Image Classification on ImageNet-1K (val) (Few-Shot and Probing Evaluation)
64.2Accuracy (1-shot)D2V2-Refined
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| D2V2-RefinedViT=H/142024.02 | 64.2 | 71.3 | 75.5 | 78.1 | 83.5 | 84.7 | 82.3 | |
| D2V2-RefinedViT=L/162024.02 | 61.7 | 69.6 | 73.9 | 78.1 | 82.1 | 83.5 | 81 | |
| MAE-RefinedViT=H/142024.02 | 59.5 | 68.5 | 73.8 | 77.4 | 82.1 | 83.7 | 82.3 | |
| MAE-RefinedViT=2B/142024.02 | 58.2 | 68.6 | 74.8 | 78.7 | 82.5 | 84.5 | 83.2 | |
| MAE-RefinedViT=L/162024.02 | 57.8 | 66.3 | 72 | 76.1 | 81.2 | 82.8 | 81.5 | |
| MugsViT=L/162024.02 | 52.9 | 62.3 | 69.4 | 76.2 | 80.3 | 82.1 | 80.4 | |
| MAE-CTViT=H/142024.02 | 49.4 | 59.6 | 67.4 | 74.4 | 81.7 | 82.3 | 79.1 | |
| iBOTViT=L/162024.02 | 48.5 | 58.2 | 66.5 | 73.3 | 79 | 81.1 | 78 | |
| D2V2ViT=L/162024.02 | 24.1 | 58.8 | 72.1 | 75.1 | 81.5 | 78.2 | 51.8 | |
| D2V2ViT=H/142024.02 | 21.6 | 60.8 | 74.2 | 77.6 | 83.3 | 80.4 | 48 | |
| MAEViT=2B/142024.02 | 17.8 | 29.1 | 62.9 | 73.6 | 82 | 79.7 | 67.1 | |
| MAEViT=L/162024.02 | 14.3 | 34.9 | 56.9 | 67.7 | 79.3 | 77.5 | 60.6 | |
| MAEViT=H/142024.02 | 7.2 | 14.1 | 40.2 | 72.8 | 81.2 | 78.2 | 58.1 |