Top
Best
New

Posted by Curiositry 12 hours ago

Chandra-OCR(github.com)
4 points | 2 comments
gumboshoes 11 hours ago|
As is too often the case, it doesn't follow what I consider to be a fundamental OCR rule: the input document type should also be a possible output document type. So, PDF in should mean that composited PDF output is possible. This is necessary for groundtruthing.
jauntywundrkind 7 hours ago|
Also at https://huggingface.co/datalab-to/chandra