Image Classification on CIFAR-10 Class imbalance Standard
56.54AccuracyREFINE
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| REFINEBackbone=CNNs2025.05 | 56.54 | 0.9103 | 0.5619 | 31.58 | |
| NoTransBackbone=CNNs2025.05 | 56.44 | 0.9055 | 0.5599 | 32.8 | |
| DistillBackbone=CNNs2025.05 | 54.89 | 0.9063 | 0.5492 | 41.96 | |
| LoRABackbone=CNNs2025.05 | 53.21 | 0.8975 | 0.5338 | 33.76 | |
| LinearProbeBackbone=CNNs2025.05 | 53.15 | 0.8883 | 0.5238 | 28.36 | |
| DANN-GateBackbone=CNNs2025.05 | 53.05 | 0.8964 | 0.5281 | 32.62 | |
| AdapterBackbone=CNNs2025.05 | 51.64 | 0.896 | 0.513 | 19.52 |