Hacker News new | ask | show | jobs
by pbhjpbhj 4372 days ago
It would have to be a author-guided OCR though as the same symbols can represent different words.
2 comments

You can probably get pretty far with machine learning.
Doesn't this have audio?