Image Classification on CIFAR-10 (test) (Adversarial Robustness Accuracy)
87.97Natural AccuracyMT
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| MTBackbone=WRN-28-102025.03 | 87.97 | 43.56 | 1 | 0.78 | 0 | 0 | |
| PUATBackbone=WRN-28-102025.03 | 81.88 | 54.4 | 46.63 | 46.7 | 44.46 | 52.12 | |
| WSCATBackbone=WRN-28-102025.03 | 80.93 | 59.62 | 58.52 | 53.15 | 52.23 | 59.4 | |
| StandardBackbone=WRN-28-102025.03 | 78.85 | 4.37 | 0 | 0 | 0 | 0 | |
| RSTBackbone=WRN-28-102025.03 | 73.22 | 49.39 | 46.1 | 43.95 | 42.94 | 49.21 | |
| RLFATBackbone=WRN-28-102025.03 | 64.95 | 32.58 | 29.52 | 26.63 | 25.72 | 31.97 | |
| WSCAT-supBackbone=WRN-28-102025.03 | 62.82 | 36.12 | 34.53 | 30.6 | 30 | 36.08 | |
| DMATBackbone=WRN-28-102025.03 | 61.05 | 36.5 | 34.75 | 30.11 | 29.68 | 35.85 | |
| TRADESBackbone=WRN-28-102025.03 | 58.95 | 30.15 | 28.09 | 26.15 | 25.44 | 30.62 |