Semi-Supervised Pixel Contrastive Learning Framework for Tissue Segmentation in Histopathological Image

Gastric Cancer Semantic Segmentation Dataset (GCSS):

Annotation: The annotation of each ROI is a pixel matrix. {1,2,3,4,5} represent tumor, lymphoid stroma, desmoplastic stroma, smooth muscle, and necrosis. {-1, 0, 6} represent others. 

Data statistics: 100 ROIs cropped from the 100 gastric cancer cases of the TCGA-STAD project.

GitHub: https://github.com/Jiangbo-Shi/SSPCL