Image Classification on CIFAR100 (IM=10) LT (test)
64.5Top-1 AccGLAG
Evaluation results
| Method | Links | |
|---|---|---|
| GLAGTraining strategy=Data augmentation2022.02 | 64.5 | |
| SADEBackbone=ResNet-32, Imbalance Ratio=102021.07 | 63.6 | |
| MISLASTraining strategy=Decoupled training2022.02 | 63.2 | |
| Balanced SoftmaxTraining strategy=End-to-end training2022.02 | 63 | |
| MISLASBackbone=ResNet-32, Imbalance Ratio=102021.07 | 62.5 | |
| SSDTraining strategy=Data augmentation2022.02 | 62.3 | |
| RIDEBackbone=ResNet-32, Imbalance Ratio=102021.07 | 61.8 | |
| LADEBackbone=ResNet-32, Imbalance Ratio=102021.07 | 61.6 | |
| Remix-DRWTraining strategy=Data augmentation2022.02 | 61.3 | |
| MetaSAug-LDAMTraining strategy=Data augmentation2022.02 | 61.3 | |
| Balanced SoftmaxBackbone=ResNet-32, Imbalance Ratio=102021.07 | 61 | |
| BBNBackbone=ResNet-32, Imbalance Ratio=102021.07 | 59.8 | |
| CausalBackbone=ResNet-32, Imbalance Ratio=102021.07 | 59.4 | |
| BBNTraining strategy=Decoupled training2022.02 | 59.1 | |
| SoftmaxBackbone=ResNet-32, Imbalance Ratio=102021.07 | 59.1 | |
| LDAM-DRWTraining strategy=End-to-end training2022.02 | 58.7 | |
| DecoupleTraining strategy=Decoupled training2022.02 | 57.9 | |
| M2mTraining strategy=Data augmentation2022.02 | 57.6 | |
| Cross-EntropyTraining strategy=End-to-end training2022.02 | 55.8 | |
| Focal LossTraining strategy=End-to-end training2022.02 | 55.8 |