Image Classification on CIFAR-10 Error-maximizing (test)
92.2Test Accuracynonlinear transformation framework
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| nonlinear transformation framework2024.06 | 92.2 | — | — | |
| AVATARdiffusion timestep=1002024.06 | 91.92 | — | — | |
| Nonlinear transformation frameworkBackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 90.75 | 15.18 | 20.09 | |
| Adversarial trainingBackbone=VGG19, Pre-trained=True (ImageNet), Defense Method=PGD2024.06 | 85.36 | — | — | |
| OPABackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 75.57 | — | — | |
| MixUp2024.06 | 53.28 | — | — | |
| CutMix2024.06 | 30.31 | — | — | |
| CutOut2024.06 | 28.3 | — | — | |
| No defenseBackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 6 | — | — |