Skip to content

Latest commit

 

History

History
2 lines (2 loc) · 353 Bytes

README.md

File metadata and controls

2 lines (2 loc) · 353 Bytes

painting-segmentation

Toolbox of three image segmentation algorithms. The first is a naive region growing/shrining approach. Next is a U-Net with a resnet backbone pretrained on the COCO dataset with options to fine-tune it using the Bob Ros Paintings Dataset. This project aimed to explore classic image segmentation algorithms applied to paintings.