Image Classification on CIFAR10 (Non-IID 2)
79.7AccuracyFedHPro
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FedHPro2026.05 | 79.7 | — | |
| FedTGPCommunication Protocol=Exchanging both prototypes and model parameters2026.05 | 77.95 | — | |
| FedSACommunication Protocol=Exchanging both prototypes and model parameters2026.05 | 77.86 | — | |
| FedRCLCommunication Protocol=Standard2026.05 | 77.51 | — | |
| FedGMKDCommunication Protocol=Standard2026.05 | 77.29 | — | |
| MOONCommunication Protocol=Standard2026.05 | 76.07 | — | |
| FedProtoCommunication Protocol=Exchanging both prototypes and model parameters2026.05 | 75.32 | — | |
| FedProxCommunication Protocol=Standard2026.05 | 74.67 | — | |
| FedAvgCommunication Protocol=Standard2026.05 | 74.48 | — | |
| FedBNNModel=ResNet10, Nc=100, FLOPs=1.11 × 10^7, Memory (MB)=0.61302026.03 | 73.82 | 73.82 | |
| FedAvgModel=ResNet10, Nc=100, FLOPs=4.40 × 10^8, Memory (MB)=19.61702026.03 | 70.62 | 12.74 | |
| FedMUDModel=ResNet10, Nc=100, FLOPs=4.40 × 10^8, Memory (MB)=19.61702026.03 | 67.22 | 18.98 | |
| FedBATModel=ResNet10, Nc=100, FLOPs=4.40 × 10^8, Memory (MB)=19.61702026.03 | 63.7 | 10.26 | |
| FedBNNModel=ConvNeXt-Tiny, Nc=100, FLOPs=6.07 × 10^7, Memory (MB)=3.48872026.03 | 63 | 63 | |
| FedBATModel=ConvNeXt-Tiny, Nc=100, FLOPs=2.98 × 10^9, Memory (MB)=111.6402026.03 | 61.88 | 16.54 | |
| FedAvgModel=ConvNeXt-Tiny, Nc=100, FLOPs=2.98 × 10^9, Memory (MB)=111.6402026.03 | 61.84 | 16.56 | |
| FedMUDModel=ConvNeXt-Tiny, Nc=100, FLOPs=2.98 × 10^9, Memory (MB)=111.6402026.03 | 38.18 | 29.5 |