go beyond adversarial perturbations, focusing instead on invariances that models should reliably uphold.

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。

↑