Clustering on ImageNet-Dogs15 1.0 (val)
87.4ACCMAE-CT
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| MAE-CTBackbone=ViT-H/16, Augmentation Strategy=minimal2023.04 | 87.4 | 88.2 | 88.2 | 82.1 | |
| MAE-CTBackbone=ViT-L/16, Augmentation Strategy=extensive2023.04 | 84.7 | 86.5 | 86.4 | 79.1 | |
| MAE-CTBackbone=ViT-H/16, Augmentation Strategy=extensive2023.04 | 83.2 | 86 | 86 | 78.2 | |
| MAE-CTBackbone=ViT-L/16, Augmentation Strategy=minimal2023.04 | 79.9 | 83.4 | 83.4 | 72.7 | |
| MSNBackbone=ViT-B/16, Augmentation Strategy=extensive2023.04 | 79.7 | 82.4 | 82.3 | 72.2 | |
| MSNBackbone=ViT-L/16, Augmentation Strategy=extensive2023.04 | 79.5 | 82.9 | 82.9 | 72.5 | |
| iBOTBackbone=ViT-B/16, Augmentation Strategy=extensive2023.04 | 75.2 | 78.6 | 78.6 | 66.9 | |
| MAE-CTBackbone=ViT-B/16, Augmentation Strategy=extensive2023.04 | 71.7 | 74.7 | 74.7 | 60.3 | |
| DINOBackbone=ViT-B/16, Augmentation Strategy=extensive2023.04 | 71.4 | 75.2 | 75.2 | 62.2 | |
| MAE-CTBackbone=ViT-B/16, Augmentation Strategy=minimal2023.04 | 67.9 | 69.2 | 69.2 | 53.5 | |
| Moco v3Backbone=ViT-B/16, Augmentation Strategy=extensive2023.04 | 66.7 | 68.8 | 68.8 | 54.2 | |
| iBOTBackbone=ViT-L/16, Augmentation Strategy=extensive2023.04 | 64.8 | 70.5 | 70.4 | 55.4 | |
| MAEBackbone=ViT-L/16, Augmentation Strategy=minimal2023.04 | 18.5 | 12.3 | 12.1 | 5.2 | |
| MAEBackbone=ViT-H/16, Augmentation Strategy=minimal2023.04 | 17.6 | 11.7 | 11.5 | 4.8 | |
| MAEBackbone=ViT-B/16, Augmentation Strategy=minimal2023.04 | 15.7 | 8.5 | 8.3 | 3.3 |