Image Classification on Red Mini-ImageNet (test) (Noise Robustness)
61.44Accuracy (Noise 0.2)Ours (CLIPCleaner + MixFix)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Ours (CLIPCleaner + MixFix)2024.08 | 61.44 | 58.42 | 53.18 | 43.82 | |
| GNL - p(Y|X) ensembleEnsemble=true2023.08 | 58.46 | 53.38 | 48.92 | 40.68 | |
| InstanceGM2024.08 | 58.38 | 52.24 | 47.96 | 39.62 | |
| GNL - p(X|Y) ensembleEnsemble=true2023.08 | 57.56 | 52.68 | 47.12 | 39.54 | |
| GNL - p(Y|X)Ensemble=false2023.08 | 54.53 | 50.23 | 45.12 | 35.78 | |
| GNL - p(X|Y)Ensemble=false2023.08 | 53.34 | 49.56 | 44.08 | 34.7 | |
| FaMUSEnsemble=false2023.08 | 51.42 | 48.06 | 45.1 | 35.5 | |
| FaMUS2024.08 | 51.42 | 48.06 | 45.1 | 35.5 | |
| MentorMixEnsemble=false2023.08 | 51.02 | 47.14 | 43.8 | 33.46 | |
| MentorMix2024.08 | 51.02 | 47.14 | 43.8 | 33.46 | |
| DivideMixEnsemble=false2023.08 | 50.96 | 46.72 | 43.14 | 34.5 | |
| DivideMix2024.08 | 50.96 | 46.72 | 43.14 | 34.5 | |
| MixupEnsemble=false2023.08 | 49.1 | 46.4 | 40.58 | 33.58 | |
| Mixup2024.08 | 49.1 | 46.4 | 40.58 | 33.58 | |
| CEEnsemble=false2023.08 | 47.36 | 42.7 | 37.3 | 29.76 | |
| CE2024.08 | 47.36 | 42.7 | 37.3 | 29.76 |