Do Images Speak Louder than Words? Investigating the Effect of Textual Misinformation in VLMs
This addresses a critical robustness limitation in VLMs for applications like visual question-answering, though it is incremental as it builds on existing research on misinformation in text-only domains.
The paper tackled the problem of Vision-Language Models (VLMs) being vulnerable to textual misinformation that conflicts with visual evidence, revealing that these models often override clear visual cues in favor of misleading text, with an average performance drop of over 48.2% after persuasive conversation.
Vision-Language Models (VLMs) have shown strong multimodal reasoning capabilities on Visual-Question-Answering (VQA) benchmarks. However, their robustness against textual misinformation remains under-explored. While existing research has studied the effect of misinformation in text-only domains, it is not clear how VLMs arbitrate between contradictory information from different modalities. To bridge the gap, we first propose the CONTEXT-VQA (i.e., Conflicting Text) dataset, consisting of image-question pairs together with systematically generated persuasive prompts that deliberately conflict with visual evidence. Then, a thorough evaluation framework is designed and executed to benchmark the susceptibility of various models to these conflicting multimodal inputs. Comprehensive experiments over 11 state-of-the-art VLMs reveal that these models are indeed vulnerable to misleading textual prompts, often overriding clear visual evidence in favor of the conflicting text, and show an average performance drop of over 48.2% after only one round of persuasive conversation. Our findings highlight a critical limitation in current VLMs and underscore the need for improved robustness against textual manipulation.