Learning with noisy labels 벤치마크
Learning with noisy labels on Red MiniImageNet 40% label noise
Test Accuracy
- 2022-02-04 — NCR (ResNet-18): Test Accuracy 64.6
| Rank | Model | Test Accuracy | Paper | Code | Year |
|---|---|---|---|---|---|
| 1 | NCR (ResNet-18) | 64.6 | Learning with Neighbor Consistency for Noisy Labels | google-research/scenic | 2022 |
| 2 | CLIPCleaner | 58.42 | CLIPCleaner: Cleaning Noisy Labels with CLIP | mrchenfeng/clipcleaner_acmmm2024 | 2024 |
| 3 | InstanceGM-SS | 56.37 | Instance-Dependent Noisy Label Learning via Graphical Modelling | arpit2412/InstanceGM | 2022 |
| 4 | InstanceGM | 52.24 | Instance-Dependent Noisy Label Learning via Graphical Modelling | arpit2412/InstanceGM | 2022 |
| 5 | NCR (ResNet-18) | 64.6 | Learning with Neighbor Consistency for Noisy Labels | google-research/scenic | 2022 |
| 6 | CLIPCleaner | 58.42 | CLIPCleaner: Cleaning Noisy Labels with CLIP | mrchenfeng/clipcleaner_acmmm2024 | 2024 |
| 7 | InstanceGM-SS | 56.37 | Instance-Dependent Noisy Label Learning via Graphical Modelling | arpit2412/InstanceGM | 2022 |
| 8 | InstanceGM | 52.24 | Instance-Dependent Noisy Label Learning via Graphical Modelling | arpit2412/InstanceGM | 2022 |