Image Classification on CIFAR-10 symmetric noise
96.1Accuracy (20% Noise, Best)DivideMix
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| DivideMix2020.02 | 96.1 | 95.7 | 94.6 | 94.4 | 93.2 | 92.9 | 76 | 75.4 | — | |
| Mixup2020.02 | 95.6 | 92.3 | 87.1 | 77.6 | 71.6 | 46.7 | 52.2 | 43.9 | — | |
| M-correction2020.02 | 94 | 93.8 | 92 | 91.9 | 86.8 | 86.6 | 69.1 | 68.7 | — | |
| Meta-Learningre-implementation=true2020.02 | 92.9 | 92 | 89.3 | 88.8 | 77.4 | 76.1 | 58.7 | 58.3 | — | |
| P-correctionre-implementation=true2020.02 | 92.4 | 92 | 89.1 | 88.7 | 77.5 | 76.5 | 58.9 | 58.2 | — | |
| Co-teaching+re-implementation=true2020.02 | 89.5 | 88.2 | 85.7 | 84.1 | 67.4 | 45.5 | 47.9 | 30.1 | — | |
| Cross-Entropy2020.02 | 86.8 | 82.7 | 79.4 | 57.9 | 62.9 | 26.1 | 42.7 | 16.8 | — | |
| Bootstrap2020.02 | 86.8 | 82.9 | 79.8 | 58.4 | 63.3 | 26.8 | 42.9 | 17 | — | |
| F-correction2020.02 | 86.8 | 83.1 | 79.8 | 59.4 | 63.3 | 26.2 | 42.9 | 18.8 | — | |
| MLCTime (ms)=2652021.04 | — | — | — | — | — | — | — | — | 84.28 | |
| MLC + FaMUSTime (ms)=842021.04 | — | — | — | — | — | — | — | — | 85 |