Our long-term goal is to build efficient and reliable 2.5B diffusion-based decoding for document OCR. MinerU-Diffusion reframes document OCR as an inverse rendering problem and replaces slow, ...
Python OCR pipeline for scanned company-list PDFs or page-level images. It processes JPG/JPEG/PNG files directly as one page each, renders PDF pages when PDFs are supplied, splits company entries by ...