FAST: Federated Active Learning with Foundation Models for Communication-efficient Sampling and Training
This work addresses communication efficiency and annotation effort in cross-silo federated learning, offering a practical solution for domains like medical imaging, though it is incremental as it builds on existing FAL methods with a novel integration of foundation models.
The paper tackles the problem of high annotation costs and communication-intensive sampling in Federated Active Learning (FAL) by introducing FAST, a two-pass framework that uses foundation models for weak labeling and refinement on uncertain samples, achieving an average performance improvement of 4.36% and reducing communication rounds eightfold with a 5% labeling budget.
Federated Active Learning (FAL) has emerged as a promising framework to leverage large quantities of unlabeled data across distributed clients while preserving data privacy. However, real-world deployments remain limited by high annotation costs and communication-intensive sampling processes, particularly in a cross-silo setting, when clients possess substantial local datasets. This paper addresses the crucial question: What is the best practice to reduce communication costs in human-in-the-loop learning with minimal annotator effort? Existing FAL methods typically rely on iterative annotation processes that separate active sampling from federated updates, leading to multiple rounds of expensive communication and annotation. In response, we introduce FAST, a two-pass FAL framework that harnesses foundation models for weak labeling in a preliminary pass, followed by a refinement pass focused exclusively on the most uncertain samples. By leveraging representation knowledge from foundation models and integrating refinement steps into a streamlined workflow, FAST substantially reduces the overhead incurred by iterative active sampling. Extensive experiments on diverse medical and natural image benchmarks demonstrate that FAST outperforms existing FAL methods by an average of 4.36% while reducing communication rounds eightfold under a limited 5% labeling budget.