FedConv: A Learning-on-Model Paradigm for Heterogeneous Federated Clients

Keyphrases

• Federated Learning System 100%
• Global Model 66%
• Federated Learning 66%
• Convolutional 66%
• Compression Method 33%
• Model Accuracy 33%
• Computation Overhead 33%
• Communication Overhead 33%
• System Resources 33%
• Starvation 33%
• Large Animal Model 33%
• Resource-constrained 33%
• Decompression 33%
• Public Dataset 33%
• Smartphone 33%
• Collaborative Training 33%
• Edge Server 33%
• One-size-fits-all 33%
• Personalized Information 33%
• Private Data 33%
• All-solution-processed 33%

Computer Science

• Federated Learning 100%
• Learning System 40%
• Model Accuracy 20%
• Edge Server 20%
• Learning Framework 20%
• Communication Overhead 20%
• System Resource 20%