Illusion-Aware Visual Preprocessing and Anti-Illusion Prompting for Classic Illusion Understanding in Vision-Language Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913106630803456 |
|---|---|
| author | Zha, Junli Wang, Jiahui Lu, Xinkai Wang, Jinbo |
| author_facet | Zha, Junli Wang, Jiahui Lu, Xinkai Wang, Jinbo |
| contents | Vision-Language Models (VLMs) exhibit systematic bias toward visual illusions, recalling memorized facts rather than perceiving actual visual differences. This paper presents a training-free framework for the 5th DataCV Challenge Task 1 at CVPR 2026, addressing this perception-versus-memory conflict through three complementary strategies:(1) illusion-aware image preprocessing that weakens illusion-inducing context via type-specific transformations (edge extraction, color isolation, morphological processing, and reference-line overlay), (2) anti-illusion prompt engineering guiding VLMs toward qualitative visual comparison, and (3) multi-vote ensemble that further improves robustness. Our method achieves 90.48% accuracy on the official 630-image test set using Claude (claude-opus-4-6) with 5-vote majority ensemble, and 98.41% on a human-verified subset. The approach requires no finetuning, relying solely on visual manipulation and prompt design. Our solution secured 2nd place in the challenge, only 0.47% behind the 1st-place solution. Code is available at https://github.com/jasminezz/sf-illusion-aware-vlm.git. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_08841 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Illusion-Aware Visual Preprocessing and Anti-Illusion Prompting for Classic Illusion Understanding in Vision-Language Models Zha, Junli Wang, Jiahui Lu, Xinkai Wang, Jinbo Computer Vision and Pattern Recognition Vision-Language Models (VLMs) exhibit systematic bias toward visual illusions, recalling memorized facts rather than perceiving actual visual differences. This paper presents a training-free framework for the 5th DataCV Challenge Task 1 at CVPR 2026, addressing this perception-versus-memory conflict through three complementary strategies:(1) illusion-aware image preprocessing that weakens illusion-inducing context via type-specific transformations (edge extraction, color isolation, morphological processing, and reference-line overlay), (2) anti-illusion prompt engineering guiding VLMs toward qualitative visual comparison, and (3) multi-vote ensemble that further improves robustness. Our method achieves 90.48% accuracy on the official 630-image test set using Claude (claude-opus-4-6) with 5-vote majority ensemble, and 98.41% on a human-verified subset. The approach requires no finetuning, relying solely on visual manipulation and prompt design. Our solution secured 2nd place in the challenge, only 0.47% behind the 1st-place solution. Code is available at https://github.com/jasminezz/sf-illusion-aware-vlm.git. |
| title | Illusion-Aware Visual Preprocessing and Anti-Illusion Prompting for Classic Illusion Understanding in Vision-Language Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.08841 |