VLM's Eye Examination: Instruct and Inspect Visual Competency of Vision Language Models
This work addresses a fundamental problem for VLM researchers by providing insights into visual competency, which is incremental as it builds on existing VLM analysis methods.
The authors tackled the limited understanding of visual perception in vision language models (VLMs) by proposing an eye examination process to investigate their sensitivity to color, shape, and semantic matching, revealing that VLMs have varying color sensitivities and are consistently insensitive to green, with shape and semantic recognition depending on LLM capacity.
Vision language models (VLMs) have shown promising reasoning capabilities across various benchmarks; however, our understanding of their visual perception remains limited. In this work, we propose an eye examination process to investigate how a VLM perceives images, specifically focusing on key elements of visual recognition, from primitive color and shape to semantic levels. To this end, we introduce a dataset named LENS to guide a VLM to follow the examination and check its readiness. Once the model is ready, we conduct the examination. Through this examination, we quantify and visualize VLMs' sensitivities to color and shape, and semantic matching. Our findings reveal that VLMs have varying sensitivity to different colors while consistently showing insensitivity to green across different VLMs. Also, we found different shape sensitivity and semantic recognition depending on LLM's capacity despite using the same fixed visual encoder. Our analyses and findings have potential to inspire the design of VLMs and the pre-processing of visual input to VLMs for improving application performance.