Loosely Coupled Federated Learning Over Generative Models

TL;DR

This paper introduces LC-FL, a federated learning framework that uses generative models to enable low-communication, heterogeneous collaborative learning across clients with different models, overcoming strict model homogeneity constraints.

Contribution

The paper proposes a novel LC-FL framework that leverages generative models for flexible, low-communication federated learning among heterogeneous clients.

Findings

Effective in real-world multiparty scenarios

Reduces communication costs significantly

Supports diverse model types across clients

Abstract

Federated learning (FL) was proposed to achieve collaborative machine learning among various clients without uploading private data. However, due to model aggregation strategies, existing frameworks require strict model homogeneity, limiting the application in more complicated scenarios. Besides, the communication cost of FL's model and gradient transmission is extremely high. This paper proposes Loosely Coupled Federated Learning (LC-FL), a framework using generative models as transmission media to achieve low communication cost and heterogeneous federated learning. LC-FL can be applied on scenarios where clients possess different kinds of machine learning models. Experiments on real-world datasets covering different multiparty scenarios demonstrate the effectiveness of our proposal.