Image Classification on CIFAR-100 Long-Tailed (rho=50) (test)
60.5Top-1 AccuracyDeiT-LT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DeiT-LTBackbone=ViT-B, Teacher=PaCo+SAM2024.04 | 60.5 | — | |
| PaCo+SAMBackbone=ResNet322024.04 | 56.6 | — | |
| DeiT-LTBackbone=ViT-B, Teacher=L-D-SAM2024.04 | 54.1 | — | |
| GCLBackbone=ResNet322024.04 | 53.6 | — | |
| MISLASBackbone=ResNet322024.04 | 52.3 | — | |
| Hybrid-SCBackbone=ResNet322024.04 | 51.9 | — | |
| ACEBackbone=ResNet322024.04 | 51.9 | — | |
| BiKD2026.05 | 51.84 | — | |
| CAMBackbone=ResNet322024.04 | 51.7 | — | |
| Teacher(CE)2026.05 | 51.59 | — | |
| DiVE2026.05 | 51.01 | — | |
| BKD2026.05 | 50.71 | — | |
| SSDBackbone=ResNet322024.04 | 50.5 | — | |
| L-D-SAMBackbone=ResNet322024.04 | 49.4 | — | |
| LFME2026.05 | 49.29 | — | |
| LDAM+DAPBackbone=ResNet322024.04 | 49.2 | — | |
| CCKD2026.05 | 48.86 | — | |
| Student(CE)2026.05 | 48.14 | — | |
| KD2026.05 | 47.87 | — | |
| RWKD2026.05 | 47.26 | — | |
| WSL2026.05 | 47.2 | — | |
| BBNBackbone=ResNet322024.04 | 47 | — | |
| CB Focal lossBackbone=ResNet322024.04 | 46.2 | — | |
| LDAM+DRWBackbone=ResNet322024.04 | 45.1 | — | |
| DeiT-IIIBackbone=ViT-B2024.04 | 44.1 | — | |
| ViT (cRT)Backbone=ViT-B2024.04 | 42.2 | — | |
| DeiTBackbone=ViT-B2024.04 | 39.1 | — | |
| ViTBackbone=ViT-B2024.04 | 39 | — | |
| A-EncoderBackbone=ResNet-32, Learning Protocol=ERM, Imbalance Ratio (rho)=502021.06 | — | 54.91 | |
| A-EncoderBackbone=ResNet-32, Learning Protocol=ERM, Imbalance Ratio (rho)=502021.06 | — | 57.85 | |
| ERMBackbone=ResNet-32, RSG=false2021.06 | — | 56.15 | |
| ERMBackbone=ResNet-32, RSG=true2021.06 | — | 54.44 | |
| ERM-RSGBackbone=ResNet-32, Learning Protocol=ERM, Imbalance Ratio (rho)=502021.06 | — | 54.44 | |
| ERM-RSGBackbone=ResNet-32, Learning Protocol=ERM, Imbalance Ratio (rho)=502021.06 | — | 56.82 | |
| Focal LossBackbone=ResNet-32, RSG=false2021.06 | — | 55.68 | |
| Focal LossBackbone=ResNet-32, RSG=true2021.06 | — | 54.85 | |
| FTLBackbone=ResNet-32, Learning Protocol=ERM, Imbalance Ratio (rho)=502021.06 | — | 55.24 | |
| FTLBackbone=ResNet-32, Learning Protocol=ERM, Imbalance Ratio (rho)=502021.06 | — | 58.03 | |
| ImaginaryBackbone=ResNet-32, Learning Protocol=ERM, Imbalance Ratio (rho)=502021.06 | — | 55.08 | |
| ImaginaryBackbone=ResNet-32, Learning Protocol=ERM, Imbalance Ratio (rho)=502021.06 | — | 58.46 | |
| LDAM-DRWBackbone=ResNet-32, RSG=false2021.06 | — | 53.38 | |
| LDAM-DRWBackbone=ResNet-32, RSG=true2021.06 | — | 51.5 | |
| M-DRWBackbone=ResNet-32, RSG=false2021.06 | — | 56.06 | |
| M-DRWBackbone=ResNet-32, RSG=true2021.06 | — | 55.3 |