-
Notifications
You must be signed in to change notification settings - Fork 15
ModelDrivenSegmentation
Kay-Michael Würzner edited this page Sep 4, 2019
·
2 revisions
- dhSegment: A generic deep-learning approach for document segmentation: Oliveira & Seguin, 2019
- https://github.com/ocr-d-modul-2-segmentierung/page-segmentation
- https://github.com/dhlab-epfl/dhSegment
- https://github.com/mjenckel/LAYoutERkennung
- https://fairyonice.github.io/Learn-about-Fully-Convolutional-Networks-for-semantic-segmentation.html
Maybe the actual work should be done in a separate component reading a (binarized/raw) image and writing a (pixel-classified) image, which merely gets wrapped for OCR-D interfaces and PAGE-XML GT.