Image Classification on CIFAR-100 (test) (Asymmetric Noise Robustness)
68.85Test Acc (asym-20%)Co-learning+HCR
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Co-learning+HCRBackbone=ResNet-18, Number of trials=5, Evaluation window=last 10 epochs2022.06 | 68.85 | 61.94 | 50.29 | |
| Co-learningBackbone=ResNet-18, Number of trials=5, Evaluation window=last 10 epochs2022.06 | 65.26 | 56.97 | 47.62 | |
| Co-teachingBackbone=ResNet-18, Number of trials=5, Evaluation window=last 10 epochs2022.06 | 59.76 | 49.53 | 40.62 | |
| Standard CEBackbone=ResNet-18, Number of trials=5, Evaluation window=last 10 epochs2022.06 | 59.36 | 51.06 | 42.49 | |
| JoCoRBackbone=ResNet-18, Number of trials=5, Evaluation window=last 10 epochs2022.06 | 58.58 | 49.04 | 39.72 | |
| DecouplingBackbone=ResNet-18, Number of trials=5, Evaluation window=last 10 epochs2022.06 | 57.97 | 49.86 | 41.51 | |
| Co-teaching+Backbone=ResNet-18, Number of trials=5, Evaluation window=last 10 epochs2022.06 | 56.11 | 47.12 | 38.98 |