Domain-Incremental Classification on NICO_U (test)
69.22Accuracy (T1)FedAvg
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| FedAvgBackbone=ResNet-18, Incremental Setup=Domain-Incremental2026.02 | 69.22 | 68.25 | 72.87 | 72.16 | 73.27 | 73.15 | |
| FedProxBackbone=ResNet-18, Incremental Setup=Domain-Incremental2026.02 | 69.22 | 68.04 | 72.96 | 73.22 | 71.47 | 73.09 | |
| FedEWCBackbone=ResNet-18, Incremental Setup=Domain-Incremental, lambda=0.12026.02 | 69.22 | 68.77 | 73.24 | 73.96 | 73.84 | 73.33 | |
| OSCAR-ILBackbone=ResNet-18, Incremental Setup=Domain-Incremental2026.02 | 59.98 | 68.35 | 70.28 | 69.24 | 70.22 | 70.26 | |
| OSCAR-RBackbone=ResNet-18, Incremental Setup=Domain-Incremental, lambda=0.12026.02 | 59.98 | 66.22 | 67.51 | 67.03 | 67.7 | 68.11 | |
| OSI-FLBackbone=ResNet-18, Incremental Setup=Domain-Incremental, retained samples per task=52026.02 | 59.98 | 69.39 | 73.5 | 73.46 | 75.75 | 73.85 |