Domain Adaptation for 2D Medical Image Segmentation

This dataset includes two 2D medical image segmentation benchmark.1. OD/OC Segmentation in Fundus ImageThis dataset conprises five sub-datasets: Drishti-GS, RIM-ONE-r3, ORIGA, REFUGE, and the validation set of REFUGE2. Each image is cropped around the optic disc area. The size of all images is 512×512. The manual pixel-wise annotation is stored as a PNG image with the same size as the corresponding fundus image with the following labels:128: Optic Disc (Grey color)0: Optic Cup (Black color)255: Background (White color)2. Polyp Segmentation in Endoscopy ImageThis dataset conprises four sub-datasets: BKAI, CVC-ClinicDB, ETIS-LaribPolypDB, and Kvasir-SEG. The manual pixel-wise annotation is stored as a PNG image with the same size as the corresponding fundus image with the following labels:255: Polyp (White color)0: Background (Black color)