Image Classification on WebVision
80.88Top-1 AccCoDiM-Sup
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CoDiM-Sup2021.11 | 80.88 | 92.48 | — | |
| CoDiM-bare2021.11 | 80.32 | 93.4 | — | |
| C2Dimplementation=re-implemented2021.11 | 80.2 | 93.22 | — | |
| CoDiM-Self2021.11 | 80.12 | 93.52 | — | |
| GJS2021.11 | 79.28 | 91.22 | — | |
| LongReMix2021.11 | 78.92 | 92.32 | — | |
| DM-AugDescimplementation=re-implemented2021.11 | 78.64 | 93.2 | — | |
| ELR+2021.11 | 77.78 | 91.68 | — | |
| DivideMix2021.11 | 77.32 | 91.64 | — | |
| fSGLDModel=ResNet-34, Wall-clock time per iteration (s/epoch)=23.72025.10 | 73.95 | 90.03 | — | |
| fSGLDModel=ResNet-50, Wall-clock time per iteration (s/epoch)=34.12025.10 | 73.54 | 90.34 | — | |
| SAMModel=ResNet-34, Wall-clock time per iteration (s/epoch)=41.32025.10 | 73.49 | 90.32 | — | |
| ASAMModel=ResNet-34, Wall-clock time per iteration (s/epoch)=41.42025.10 | 73.46 | 90.14 | — | |
| SAMModel=ResNet-50, Wall-clock time per iteration (s/epoch)=60.72025.10 | 72.52 | 89.53 | — | |
| ASAMModel=ResNet-50, Wall-clock time per iteration (s/epoch)=60.92025.10 | 71.92 | 88.48 | — | |
| SGDModel=ResNet-34, Wall-clock time per iteration (s/epoch)=22.02025.10 | 71.87 | 89.33 | — | |
| SGDModel=ResNet-50, Wall-clock time per iteration (s/epoch)=31.92025.10 | 71.11 | 88.31 | — | |
| AdamWModel=ResNet-50, Wall-clock time per iteration (s/epoch)=32.32025.10 | 69.92 | 87.97 | — | |
| AdamWModel=ResNet-34, Wall-clock time per iteration (s/epoch)=22.52025.10 | 68.69 | 87.01 | — | |
| Iterative-CV2021.11 | 65.24 | 85.34 | — | |
| Co-teaching2021.11 | 63.58 | 85.2 | — | |
| CE2022.03 | — | — | 66.96 | |
| Co-teaching2022.03 | — | — | 63.58 | |
| CurrNetBackbone=Inception-ResNetV22022.02 | — | — | 79.3 | |
| Decoupling2022.03 | — | — | 62.54 | |
| DivideMix2022.03 | — | — | 77.32 | |
| HeteroscedasticBackbone=Inception-ResNetV22022.02 | — | — | 76.6 | |
| Iterative-CV2022.03 | — | — | 65.24 | |
| MentorNet2022.03 | — | — | 63 | |
| MILeBackbone=ResNet-502022.02 | — | — | 75.2 | |
| MixupBackbone=ResNet-502022.02 | — | — | 75.5 | |
| MoProBackbone=ResNet-502022.02 | — | — | 73.9 | |
| NCRBackbone=ResNet-502022.02 | — | — | 75.7 | |
| NCR+MixupBackbone=ResNet-502022.02 | — | — | 75.4 | |
| NCR+Mixup+DABackbone=ResNet-502022.02 | — | — | 76.8 | |
| SPR2022.03 | — | — | 78.12 | |
| StandardBackbone=ResNet-502022.02 | — | — | 74.9 |