Image Classification on CIFAR10 (test) (Symmetric/Asymmetric Robustness)
91.15Top-1 Acc (Sym, 0.2)ABSGD(SCE)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| ABSGD(SCE)Backbone=ResNet-50, Pre-trained=ImageNet, Optimizer=ABSGD, Loss Function=SCE2020.12 | 91.15 | 88.65 | 91.04 | 86.1 | |
| CIW(SCE)Backbone=ResNet-50, Pre-trained=ImageNet, Optimizer=CIW, Loss Function=SCE2020.12 | 90.79 | 88.21 | 89.97 | 85.13 | |
| ABSGD(CE)Backbone=ResNet-50, Pre-trained=ImageNet, Optimizer=ABSGD, Loss Function=CE2020.12 | 90.64 | 88.31 | 90.15 | 87.84 | |
| CIW(CE)Backbone=ResNet-50, Pre-trained=ImageNet, Optimizer=CIW, Loss Function=CE2020.12 | 90.29 | 87.92 | 88.99 | 87.18 | |
| EG(SCE)Backbone=ResNet-50, Pre-trained=ImageNet, Optimizer=EG, Loss Function=SCE2020.12 | 90.25 | 88.13 | 90.55 | 84.47 | |
| SGD(SCE)Backbone=ResNet-50, Pre-trained=ImageNet, Optimizer=SGD, Loss Function=SCE2020.12 | 90.05 | 87.83 | 90.25 | 81.91 | |
| SGD(TCE)Backbone=ResNet-50, Pre-trained=ImageNet, Optimizer=SGD, Loss Function=TCE2020.12 | 89.87 | 86.84 | 88.97 | 80.85 | |
| SGD(CE)Backbone=ResNet-50, Pre-trained=ImageNet, Optimizer=SGD, Loss Function=CE2020.12 | 88.59 | 85.75 | 86.62 | 80.81 |