Image Classification on CIFAR-10 (test) (Standard and Robust Accuracy)
95.19Standard AccuracyBase Classifier
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Base ClassifierThreat Model=l2(ϵ = 0.5)2023.07 | 95.19 | 0.3 | |
| ScoreOpt-NThreat Model=l2(ϵ = 0.5)2023.07 | 93.1 | 92.52 | |
| ScoreOpt-OThreat Model=l2(ϵ = 0.5)2023.07 | 91.85 | 90.55 | |
| Gowal et al.Threat Model=l2(ϵ = 0.5)2023.07 | 90.9 | 74.5 | |
| Rony et al.Threat Model=l2(ϵ = 0.5)2023.07 | 89.05 | 67.6 | |
| Rice et al.Threat Model=l2(ϵ = 0.5)2023.07 | 88.67 | 71.6 | |
| Wu et al.Threat Model=l2(ϵ = 0.5)2023.07 | 88.51 | 73.66 | |
| Ding et al.Threat Model=l2(ϵ = 0.5)2023.07 | 88.02 | 66.18 |