Image Classification on SVHN (Label Subset Error Rate)
1.85Error Rate (1000 labels)CoMatch + EPASS
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CoMatch + EPASS2023.10 | 1.85 | 2.98 | 1.93 | |
| UDA2023.10 | 1.91 | 2.39 | 1.99 | |
| CoMatch2023.10 | 2.01 | 8.2 | 2.16 | |
| FixMatch2023.10 | 2.02 | 3.37 | 1.97 | |
| AdaMatch2023.10 | 2.02 | 6.14 | 2.13 | |
| SimMatch + EPASS2023.10 | 2.02 | 2.31 | 2.04 | |
| SimMatch2023.10 | 2.05 | 7.6 | 2.48 | |
| Dash2023.10 | 2.08 | 5.26 | 2.01 | |
| SoftMatch2023.10 | 2.09 | 2.46 | 2.15 | |
| Fully-Supervised2023.10 | 2.13 | — | — | |
| MixMatch2023.10 | 5.01 | 26.61 | 4.48 | |
| ReMixMatch2023.10 | 5.36 | 16.43 | 5.65 | |
| FreeMatch2023.10 | 7.56 | 10.43 | 8.23 | |
| FlexMatch2023.10 | 7.9 | 5.42 | 8.74 |