Image Classification on CIFAR-100 (test) (Symmetric/Asymmetric Noise Robustness)
79.27Accuracy (Clean)GJS
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| GJSBackbone=ResNet-342021.05 | 79.27 | 78.05 | 75.71 | 70.15 | 31.4 | 74.6 | 63.7 | |
| LSBackbone=ResNet-342021.05 | 78.6 | 74.88 | 68.41 | 54.58 | 26.98 | 73.17 | 57.2 | |
| SCEBackbone=ResNet-342021.05 | 78.29 | 74.21 | 68.23 | 59.28 | 26.8 | 70.86 | 51.12 | |
| JSBackbone=ResNet-342021.05 | 77.95 | 75.41 | 71.12 | 64.36 | 15.65 | 71.7 | 49.36 | |
| BSBackbone=ResNet-342021.05 | 77.65 | 72.92 | 68.52 | 53.8 | 13.83 | 73.79 | 64.67 | |
| GCEBackbone=ResNet-342021.05 | 77.65 | 75.02 | 71.54 | 65.21 | 49.68 | 72.13 | 51.5 | |
| CEBackbone=ResNet-342021.05 | 77.6 | 65.74 | 55.77 | 44.42 | 10.74 | 66.85 | 49.45 | |
| NCE+RCEBackbone=ResNet-342021.05 | 74.66 | 72.39 | 68.79 | 62.18 | 31.63 | 71.35 | 57.8 |