Methods | Backbone | Sup | CRAG | GlaS | GlaS-A | GlaS_B | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ObjF1 | ObjDice | Dice | HD | ObjF1 | ObjDice | Dice | HD | ObjF1 | ObjDice | HD | ObjF1 | ObjDice | HD | |||
DCAN(report) | DeepLab-v1(VGG) | FullSup | 0.736 | 0.794 | 218.76 | 0.856 | 0.850 | 94.51 | ||||||||
DCAN | DeepLab-v1(VGG) | FullSup | 0.754 | 0.806 | 253.93 | 0.827 | 0.853 | 87.85 | ||||||||
MILD-Net(report) | CNN(self-design) | FullSup | 0.825 | 0.875 | 160.14 | 0.879 | 0.874 | 73.715 | 0.914 | 0.913 | 41.54 | 0.844 | 0.836 | 105.89 | ||
MILD-Net | CNN(self-design) | FullSup | 0.807 | 0.843 | 86.423 | |||||||||||
MedT | Transformer(self-design) | FullSup | ||||||||||||||
SwinUnet | Transformer(swinTransformer) | FullSup | 0.797 | 0.845 | 0.935 | 184.35 | 0.824 | 0.867 | 0.924 | 72.86 | ||||||
SwinUnet-Semi-round1 | Transformer(swinTransformer) | SemiSup-10% | 0.579 | 0.684 | 0.857 | 359.76 | 0.781 | 0.804 | 0.891 | 118.93 | ||||||
SwinUnet-Semi-round2 | Transformer(swinTransformer) | SemiSup-10% | 0.678 | 0.765 | 0.885 | 265.98 | 0.827 | 0.843 | 0.896 | 83.08 | ||||||
I2CS-B1(report) | EfficientNet-B1 | FullSup | 0.834 | 0.877 | 121.42 | 0.860 | 0.881 | 61.78 | ||||||||
I2CS-B1 | EfficientNet-B1 | FullSup | 0.764 | 0.851 | 165.74 | 0.812 | 0.867 | 69.89 |
dong-ma-zi / glandsegbenchmarks Goto Github PK
View Code? Open in Web Editor NEWBaseline methods for gland segmantation