Robust Image Classification on CIFAR-10 (val) (Noise Levels 20%-80%)
95.6Acc (20% noise)MentorMix
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| MentorMixBackbone=ResNet-322019.11 | 95.6 | 94.2 | 91.3 | 81 | |
| M-DYR-HArchitecture=PRN-182019.04 | 94 | 92.8 | 90.3 | 46.3 | |
| Arazo et al.2019.11 | 94 | 92.8 | 90.3 | 74.1 | |
| Mixup2019.11 | 94 | 91.5 | 86.8 | 76.9 | |
| MD-DYR-SHArchitecture=PRN-182019.04 | 93.8 | 92.3 | 86.1 | 74.1 | |
| Huang et al.2019.11 | 92.6 | 90.3 | 43.4 | — | |
| Jiang et al. (2018b)Architecture=WRN-1012019.04 | 92 | 89 | — | 49 | |
| MentorNet2019.11 | 92 | 91.2 | 74.2 | 60 | |
| Zhang & Sabuncu2019.11 | 89.7 | 87.6 | 82.7 | 67.9 | |
| Chen et al.2019.11 | 89.7 | — | — | 52.3 | |
| Lee et al.2019.11 | 87.1 | 81.8 | 75.4 | — | |
| Ma et al. (2018)Architecture=GCNN-122019.04 | 85.1 | 83.4 | 72.8 | — | |
| Wang et al. (2018b)Architecture=GCNN-72019.04 | 81.4 | 78.2 | — | — | |
| Ren et al. (2018)Architecture=WRN-282019.04 | — | 86.9 | — | — |