ResearchBenchmarksRobust Image Classification on CIFAR-10 (test) (Input Noise Levels)Robust Image Classification on CIFAR-10 (test) (Input Noise Levels)Follow95.9Accuracy (10% Noise)TCL+93.71694.28394.8595.417Mar 13, 2023Evaluation resultsMethodMethodLinksAccuracy (10% Noise)Accuracy (20% Noise)Accuracy (30% Noise)TCL+Label Noise Type=asymm...Label Noise Type=asymmetric2023.0395.995.394.8Sel-CL+Label Noise Type=asymm...Label Noise Type=asymmetric2023.0395.695.294.5TCLLabel Noise Type=asymm...Label Noise Type=asymmetric2023.0395.194.794.4ELRLabel Noise Type=asymm...Label Noise Type=asymmetric2023.0394.493.391.5DivideMixLabel Noise Type=asymm...Label Noise Type=asymmetric2023.0393.893.292.5