Image Classification on ImageNet-1K 20% synthetic label noise (test)
60AccuracyPr4ReL_B
Evaluation Results
| Method | Links | |
|---|---|---|
| Pr4ReL_BRe-label Model=SOP+, Selection Ratio=0.42023.11 | 60 | |
| UniformRe-label Model=SOP+, Selection Ratio=0.42023.11 | 59.2 | |
| ForgetRe-label Model=SOP+, Selection Ratio=0.42023.11 | 57.2 | |
| SmallLRe-label Model=SOP+, Selection Ratio=0.42023.11 | 54.4 | |
| Pr4ReL_BRe-label Model=SOP+, Selection Ratio=0.22023.11 | 53.5 | |
| UniformRe-label Model=SOP+, Selection Ratio=0.22023.11 | 52.7 | |
| ForgetRe-label Model=SOP+, Selection Ratio=0.22023.11 | 50.6 | |
| Pr4ReL_BRe-label Model=SOP+, Selection Ratio=0.12023.11 | 44.3 | |
| SmallLRe-label Model=SOP+, Selection Ratio=0.22023.11 | 42.7 | |
| UniformRe-label Model=SOP+, Selection Ratio=0.12023.11 | 42.5 | |
| SmallLRe-label Model=SOP+, Selection Ratio=0.12023.11 | 31.4 | |
| Pr4ReL_BRe-label Model=SOP+, Selection Ratio=0.052023.11 | 30.2 | |
| UniformRe-label Model=SOP+, Selection Ratio=0.052023.11 | 27.8 | |
| SmallLRe-label Model=SOP+, Selection Ratio=0.052023.11 | 22.8 | |
| ForgetRe-label Model=SOP+, Selection Ratio=0.12023.11 | 8.3 | |
| ForgetRe-label Model=SOP+, Selection Ratio=0.052023.11 | 4.1 |