IllusionBench+: A Large-scale and Comprehensive Benchmark for Visual Illusion Understanding in Vision-Language Models
This addresses the issue of limited perceptual abilities and hallucinations in VLMs for researchers and developers, though it is incremental as it builds on existing benchmark efforts.
The authors tackled the problem of visual illusion understanding in vision-language models by introducing IllusionBench+, a large-scale benchmark with real-world and trap illusions, and found that top models like GPT-4o achieve up to 80.59% accuracy but still lag behind human performance and show hallucinations.
Current Visual Language Models (VLMs) show impressive image understanding but struggle with visual illusions, especially in real-world scenarios. Existing benchmarks focus on classical cognitive illusions, which have been learned by state-of-the-art (SOTA) VLMs, revealing issues such as hallucinations and limited perceptual abilities. To address this gap, we introduce IllusionBench, a comprehensive visual illusion dataset that encompasses not only classic cognitive illusions but also real-world scene illusions. This dataset features 1,051 images, 5,548 question-answer pairs, and 1,051 golden text descriptions that address the presence, causes, and content of the illusions. We evaluate ten SOTA VLMs on this dataset using true-or-false, multiple-choice, and open-ended tasks. In addition to real-world illusions, we design trap illusions that resemble classical patterns but differ in reality, highlighting hallucination issues in SOTA models. The top-performing model, GPT-4o, achieves 80.59% accuracy on true-or-false tasks and 76.75% on multiple-choice questions, but still lags behind human performance. In the semantic description task, GPT-4o's hallucinations on classical illusions result in low scores for trap illusions, even falling behind some open-source models. IllusionBench is, to the best of our knowledge, the largest and most comprehensive benchmark for visual illusions in VLMs to date.