Image Classification on ImageNet (test) (Accuracy)
77.35AccuracyF-SAM
Evaluation results
| Method | Links | |
|---|---|---|
| F-SAMBackbone=ResNet-502024.03 | 77.35 | |
| SAMBackbone=ResNet-502024.03 | 77.14 | |
| ASAMBackbone=ResNet-502024.03 | 77.1 | |
| SGDBackbone=ResNet-502024.03 | 76.62 | |
| FedDPSDSetup=Federated learning, Data partition=Non-IID (Dirichlet, alpha=0.5), Privacy budget (epsilon)=102026.01 | 39.13 | |
| DENSESetup=Federated learning, Data partition=Non-IID (Dirichlet, alpha=0.5), Privacy budget (epsilon)=102026.01 | 32.34 | |
| FedADISetup=Federated learning, Data partition=Non-IID (Dirichlet, alpha=0.5), Privacy budget (epsilon)=102026.01 | 30.21 | |
| FedDAFLSetup=Federated learning, Data partition=Non-IID (Dirichlet, alpha=0.5), Privacy budget (epsilon)=102026.01 | 28.22 | |
| FedDFSetup=Federated learning, Data partition=Non-IID (Dirichlet, alpha=0.5), Privacy budget (epsilon)=102026.01 | 27.43 | |
| FedAVGSetup=Federated learning, Data partition=Non-IID (Dirichlet, alpha=0.5), Privacy budget (epsilon)=102026.01 | 11.89 |