Yichi Zhang, Jiayi Pan, Yuchen Zhou, Rui Pan, Joyce Chai. EMNLP 2023.
In this paper, we ask, do Vision-Language Models (VLMs), an emergent human-computer interface, perceive visual illusions like humans? Or do they faithfully represent reality. We built VL-Illusion, a new dataset that systematically evaluate the problem. And among all other exciting findings, we found that although model's humanlike rate is low under illusion, larger models are more susceptible to visual illusions, and closer to human perception.
The 2023 Conference on Empirical Methods in Natural Language Processing