U-Net: Convolutional Networks for Biomedical Image Segmentation

Benchmark Model Rank Results
colorectal-gland-segmentation-on-cragU-Net (e)#5Dice: 0.844Hausdorff Distance (mm): 196.9
colorectal-gland-segmentation-on-cragFCN8 (e)#10Hausdorff Distance (mm): 199.5
crack-segmentation-on-crackvision12kUNet#2mIoU: 0.60333
dichotomous-image-segmentation-on-dis-te1UNet#18max F-Measure: 0.625weighted F-measure: 0.514MAE: 0.106
dichotomous-image-segmentation-on-dis-te2UNet#17max F-Measure: 0.703weighted F-measure: 0.597MAE: 0.107
dichotomous-image-segmentation-on-dis-te3UNet#14max F-Measure: 0.748weighted F-measure: 0.644MAE: 0.098HCE: 883
dichotomous-image-segmentation-on-dis-te4UNet#11max F-Measure: 0.759weighted F-measure: 0.659MAE: 0.102
dichotomous-image-segmentation-on-dis-vdUNet#16max F-Measure: 0.692weighted F-measure: 0.586MAE: 0.113
lesion-segmentation-on-university-of-waterlooU-Net#3Dice score: 0.836 ±0.132
medical-image-segmentation-on-cvc-clinicdbU-Net#40mean Dice: 0.8230
medical-image-segmentation-on-kvasir-segU-Net#47mean Dice: 0.8180Average MAE: 0.055S-Measure: 0.858
multi-tissue-nucleus-segmentation-on-kumarU-Net (e)#11Dice: 0.758Hausdorff Distance (mm): 47.8
retinal-vessel-segmentation-on-chase-db1U-Net#12AUC: 0.9772
retinal-vessel-segmentation-on-driveU-Net#14AUC: 0.9755F1 score: 0.8142
retinal-vessel-segmentation-on-rose-1-dvcU-Net#2Dice Score: 66.05
retinal-vessel-segmentation-on-rose-1-svcU-Net#5Dice Score: 71.16
retinal-vessel-segmentation-on-rose-1-svc-dvcU-Net#5Dice Score: 70.12
retinal-vessel-segmentation-on-rose-2U-Net#5Dice Score: 65.64
retinal-vessel-segmentation-on-stareU-Net#6AUC: 0.7783F1 score: 0.8373
semantic-segmentation-on-bjroadUNet#6IoU: 54.88
semantic-segmentation-on-event-basedU-Net#5mIoU: 64.7
semantic-segmentation-on-fine-grained-grassUNet#5mIoU: 48.17
semantic-segmentation-on-selmaUNet#7mIoU: 36.2
semantic-segmentation-on-skyscapes-dense-1U-Net#6Mean IoU: 14.15
semantic-segmentation-on-trans10kU-Net#15GFLOPs: 124.55
semantic-segmentation-on-urbanlfOCR (HRNetV2-W48)#6mIoU (Syn): 79.36mIoU (Real): 78.60
thermal-image-segmentation-on-mfn-datasetUNet#40mIOU: 45.1
thermal-image-segmentation-on-pst900UNet#16mIoU: 52.8
video-polyp-segmentation-on-sun-seg-easyUNet#14Sensitivity: 0.420
video-polyp-segmentation-on-sun-seg-easy-1UNet#6S measure: 0.669mean E-measure: 0.677
video-polyp-segmentation-on-sun-seg-hardUNet#14Sensitivity: 0.429
video-polyp-segmentation-on-sun-seg-hard-1UNet#6Dice: 0.542S-Measure: 0.670