Back to Search Start Over

IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models

Authors :
Shahgir, Haz Sameen
Sayeed, Khondker Salman
Bhattacharjee, Abhik
Ahmad, Wasi Uddin
Dong, Yue
Shahriyar, Rifat
Publication Year :
2024

Abstract

The advent of Vision Language Models (VLM) has allowed researchers to investigate the visual understanding of a neural network using natural language. Beyond object classification and detection, VLMs are capable of visual comprehension and common-sense reasoning. This naturally led to the question: How do VLMs respond when the image itself is inherently unreasonable? To this end, we present IllusionVQA: a diverse dataset of challenging optical illusions and hard-to-interpret scenes to test the capability of VLMs in two distinct multiple-choice VQA tasks - comprehension and soft localization. GPT4V, the best performing VLM, achieves 62.99% accuracy (4-shot) on the comprehension task and 49.7% on the localization task (4-shot and Chain-of-Thought). Human evaluation reveals that humans achieve 91.03% and 100% accuracy in comprehension and localization. We discover that In-Context Learning (ICL) and Chain-of-Thought reasoning substantially degrade the performance of Gemini-Pro in the localization task. Tangentially, we discover a potential weakness in the ICL capabilities of VLMs: they fail to locate optical illusions even when the correct answer is in the context window as a few-shot example.

Details

Database :
arXiv
Publication Type :
Report
Accession number :
edsarx.2403.15952
Document Type :
Working Paper