Classification on VisDA-C (test)
91.71Accuracy (3D trainset)WATT-S
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| WATT-SVisual Encoder=ViT-L/142024.06 | 91.71 | 86.8 | 89.26 | |
| TENTVisual Encoder=ViT-L/142024.06 | 91.4 | 85.77 | 88.59 | |
| CLIPARTTVisual Encoder=ViT-L/142024.06 | 91.34 | 85.61 | 88.48 | |
| CLIPVisual Encoder=ViT-L/142024.06 | 91.24 | 85.62 | 88.43 | |
| TPTVisual Encoder=ViT-L/142024.06 | 90.65 | 85.41 | 88.03 | |
| WATT-SVisual Encoder=ViT-B/162024.06 | 87.72 | 86.75 | 87.24 | |
| WATT-PVisual Encoder=ViT-B/162024.06 | 87.61 | 86.66 | 87.14 | |
| CLIPARTTVisual Encoder=ViT-B/162024.06 | 87.58 | 86.6 | 87.09 | |
| TENTVisual Encoder=ViT-B/162024.06 | 87.57 | 86.81 | 87.19 | |
| CLIPVisual Encoder=ViT-B/162024.06 | 87.16 | 86.61 | 86.89 | |
| TPTVisual Encoder=ViT-B/162024.06 | 84.04 | 85.9 | 84.97 |