WebAccessVL: Violation-Aware VLM for Web Accessibility
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910048350896128 |
|---|---|
| author | Zheng, Amber Yijia Lee, Jae Joong Benes, Bedrich Yeh, Raymond A. |
| author_facet | Zheng, Amber Yijia Lee, Jae Joong Benes, Bedrich Yeh, Raymond A. |
| contents | We present a vision-language model (VLM) that automatically edits website HTML to address violations of the Web Content Accessibility Guidelines 2 (WCAG2) while preserving the original design. We formulate this as a supervised image-conditioned program synthesis task, where the model learns to correct HTML given both the code and its visual rendering. We create WebAccessVL, a website dataset with manually corrected accessibility violations. We then propose a violation-conditioned VLM that further takes the detected violations' descriptions from a checker as input. This conditioning enables an iterative checker-in-the-loop refinement strategy at test time. We conduct extensive evaluation on both open API and open-weight models. Empirically, our method achieves 0.211 violations per website, a 96.0\% reduction from the 5.34 violations in raw data and 87\% better than GPT-5. A perceptual study also confirms that our edited websites better maintain the original visual appearance and content. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_03850 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | WebAccessVL: Violation-Aware VLM for Web Accessibility Zheng, Amber Yijia Lee, Jae Joong Benes, Bedrich Yeh, Raymond A. Human-Computer Interaction Artificial Intelligence Computer Vision and Pattern Recognition We present a vision-language model (VLM) that automatically edits website HTML to address violations of the Web Content Accessibility Guidelines 2 (WCAG2) while preserving the original design. We formulate this as a supervised image-conditioned program synthesis task, where the model learns to correct HTML given both the code and its visual rendering. We create WebAccessVL, a website dataset with manually corrected accessibility violations. We then propose a violation-conditioned VLM that further takes the detected violations' descriptions from a checker as input. This conditioning enables an iterative checker-in-the-loop refinement strategy at test time. We conduct extensive evaluation on both open API and open-weight models. Empirically, our method achieves 0.211 violations per website, a 96.0\% reduction from the 5.34 violations in raw data and 87\% better than GPT-5. A perceptual study also confirms that our edited websites better maintain the original visual appearance and content. |
| title | WebAccessVL: Violation-Aware VLM for Web Accessibility |
| topic | Human-Computer Interaction Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.03850 |