Image Classification on CIFAR-10-LT (Top-1 Accuracy by IF)
86.3Top-1 AccuracyOurs
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| OursBackbone=ResNet-322026.03 | 86.3 | — | — | — | |
| FeatReconBackbone=ResNet-322026.03 | 85.2 | — | — | — | |
| GBGBackbone=ResNet-322026.03 | 85.1 | — | — | — | |
| DiffuLT + BBNBackbone=ResNet-322026.03 | 85 | — | — | — | |
| DiffuLTBackbone=ResNet-322026.03 | 84.7 | — | — | — | |
| Meta-CALABackbone=ResNet-322026.03 | 84.7 | — | — | — | |
| SELBackbone=ResNet-322026.03 | 84.4 | — | — | — | |
| BCLBackbone=ResNet-322026.03 | 84.3 | — | — | — | |
| Heuristic-CALABackbone=ResNet-322026.03 | 83.9 | — | — | — | |
| ADRWBackbone=ResNet-322026.03 | 83.6 | — | — | — | |
| SAMBackbone=ResNet-322026.03 | 81.9 | — | — | — | |
| BBNBackbone=ResNet-322026.03 | 79.8 | — | — | — | |
| LDAM-DRWBackbone=ResNet-322026.03 | 77 | — | — | — | |
| cRTBackbone=ResNet-322026.03 | 75.7 | — | — | — | |
| CAM-BSBackbone=ResNet-322026.03 | 75.4 | — | — | — | |
| CEBackbone=ResNet-322026.03 | 70.4 | — | — | — | |
| Focal LossBackbone=ResNet-322026.03 | 70.4 | — | — | — | |
| BBNBackbone=ResNet-32, Method Category=rebalance classifier2023.05 | — | 79.82 | 82.18 | 88.32 | |
| BCLBackbone=ResNet-32, Method Category=self-supervised pretraining2023.05 | — | 84.32 | 87.24 | 91.12 | |
| CB-FocalBackbone=ResNet-32, Method Category=rebalance classifier2023.05 | — | 74.6 | 79.3 | 87.1 | |
| CEBackbone=ResNet-322023.05 | — | 70.4 | 74.8 | 86.4 | |
| GLMCBackbone=ResNet-32, Training Stage=one-stage training2023.05 | — | 87.75 | 90.18 | 94.04 | |
| GLMC + MaxNormBackbone=ResNet-32, Training Stage=finetune classifier2023.05 | — | 87.57 | 90.22 | 94.03 | |
| KCLBackbone=ResNet-32, Method Category=self-supervised pretraining2023.05 | — | 77.6 | 81.7 | 88 | |
| LogitAjustBackbone=ResNet-32, Method Category=rebalance classifier2023.05 | — | 80.92 | — | — | |
| MixupBackbone=ResNet-32, Method Category=augmentation2023.05 | — | 73.06 | 77.82 | 87.1 | |
| RISDABackbone=ResNet-32, Method Category=augmentation2023.05 | — | 79.89 | 79.89 | 79.89 | |
| TSCBackbone=ResNet-32, Method Category=self-supervised pretraining2023.05 | — | 79.7 | 82.9 | 88.7 |