Prostate cancer grading on SICAP target-imbalanced KD v2 (test)
91.75Overall AccuracyUMKD
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| UMKDMethod Type=KD, Expert Backbone=ResNet-50, Student Backbone=ResNet-182025.05 | 91.75 | 90.72 | 91.72 | 0.1199 | |
| KDMethod Type=KD, Expert Backbone=ResNet-50, Student Backbone=ResNet-182025.05 | 90.97 | 89.44 | 90.9 | 0.1368 | |
| RKDMethod Type=Feature-based KD, Expert Backbone=ResNet-50, Student Backbone=ResNet-182025.05 | 90.79 | 88.3 | 90.67 | 0.1369 | |
| Resnet18 (Stu)Method Type=Individually trained model, Backbone=ResNet-182025.05 | 90.36 | 88.11 | 90.23 | 0.1318 | |
| SDDMethod Type=KD, Expert Backbone=ResNet-50, Student Backbone=ResNet-182025.05 | 89.93 | 88.81 | 89.85 | 0.1447 | |
| Resnet50 (Exp2)Method Type=Individually trained model, Backbone=ResNet-502025.05 | 89.71 | 89.78 | 89.61 | 0.1322 | |
| Resnet50 (Exp1)Method Type=Individually trained model, Backbone=ResNet-502025.05 | 89.19 | 89.44 | 89.13 | 0.1332 | |
| DKDMethod Type=KD, Expert Backbone=ResNet-50, Student Backbone=ResNet-182025.05 | 89.19 | 87.11 | 89.12 | 0.1476 | |
| FitNetMethod Type=Feature-based KD, Expert Backbone=ResNet-50, Student Backbone=ResNet-182025.05 | 79.06 | 55.77 | 77.12 | 0.3694 |