Image Classification on CIFAR-10 rho=100 long-tailed (test)
87.5AccuracyDeiT-LT
Evaluation results
| Method | Links | |
|---|---|---|
| DeiT-LTBackbone=ViT-B, Teacher=PaCo+SAM2024.04 | 87.5 | |
| PaCo+SAMBackbone=ResNet322024.04 | 86.8 | |
| DeiT-LTBackbone=ViT-B, Teacher=L-D-SAM2024.04 | 84.8 | |
| GCLBackbone=ResNet322024.04 | 82.7 | |
| VS+SAMBackbone=ResNet322024.04 | 82.4 | |
| MISLASBackbone=ResNet322024.04 | 82.1 | |
| L-D-SAMBackbone=ResNet322024.04 | 81.9 | |
| Hybrid-SCBackbone=ResNet322024.04 | 81.4 | |
| ACEBackbone=ResNet322024.04 | 81.4 | |
| LDAM+DAPBackbone=ResNet322024.04 | 80 | |
| CAMBackbone=ResNet322024.04 | 80 | |
| BBNBackbone=ResNet322024.04 | 79.8 | |
| VSBackbone=ResNet322024.04 | 78.6 | |
| Log. Adj.Backbone=ResNet322024.04 | 77.7 | |
| LDAM+DRWBackbone=ResNet322024.04 | 77 | |
| CB Focal lossBackbone=ResNet322024.04 | 74.6 | |
| DeiTBackbone=ViT-B2024.04 | 70.2 | |
| ViT (cRT)Backbone=ViT-B2024.04 | 68.9 | |
| ViTBackbone=ViT-B2024.04 | 62.6 | |
| DeiT-IIIBackbone=ViT-B2024.04 | 59.1 |