The current environment has the OCR library, Tesseract installed, but although this is the most frequently used, it fails to have all it’s language packages installed. Also it is not the best for let’s say handwritten notes, which GPT’s certainly could make use of in uploaded pictures of notes and such (writings on black board, etc.).
Would it be possible to implement another library (with full language support), like fx. Kraken OCR?
[GitHub - mittagessen/kraken: OCR engine for all the languages], which specializes in old writings and handwritings, etc.
Another option is OCRopus, or even making sure all the language packages for Tesseract is available in the Python environment.