WSI-level retrieval on BRACS (test)
59Macro Avg F1 ScoreKimiaNet
Evaluation Results
| Method | Links | |
|---|---|---|
| KimiaNetPre-training Data Domain=Histopathology, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 59 | |
| PathDino-512Pre-training Data Domain=Histopathology, Model Architecture Type=Transformer, Resolution=512, Majority Vote Top-k=MV@52023.11 | 59 | |
| DinoV1-ViT-b16Pre-training Data Domain=Natural, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 57 | |
| Barlow-Twins-ResNet50Pre-training Data Domain=Histopathology, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 57 | |
| PLIPPre-training Data Domain=Histopathology, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 57 | |
| iBOT-PathPre-training Data Domain=Histopathology, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 57 | |
| DinoSSLPathology-8Pre-training Data Domain=Histopathology, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 57 | |
| EfficientNet-b5Pre-training Data Domain=Natural, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 56 | |
| ConvNext-xlargePre-training Data Domain=Natural, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 56 | |
| DinoV1-ViT-s16Pre-training Data Domain=Natural, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 56 | |
| MoCoV2-ResNet50Pre-training Data Domain=Histopathology, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 56 | |
| PathDino-224Pre-training Data Domain=Histopathology, Model Architecture Type=Transformer, Resolution=224, Majority Vote Top-k=MV@52023.11 | 56 | |
| ConvNext-b-224Pre-training Data Domain=Natural, Model Architecture Type=CNN-based, Resolution=224, Majority Vote Top-k=MV@52023.11 | 55 | |
| CLIP - ViT-B/16Pre-training Data Domain=Natural, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 55 | |
| ResNet50Pre-training Data Domain=Natural, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 54 | |
| BiomedCLIPPre-training Data Domain=Histopathology, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 54 | |
| DenseNet121Pre-training Data Domain=Natural, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 52 | |
| MuDiPath-DenseNet-101Pre-training Data Domain=Histopathology, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 52 | |
| EfficientNet-b3-288Pre-training Data Domain=Natural, Model Architecture Type=CNN-based, Resolution=288, Majority Vote Top-k=MV@52023.11 | 51 | |
| ViT-b16-224Pre-training Data Domain=Natural, Model Architecture Type=Transformer, Resolution=224, Majority Vote Top-k=MV@52023.11 | 50 | |
| DinoV2-ViT-b14Pre-training Data Domain=Natural, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 49 | |
| MuDiPath-ResNet50Pre-training Data Domain=Histopathology, Model Architecture Type=CNN-based, Majority Vote Top-k=MV@52023.11 | 49 | |
| HIPT-ViT-s16Pre-training Data Domain=Histopathology, Model Architecture Type=Transformer, Majority Vote Top-k=MV@52023.11 | 48 |