Image Classification on CIFAR-10 OPS (test)
88.1Test AccuracyOPA
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| OPABackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 88.1 | — | — | |
| Nonlinear transformation frameworkBackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 87.76 | -0.34 | -0.39 | |
| No defenseBackbone=VGG19, Pre-trained=True (ImageNet)2024.06 | 15 | — | — | |
| Adversarial trainingBackbone=VGG19, Pre-trained=True (ImageNet), Defense Method=PGD2024.06 | 13.01 | — | — |