General Federated Class-Incremental Learning With Lightweight Generative Replay
- 29 July 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (20) , 33927-33939
Abstract
Federated class-incremental learning (FCIL) aims to allow federated learning (FL) systems to consistently learn new tasks with classes that change dynamically, without forgetting knowledge from previous classes. In FCIL scenarios, both heterogeneity in both label and data distribution across clients and catastrophic forgetting caused by continual emergence of new classes can significantly affect the performance of a FL system. Existing FCIL methods assume only changes in class distribution over time for each single client while ignoring class-specific domain distribution. Furthermore, these methods often rely on storing old class exemplars to mitigate catastrophic forgetting, potentially raising privacy concerns and computational burdens. In this article, we propose a FCIL framework called generative federated class-incremental learning (GenFCIL) that effectively addresses the aforementioned challenges. First, we introduce a lightweight generator that promotes knowledge sharing among clients and preserves the accumulated knowledge from all clients. By collecting classes and their associated data from each client, the generator effectively tackles data heterogeneity, facilitating information transfer across clients, and mitigating catastrophic forgetting in a replay-free manner. Importantly, the lightweight nature of the generator ensures that it does not impose excessive memory and computation requirements. Second, to tackle challenges from shifts in both class distribution and class-specific domain distribution in general FCIL scenarios, which may exacerbate catastrophic forgetting, we incorporate and update multiple logit scores from clients focusing on their old and new overlapping classes to incorporate more intraclass information. Experimental results show that GenFCIL effectively alleviates the impact of catastrophic forgetting and heterogeneity.Keywords
Funding Information
- National Natural Science Foundation of China (62106167)
- National Research Foundation Singapore and DSO National Laboratories under the AI Singapore Programme (AISG2-RP-2020-019)
- Agency for Science, Technology and Research under its RIE2025 Industry Alignment Fund—Industry Collaboration Projects (I2301E0026)
- Alibaba Group
- NTU Singapore
This publication has 40 references indexed in Scilit:
- Rethinking few-shot class-incremental learning: A lazy learning baselineExpert Systems with Applications, 2024
- Towards optimal learning: Investigating the impact of different model updating strategies in federated learningExpert Systems with Applications, 2024
- Federated recommender systems based on deep learning: The experimental comparisons of deep learning algorithms and federated learning aggregation strategiesExpert Systems with Applications, 2023
- Modeling global distribution for federated learning with label distribution skewPattern Recognition, 2023
- Robust federated learning under statistical heterogeneity via Hessian spectral decompositionPattern Recognition, 2023
- Exemplar-free class incremental learning via discriminative and comparable parallel one-class classifiersPattern Recognition, 2023
- Knowledge aggregation networks for class incremental learningPattern Recognition, 2023
- Multi-center federated learning: clients clustering for better personalizationWorld Wide Web, 2022
- Trust-driven reinforcement selection strategy for federated learning on IoT devicesComputing, 2022
- Memory Aware Synapses: Learning What (not) to Forget2018