UNet++: A Nested U-Net Architecture for Medical Image Segmentation

Benchmark Model Rank Results
camouflaged-object-segmentation-on-pcod-1200UNet++#15S-Measure: 0.801
medical-image-segmentation-on-2018-dataUnet++#8Dice: 0.8974mIoU: 0.9255
medical-image-segmentation-on-cvc-clinicdbU-Net++#42mean Dice: 0.7940
medical-image-segmentation-on-kvasir-segU-Net++#45mean Dice: 0.8210Average MAE: 0.048S-Measure: 0.862
semantic-segmentation-on-cityscapes-valUNet++ (ResNet-101)#68mIoU: 75.5
video-polyp-segmentation-on-sun-seg-easyUNet++#13Sensitivity: 0.457
video-polyp-segmentation-on-sun-seg-easy-1UNet++#5Dice: 0.559S measure: 0.684mean E-measure: 0.687
video-polyp-segmentation-on-sun-seg-hardUNet++#13Sensitivity: 0.467
video-polyp-segmentation-on-sun-seg-hard-1UNet++#5Dice: 0.554S-Measure: 0.685mean E-measure: 0.697