Federated Class-Incremental Learning on FCIL-H
95.06Average Accuracy (Aavg)UP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| UPHeterogeneity alpha=1.02026.03 | 95.06 | 95.87 | |
| UPHeterogeneity alpha=0.52026.03 | 95.06 | 95.87 | |
| FeDMRAHeterogeneity alpha=1.02026.03 | 91.77 | 87.34 | |
| FeDMRAHeterogeneity alpha=0.52026.03 | 88.48 | 82.95 | |
| MFCLHeterogeneity alpha=1.02026.03 | 79.27 | 51.09 | |
| DDDRHeterogeneity alpha=1.02026.03 | 78.23 | 64 | |
| MFCLHeterogeneity alpha=0.52026.03 | 77.91 | 51.2 | |
| Re-FedHeterogeneity alpha=1.02026.03 | 76.98 | 73.44 | |
| Re-FedHeterogeneity alpha=0.52026.03 | 76.98 | 73.44 | |
| DDDRHeterogeneity alpha=0.52026.03 | 73.21 | 59.02 | |
| UACL+FLHeterogeneity alpha=1.02026.03 | 66.59 | 48.44 | |
| UACL+FLHeterogeneity alpha=0.52026.03 | 66.59 | 48.44 | |
| iCaRL+FLHeterogeneity alpha=1.02026.03 | 61.89 | 46.51 | |
| iCaRL+FLHeterogeneity alpha=0.52026.03 | 61.89 | 46.51 |