Image Classification on CIFAR-100 Non-I.I.D-Dirichlet 0.3 (test)
70.36Best Test AccuracyFedGR
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FedGRNoise Type=Clean, phi=0.02024.11 | 70.36 | — | |
| FedGRNoise Type=Mixed, phi=0.62024.11 | 66.54 | — | |
| FedGRNoise Type=Asym, phi=0.62024.11 | 66.34 | — | |
| FedCorrNoise Type=Clean, phi=0.02024.11 | 65.38 | — | |
| FedGRNoise Type=Mixed, phi=1.02024.11 | 61.01 | — | |
| FedGRNoise Type=Sym, phi=0.62024.11 | 59.07 | — | |
| FedGRNoise Type=Asym, phi=1.02024.11 | 58 | — | |
| FedAvgNoise Type=Sym, phi=0.62024.11 | 33.85 | — | |
| FedGRNoise Type=Sym, phi=1.02024.11 | 31.89 | — | |
| FedGRMetric Scope=Avg2024.11 | — | 59.03 |