Decision Pattern Similarity on Frequency Preference Examination N=21
84.75Accuracy Map Similarity (Γ)FAN-B-AdaNCA
Evaluation Results
| Method | Links | |
|---|---|---|
| FAN-B-AdaNCABackbone=FAN-B, Training Strategy=AdaNCA2024.06 | 84.75 | |
| FAN-B-HybridBackbone=FAN-B, Training Strategy=Hybrid2024.06 | 83.52 | |
| RVT-B-AdaNCABackbone=RVT-B, Training Strategy=AdaNCA2024.06 | 82.17 | |
| TAPADL-RVTBackbone=RVT, Training Strategy=TAPADL2024.06 | 81.91 | |
| ConViT-B-AdaNCABackbone=ConViT-B, Training Strategy=AdaNCA2024.06 | 81.56 | |
| Swin-B-AdaNCABackbone=Swin-B, Training Strategy=AdaNCA2024.06 | 81.34 | |
| ConViT-BBackbone=ConViT-B, Training Strategy=Standard2024.06 | 79.69 | |
| Swin-BBackbone=Swin-B, Training Strategy=Standard2024.06 | 77.07 | |
| ResNet50Backbone=ResNet50, Training Strategy=Standard2024.06 | 72.23 | |
| ResNet50-AdvBackbone=ResNet50, Training Strategy=L2-adversarial training2024.06 | 52.12 |