ACM Multimedia 2023: DocDiff: Document Enhancement via Residual Diffusion Models. Also contains 1597 red seals in Chinese scenes, along with their corresponding binary masks.
ocr deep-learning documentation-tool pytorch dataset super-resolution image-to-image image-translation deblurring document-binarization low-level-vision table-recognition img2img diffusion-models math-ocr seal-removal
-
Updated
Aug 22, 2024 - Python