Image Classification on CIFAR-100 Long-tailed Imbalance 10 (train)
58.74Top-1 AccuracyCDB-CE
Evaluation results
| Method | Links | |
|---|---|---|
| CDB-CEBackbone=ResNet-32, tau=dynamically updated2020.10 | 58.74 | |
| LDAM-DRWBackbone=ResNet-322020.10 | 58.71 | |
| Meta-Weight NetBackbone=ResNet-322020.10 | 58.46 | |
| Equalization lossBackbone=ResNet-32, Data Augmentation=Without additional augmentation2020.10 | 58.32 | |
| Class-BalancedBackbone=ResNet-322020.10 | 57.99 | |
| Focal lossBackbone=ResNet-322020.10 | 55.78 | |
| L2RWBackbone=ResNet-322020.10 | 53.73 |