The majority of scientific papers are distributed in PDF, which pose challenges for accessibility, especially for blind and low vision (BLV) readers.<p>SciA11y incorporates several machine learning models to extract the content of scientific PDFs and render this content as accessible HTML, with added novel navigational features to support screen reader users.<p>Preprint: <a href="https://arxiv.org/abs/2105.00076" rel="nofollow">https://arxiv.org/abs/2105.00076</a><p>Blog post: <a href="https://medium.com/ai2-blog/scia11y-improving-the-accessibility-of-scientific-documents-1-of-2-48e1a7025a09" rel="nofollow">https://medium.com/ai2-blog/scia11y-improving-the-accessibil...</a><p>Demo: <a href="https://scia11y.org/" rel="nofollow">https://scia11y.org/</a><p>Disclosure: I work for the Semantic Scholar team at the Allen Institute for AI, the nonprofit research organization working on SciA11y.