Using LLama2 to Correct OCR Errors
Enhances Tesseract OCR output using LLMs (local or API) for error correction, smart chunking, and markdown formatting of scanned PDFs. I've been disappointed by the very poor quality of results that I generally get when trying to run OCR on older scanned documents, especially ones that are typewritten or otherwise have unusual or irregular typography.
Open source
Reactions (1)
Similar products
by description similarity
LLM-aided OCR
github.comCorrecting Tesseract OCR errors with LLMs
🔍 Better text detection by combining multiple OCR engines (EasyOCR, Tesseract, and Pororo) with 🧠 LLM.
LLM Aided Transcription Improvement
github.comI recently submitted another project for using LLMs to correct errors and improve formatting of OCRed documents which was well received.
PDF to MD by LLMs
github.comExtract Text/Tables/Image Descriptives by GPT4o
(YC S23) OCR that works on the messiest PDFs
trycardinal.ai