4 ms·
> I extracted the embedded text from the PDF What did you use to extract the embedded text during this step? Other than some other OCR tech
by nbbaier 2y ago
> I extracted the embedded text from the PDF
What did you use to extract the embedded text during this step? Other than some other OCR tech
- hugodutka 2y agoPyMuPDF, a PDF library for Python.
- jimmySixDOF 2y agoA different approach from vanilla OCR/parsing seems to be this mixed ColPali integrating a purposed small vision models and a ColBERT type indexing for retrieval. So - if search is the intended use case - it can skip the whole OCR step entirely. [1] https://huggingface.co/blog/manu/colpali https://huggingface.co/blog/manu/colpali