Tesseract is way outdated though, to the point of being borderline useless when compared to alternatives. What’s the current deep learning based FOSS SOTA, does anyone know? I want something that does what FineReader does - create a high quality searchable text underlay for scanned PDFs.
