The majority of scientific papers are distributed in PDF, which pose challenges for accessibility, especially for blind and low vision (BLV) readers.
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.
Adriana here from the Semantic Scholar team at AI2.
Thanks for your interest in the Semantic Reader beta! We're actively testing and developing new features and welcome your feedback (https://www.surveymonkey.com/r/Q5K2XPK).
I work on the Semantic Scholar team at the Allen Institute for AI (allenai.org). We're working on solving the problems described in the post.
We're investing considerable effort into making it easier for researchers to find and consume scientific literature. Our team is made up of engineers and researchers who have felt the pain points firsthand and are very motivated to design and build solutions that fix them. Our software is free to use, and always will be.
Our search engine uses things like citation intent, citation influence, and figure and table extraction to make filtering through papers easier. We’re also currently prototyping an augmented reading experience that aims to embed contextual information directly into the reading experience so that it's easier to consume and comprehend academic literature. Give it a try: semanticscholar.org.
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.
Preprint: https://arxiv.org/abs/2105.00076
Blog post: https://medium.com/ai2-blog/scia11y-improving-the-accessibil...
Demo: https://scia11y.org/
Disclosure: I work for the Semantic Scholar team at the Allen Institute for AI, the nonprofit research organization working on SciA11y.