The widespread use of image tables presents significant accessibility challenges for blind and low vision (BLV) people, limiting their access to critical data. Despite advancements in artificial intelligence (AI) for interpreting image tables, current solutions often fail to consider the specific needs of BLV users, leading to a poor user experience. To address these issues, we introduce TableNarrator, an innovative system designed to enhance the accessibility of image tables. Informed by accessibility standards and user feedback, TableNarrator leverages AI to generate alternative text tailored to the cognitive and reading preferences of BLV users. It streamlines access through a simple interaction mode and offers personalized options. Our evaluations, from both technical and user perspectives, demonstrate that TableNarrator not only provides accurate and comprehensive table information but also significantly enhances the user experience for BLV people.
https://dl.acm.org/doi/10.1145/3706598.3714329
The ACM CHI Conference on Human Factors in Computing Systems (https://chi2025.acm.org/)