Image Classification on DTD (Adversarial Robustness)
40.64Clean AccuracyCLIP
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| CLIP2025.12 | 40.64 | 2.22 | 2.55 | 0.19 | |
| PMG-AFTFine-tuning dataset=TinyImageNet, Adversarial training=2-step PGD, Epsilon=1/2552025.12 | 32.45 | 20.05 | 12.55 | 3.13 | |
| UCATFine-tuning dataset=TinyImageNet, Adversarial training=2-step PGD, Epsilon=1/2552025.12 | 31.92 | 19.47 | 17.87 | 18.14 | |
| TGA-ZSRFine-tuning dataset=TinyImageNet, Adversarial training=2-step PGD, Epsilon=1/2552025.12 | 25.69 | 17.82 | 16.97 | 17.02 | |
| TeCoAFine-tuning dataset=TinyImageNet, Adversarial training=2-step PGD, Epsilon=1/2552025.12 | 24.95 | 18.3 | 17.23 | 17.29 | |
| FAREFine-tuning dataset=TinyImageNet, Adversarial training=2-step PGD, Epsilon=1/2552025.12 | 23.09 | 5.75 | 4.52 | 1.92 |