Adversarial Robustness on CIFAR-10 (L2, eps=8/255)
77.64Robust Acc (AA)Stutz et al. (2020)
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Stutz et al. (2020)author_provided=true, epsilon=0.032021.02 | 77.64 | 39.54 | -38.1 | 101 | 108 | 0.94 | 296 | |
| Stutz et al. (2020)attack_sequence_length_A3=82021.02 | 77.64 | 26.87 | -50.77 | 101 | 205 | 0.49 | 659 | |
| Wu et al. (2020) RTS + Das et al. (2017) + Lee et al. (2018)non-differentiable=true2021.02 | 60.72 | 60.04 | -0.68 | 621 | 210 | 2.96 | 585 | |
| Wang et al. (2019)2021.02 | 50.84 | 50.81 | -0.03 | 734 | 372 | 1.97 | 755 | |
| Wu et al. (2020) RTS + Random Rotation2021.02 | 49.53 | 41.99 | -7.54 | 255 | 462 | 0.55 | 900 | |
| Zhang & Wang (2019)author_provided=true2021.02 | 36.74 | 37.11 | 0.37 | 381 | 302 | 1.26 | 726 | |
| Grathwohl et al. (2020) + Random Rotation2021.02 | 22.29 | 13.45 | -8.84 | 114 | 374 | 0.3 | 1,023 | |
| Grathwohl et al. (2020) + Das et al. (2017) + Lee et al. (2018)2021.02 | 15.27 | 5.24 | -10.03 | 261 | 79 | 3.3 | 746 | |
| Hu et al. (2019)2021.02 | 6.25 | 3.07 | -3.18 | 110 | 56 | 1.96 | 502 | |
| Xiao et al. (2020) ADVtraining=adversarial2021.02 | 5.4 | 2.31 | -3.09 | 95 | 146 | 0.65 | 828 | |
| Grathwohl et al. (2020)2021.02 | 5.15 | 5.16 | 0.01 | 107 | 114 | 0.94 | 749 |