Overcoming Catastrophic Forgetting in Federated Class-Incremental Learning via Federated Global Twin Generator

TL;DR

This paper introduces FedGTG, a federated learning framework that uses a global generative model to mitigate catastrophic forgetting in class-incremental learning without accessing private client data.

Contribution

FedGTG is a novel federated class-incremental learning approach that trains a global generator to produce synthetic data, improving knowledge retention and task learning.

Findings

FedGTG outperforms previous methods in accuracy and forgetting metrics.

The framework achieves better calibration and convergence to flat minima.

Experimental results on CIFAR-10, CIFAR-100, and tiny-ImageNet validate its effectiveness.

Abstract

Federated Class-Incremental Learning (FCIL) increasingly becomes important in the decentralized setting, where it enables multiple participants to collaboratively train a global model to perform well on a sequence of tasks without sharing their private data. In FCIL, conventional Federated Learning algorithms such as FedAVG often suffer from catastrophic forgetting, resulting in significant performance declines on earlier tasks. Recent works, based on generative models, produce synthetic images to help mitigate this issue across all classes, but these approaches' testing accuracy on previous classes is still much lower than recent classes, i.e., having better plasticity than stability. To overcome these issues, this paper presents Federated Global Twin Generator (FedGTG), an FCIL framework that exploits privacy-preserving generative-model training on the global side without accessing client data. Specifically, the server trains a data generator and a feature generator to create two types of information from all seen classes, and then it sends the synthetic data to the client side. The clients then use feature-direction-controlling losses to make the local models retain knowledge and learn new tasks well. We extensively analyze the robustness of FedGTG on natural images, as well as its ability to converge to flat local minima and achieve better-predicting confidence (calibration). Experimental results on CIFAR-10, CIFAR-100, and tiny-ImageNet demonstrate the improvements in accuracy and forgetting measures of FedGTG compared to previous frameworks.