CUNY Academic Commons

In today's meeting we decided to explore an achievable way to automatically recognize (and even process/convert) non-OCR PDFs uploaded to the Commons, one of our key accessibility challenges.

This CUNY AI Lab tool works well for conversion: https://ailab.gc.cuny.edu/

Boone also found https://verapdf.org/ which might be a pretty robust tool. He suggested we could set up a private instance of it and do validation against it when PDFs are uploaded.

At the very least, if we aren't able to implement an automatic solution, we should improve our documentation with our best suggested OCR solution and promote it to users as much as possible.