search

facebookresearch / nougat

Implementation of Nougat Neural Optical Understanding for Academic Documents
https://facebookresearch.github.io/nougat/
MIT License
8.98k stars 567 forks source link

Can it be used to ocr non-English text? #3

Closed nissansz closed 1 year ago

nissansz commented 1 year ago

Can it be used to ocr non-English text?

lukas-blecher commented 1 year ago

We've done some experiments with other Latin-based languages. The results were mostly satisfactory, although any special characters from these languages will be replaced with the closest equivalent from the Latin alphabet. Non-Latin script languages are not supported by the training set at all, so it won't work in these cases

nissansz commented 1 year ago

For text line position detection, is mupdf better than cptn?

lukas-blecher commented 1 year ago

I have no insight on that matter. We are not performing a text detection step