Image Classification on CIFAR-10 (test) (Robustness Symm 0.2 & 0.5)
95.9Accuracy (Symm 0.2)CORES2
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CORES2Backbone=Pre-ResNet182020.10 | 95.9 | 94.5 | |
| DivideMixBackbone=Pre-ResNet182020.10 | 95.7 | 94.4 | |
| M-correctionBackbone=Pre-ResNet182020.10 | 93.8 | 91.9 | |
| MixupBackbone=Pre-ResNet182020.10 | 92.3 | 77.6 | |
| P-correctionBackbone=Pre-ResNet182020.10 | 92 | 88.7 | |
| Meta-LearningBackbone=Pre-ResNet182020.10 | 92 | 88.8 | |
| Co-teaching+Backbone=Pre-ResNet182020.10 | 88.2 | 84.1 | |
| Forward TBackbone=Pre-ResNet182020.10 | 83.1 | 59.4 | |
| BootstrapBackbone=Pre-ResNet182020.10 | 82.9 | 58.4 | |
| CEBackbone=Pre-ResNet182020.10 | 82.7 | 57.9 |