Image Classification on Tiny-ImageNet Non-IID 2 (Accuracy, Binarized Accuracy)
54.54AccuracyFedAvg
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FedAvgModel=ResNet18, Nc=100, FLOPs=4.44 × 10^9, Memory (MB)=45.0902026.03 | 54.54 | 0.56 | |
| FedMUDModel=ResNet18, Nc=100, FLOPs=4.44 × 10^9, Memory (MB)=45.0902026.03 | 46.06 | 15.6 | |
| FedBNNModel=ResNet18, Nc=100, FLOPs=9.05 × 10^7, Memory (MB)=1.412026.03 | 45.74 | 45.74 | |
| FedHPro2026.05 | 40.52 | — | |
| FedRCLCommunication Protocol=Standard2026.05 | 38.76 | — | |
| FedSACommunication Protocol=Exchanging both prototypes and model parameters2026.05 | 38.35 | — | |
| FedTGPCommunication Protocol=Exchanging both prototypes and model parameters2026.05 | 38.21 | — | |
| FedGMKDCommunication Protocol=Standard2026.05 | 37.94 | — | |
| MOONCommunication Protocol=Standard2026.05 | 37.15 | — | |
| FedProtoCommunication Protocol=Exchanging both prototypes and model parameters2026.05 | 36.86 | — | |
| FedProxCommunication Protocol=Standard2026.05 | 36.62 | — | |
| FedAvgCommunication Protocol=Standard2026.05 | 36.2 | — | |
| FedBATModel=ResNet18, Nc=100, FLOPs=4.44 × 10^9, Memory (MB)=45.0902026.03 | 20.9 | 0.8 |