Image Classification on Tiny-ImageNet Held-out Clients (test)
90.6AccuracyPEP-FedPT
Evaluation results
| Method | Links | |
|---|---|---|
| PEP-FedPTPartitioning=Pathological, Classes per client=102025.10 | 90.6 | |
| SGPTPartitioning=Pathological, Classes per client=102025.10 | 78.1 | |
| FedVPT-DPartitioning=Pathological, Classes per client=102025.10 | 76.89 | |
| FedVPTPartitioning=Pathological, Classes per client=102025.10 | 74.15 | |
| HeadPartitioning=Pathological, Classes per client=102025.10 | 68.97 | |
| FedPRPartitioning=Pathological, Classes per client=102025.10 | 68.23 | |
| P-PTPartitioning=Pathological, Classes per client=102025.10 | 60.41 |