Image Classification on CIFAR-10 ρ=0.2 (test)
83.77AccuracyHARMONY
Evaluation results
| Method | Links | ||
|---|---|---|---|
| HARMONYBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=22026.05 | 83.77 | 31.1 | |
| SplitGP + FTBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=2, Personalization Intensity (λ)=0.2, Evaluation Protocol (Fine-tuning)=on-device fine-tuning2026.05 | 81.49 | 31.27 | |
| SplitGPBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=2, Personalization Intensity (λ)=0.22026.05 | 80.87 | 25.78 | |
| SplitGPBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=2, Personalization Intensity (λ)=0.92026.05 | 78.12 | 38.59 | |
| Multi-exit SFLBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=22026.05 | 64.69 | 48.57 | |
| SplitFedBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=22026.05 | 46.83 | 125.76 |