Image Classification on CIFAR-100 (test) (Symmetric and Asymmetric Noise Robustness)
68.75Symmetric Noise (η=0.2) AccuracyFasTEN
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| FasTENBackbone=ResNet-342021.11 | 68.75 | — | 63.82 | 55.22 | 37.36 | 70.35 | — | 67.93 | |
| MSLCBackbone=ResNet-342021.11 | 68.62 | — | 63.3 | 53.83 | 21.07 | 70.86 | — | 66.99 | |
| MLOCBackbone=ResNet-342021.11 | 68.16 | — | 62.09 | 54.49 | 20.23 | 69.2 | — | 66.48 | |
| MW-NetBackbone=ResNet-342021.11 | 66.73 | — | 59.44 | 49.19 | 19.04 | 67.9 | — | 64.5 | |
| GLCBackbone=ResNet-342021.11 | 60.99 | — | 49 | 33.38 | 20.38 | 64.43 | — | 54.2 | |
| SL2019.08 | 60.01 | 66.75 | 58 | 41.47 | 15 | 65.58 | 65.14 | 63.1 | |
| Forward2019.08 | 59.75 | 63.99 | 53.13 | 24.7 | 2.65 | 64.09 | 64 | 60.91 | |
| Deep kNNBackbone=ResNet-342021.11 | 59.6 | — | 52.48 | 39.9 | 23.39 | 57.71 | — | 50.23 | |
| CE2019.08 | 59.26 | 64.34 | 50.82 | 25.39 | 5.27 | 62.97 | 63.12 | 61.85 | |
| D2L2019.08 | 59.2 | 64.6 | 52.01 | 35.27 | 5.33 | 62.43 | 63.2 | 61.35 | |
| GCE2019.08 | 59.06 | 64.43 | 53.25 | 36.16 | 8.43 | 63.03 | 63.17 | 61.69 | |
| LSR2019.08 | 58.83 | 63.68 | 50.05 | 24.68 | 5.22 | 63.03 | 62.32 | 61.59 | |
| Bootstrap2019.08 | 57.91 | 63.26 | 48.17 | 12.27 | 1 | 63.44 | 63.18 | 62.08 | |
| L2RWBackbone=ResNet-342021.11 | 57.79 | — | 44.82 | 30.01 | 10.71 | 59.11 | — | 55.12 | |
| MLaCBackbone=ResNet-342021.11 | 49.81 | — | 35.15 | 20.15 | 12.85 | 56.46 | — | 49.2 |