Federated Active Learning Under Extreme Non-IID and Global Class Imbalance

ArXi:2603.10341v1 Announce Type: cross Federated active learning (FAL) seeks to reduce annotation cost under privacy constraints, yet its effectiveness degrades in realistic settings with severe global class imbalance and highly heterogeneous clients. We conduct a systematic study of query-model selection in FAL and uncover a central insight: the model that achieves class-balanced sampling, especially for minority classes, consistently leads to better final performance.