ResearchBenchmarksImage Classification on CIFAR-10 (test) (Symmetric/Asymmetric Metrics)Image Classification on CIFAR-10 (test) (Symmetric/Asymmetric Metrics)Follow96.3Symmetric Accuracy (20%)SOP+86.83689.29391.7594.207Feb 28, 2022Evaluation resultsMethodMethodLinksSymmetric Accuracy (20%)Symmetric Accuracy (50%)Symmetric Accuracy (80%)Asymmetric Accuracy (40%)SOP+Backbone=ResNet34Backbone=ResNet342022.0296.395.59493.8DivideMixBackbone=ResNet34Backbone=ResNet342022.0296.194.693.293.4ELR+Backbone=ResNet34Backbone=ResNet342022.0295.894.893.393MixUpBackbone=ResNet34Backbone=ResNet342022.0293.587.972.3—CEBackbone=ResNet34Backbone=ResNet342022.0287.280.765.882.2