Unsupervised Object Localization on ImageNet 1K
63.7MaxBoxAccV2EP
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| EPBackbone=iBOT, Architecture=ViT-B/162025.06 | 63.7 | 6.1 | |
| EPBackbone=MAE, Architecture=ViT-L/162025.06 | 61.2 | 14.3 | |
| EPBackbone=BEiTv2, Architecture=ViT-B/162025.06 | 61.2 | 14.2 | |
| EPBackbone=MAE, Architecture=ViT-B/162025.06 | 60.4 | 6.2 | |
| EPBackbone=SimMIM, Architecture=ViT-B/162025.06 | 60 | 14.8 | |
| EPBackbone=MAE, Architecture=ViT-S/162025.06 | 58.5 | 12.5 | |
| LPBackbone=iBOT, Architecture=ViT-B/162025.06 | 57.6 | — | |
| LPBackbone=MAE, Architecture=ViT-B/162025.06 | 54.2 | — | |
| LPBackbone=BEiTv2, Architecture=ViT-B/162025.06 | 47 | — | |
| LPBackbone=MAE, Architecture=ViT-L/162025.06 | 46.9 | — | |
| LPBackbone=MAE, Architecture=ViT-S/162025.06 | 46 | — | |
| LPBackbone=SimMIM, Architecture=ViT-B/162025.06 | 45.2 | — | |
| EPBackbone=SigLIP, Architecture=ViT-L/162025.06 | 44.2 | 0.2 | |
| LPBackbone=SigLIP, Architecture=ViT-L/162025.06 | 44 | — |