Balanced Multi-modal Federated Learning via Cross-Modal Infiltration

Federated learning (FL) underpins advancements in privacy-preserving distributed computing by collaboratively training neural networks without exposing clients' raw data.