Image Classification on CIFAR-10 Deepconfuse (test)
90.29Test AccuracyAVATAR
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| AVATARdiffusion timestep=1002024.06 | 90.29 | — | — | |
| nonlinear transformation framework2024.06 | 87 | — | — | |
| Nonlinear transformation frameworkBackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 86.93 | 11.65 | 15.48 | |
| Adversarial trainingBackbone=VGG19, Pre-trained=True (ImageNet), Defense Method=PGD2024.06 | 84.92 | — | — | |
| OPABackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 75.28 | — | — | |
| No defenseBackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 29 | — | — | |
| MixUp2024.06 | 26.65 | — | — | |
| CutMix2024.06 | 17.31 | — | — | |
| CutOut2024.06 | 14.86 | — | — |