Image Classification on CIFAR10 12/255 noise (test)
81.09Natural AccuracyMultiGrad
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MultiGradBackbone=PreActResNet-182024.04 | 81.09 | 9.95 | |
| PGD-2Backbone=PreActResNet-182024.04 | 78.97 | 32.99 | |
| ZeroGradBackbone=PreActResNet-182024.04 | 77.52 | 27.34 | |
| RS-AAERBackbone=PreActResNet-182024.04 | 74.4 | 32.17 | |
| Grad AlignBackbone=PreActResNet-182024.04 | 74.17 | 34.87 | |
| PGD-10(20)Backbone=PreActResNet-182024.04 | 72.37 | 38.6 | |
| N-FGSMBackbone=PreActResNet-182024.04 | 71.3 | 36.23 | |
| N-AAERBackbone=PreActResNet-182024.04 | 71.15 | 36.52 | |
| FreeATBackbone=PreActResNet-18, Threat Model=L_infinity2024.04 | 68.07 | 33.14 | |
| RS-FGSMBackbone=PreActResNet-182024.04 | 66.46 | 0 |