Image Classification on CIFAR-100 (test) (0%, 40%, 80% Noise Accuracy)
77.44Accuracy (40% Noise)RTE (PBT)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RTE (PBT)#Params=13.1M, Configuration=Population Based Training2021.09 | 77.44 | — | 66.43 | |
| RTE (Manual)#Params=13.1M, Configuration=Manual2021.09 | 76.7 | 79.71 | 64.02 | |
| CE+JSD+ECR#Params=13.1M2021.09 | 71.89 | — | 40.43 | |
| GCE-only#Params=13.1M2021.09 | 69.73 | — | 39.19 | |
| CE-only#Params=13.1M2021.09 | 65.98 | — | 35.8 | |
| DivideMix#Params=11.2M2021.09 | 60.2 | 75.2 | — | |
| Reed-Hard2021.09 | 51.34 | — | — | |
| S-Model2021.09 | 49.1 | — | — | |
| M-DYR-H#Params=11.2M2021.09 | 48.2 | — | — | |
| Meta-Learning2021.09 | 42.4 | — | — | |
| MixUp2021.09 | 30.8 | — | — | |
| GCE#Params=21.8M2021.09 | 29.16 | 61.77 | — | |
| Co-Teaching2021.09 | 23.22 | 55.95 | — | |
| Forward T2021.09 | 8.9 | 31.05 | — | |
| Bi-Level#Params=11.2M2021.09 | — | 61.6 | — | |
| MentorNet PD#Params=84M2021.09 | — | 56 | 14 | |
| PENCIL#Params=21.8M2021.09 | — | 69.12 | — | |
| Rand Weights#Params=36.4M2021.09 | — | 58.01 | — | |
| RoG2021.09 | — | 55.68 | — | |
| SELF#Params=25.0M2021.09 | — | 71.98 | 42.09 | |
| SELFIE2021.09 | — | 62.9 | — |