Image Classification on Tiny ImageNet (Adversarial Robustness Evaluation)
11.9Accuracy (FGSM)Poincare ResNet-32
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Poincare ResNet-32Model=PRN 32, ε=2.4/255 (ε1)2025.11 | 11.9 | 9.56 | 7.03 | 5.74 | — | |
| Poincare ResNet-20Model=PRN 20, ε=2.4/255 (ε1)2025.11 | 11.73 | 8.9 | 7.31 | 5.93 | — | |
| Poincare ResNet-32Model=PRN 32, ε=3.2/255 (ε2)2025.11 | 10.71 | 8.23 | 6.61 | 5.49 | 30.46 | |
| Poincare ResNet-20Model=PRN 20, ε=3.2/255 (ε2)2025.11 | 10.5 | 7.78 | 6.62 | 5.49 | 30.48 | |
| Poincare ResNet-32Model=PRN 32, ε=8.0/255 (ε3)2025.11 | 8.02 | 5.57 | 5.69 | 5 | — | |
| Poincare ResNet-20Model=PRN 20, ε=8.0/255 (ε3)2025.11 | 7.44 | 5.43 | 5.66 | 4.63 | — |