Breast Cancer Metastases Classification on Camelyon17-WILDS (test)
96.85Center 1 AccuracyVL2V* + DIPT
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| VL2V* + DIPTBackbone=ViT-based student2025.11 | 96.85 | 96.86 | 96.45 | 96.4 | 94.47 | 94.18 | 93.66 | 93.28 | |
| VL2V*Backbone=ViT-based student2025.11 | 96.32 | 96.31 | 93.81 | 93.49 | 95.24 | 95.06 | 88.66 | 87.23 | |
| RISE + DIPTBackbone=ResNet-502025.11 | 95.98 | 95.87 | 91.23 | 90.51 | 94.13 | 93.84 | 83.37 | 80.14 | |
| KDBackbone=ResNet-502025.11 | 93.94 | 93.61 | 88.57 | 87.39 | 91.42 | 90.75 | 83.32 | 80.2 | |
| VL2V + DIPTBackbone=ResNet-502025.11 | 93.85 | 93.46 | 93.87 | 93.55 | 90.9 | 89.99 | 93.85 | 93.46 | |
| VL2VBackbone=ResNet-502025.11 | 93.24 | 92.76 | 90.43 | 89.47 | 90.99 | 90.11 | 87.42 | 85.78 | |
| RISEBackbone=ResNet-502025.11 | 90.16 | 89.16 | 93.07 | 92.85 | 90.51 | 89.58 | 81.67 | 77.75 | |
| Agg. prompt Zero-ShotBackbone=ResNet-50, Evaluation Protocol=Zero-Shot2025.11 | 81.89 | 81.79 | 89.38 | 89.64 | 77.63 | 77.27 | 75.58 | 79.98 |