Image Classification on ImageNet (val) (Adversarial Robustness)
38.41Robust AccuracyTTC
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| TTCMode=Test-time Defence, Attack=10-step PGD, epsilon=1/2552025.03 | 38.41 | 49.39 | — | — | — | 37.26 | |
| TTEMode=Test-time Defence, Attack=10-step PGD, epsilon=1/2552025.03 | 31.01 | 60.02 | — | — | — | — | |
| PMG-AFTMode=Adversarial Finetuning, Attack=10-step PGD, epsilon=1/2552025.03 | 21.43 | 36.12 | — | — | — | — | |
| TeCoAMode=Adversarial Finetuning, Attack=10-step PGD, epsilon=1/2552025.03 | 18.89 | 34.89 | — | — | — | — | |
| FAREMode=Adversarial Finetuning, Attack=10-step PGD, epsilon=1/2552025.03 | 14 | 48.79 | — | — | — | — | |
| Anti-advMode=Test-time Defence, Attack=10-step PGD, epsilon=1/2552025.03 | 8.67 | 54.27 | — | — | — | — | |
| HDMode=Test-time Defence, Attack=10-step PGD, epsilon=1/2552025.03 | 6.63 | 54.54 | — | — | — | — | |
| RNMode=Test-time Defence, Attack=10-step PGD, epsilon=1/2552025.03 | 1.77 | 59.34 | — | — | — | — | |
| CLIPAttack=10-step PGD, epsilon=1/2552025.03 | 1.15 | 59.69 | — | — | — | — | |
| CLIP-FTMode=Adversarial Finetuning, Attack=10-step PGD, epsilon=1/2552025.03 | 0.93 | 54.24 | — | — | — | — |