STSA: Federated Class-Incremental Learning via Spatial-Temporal Statistics Aggregation
This work solves practical deployment challenges in federated learning for incremental tasks, though it is incremental as it builds on existing FCIL methods.
The paper tackles the problem of federated class-incremental learning (FCIL) by addressing spatial-temporal client drift and high overhead from data heterogeneity, proposing STSA to aggregate feature statistics across clients and stages, with results showing outperformance in performance, flexibility, and efficiency on three datasets.
Federated Class-Incremental Learning (FCIL) enables Class-Incremental Learning (CIL) from distributed data. Existing FCIL methods typically integrate old knowledge preservation into local client training. However, these methods cannot avoid spatial-temporal client drift caused by data heterogeneity and often incur significant computational and communication overhead, limiting practical deployment. To address these challenges simultaneously, we propose a novel approach, Spatial-Temporal Statistics Aggregation (STSA), which provides a unified framework to aggregate feature statistics both spatially (across clients) and temporally (across stages). The aggregated feature statistics are unaffected by data heterogeneity and can be used to update the classifier in closed form at each stage. Additionally, we introduce STSA-E, a communication-efficient variant with theoretical guarantees, achieving similar performance to STSA-E with much lower communication overhead. Extensive experiments on three widely used FCIL datasets, with varying degrees of data heterogeneity, show that our method outperforms state-of-the-art FCIL methods in terms of performance, flexibility, and both communication and computation efficiency.