Taming Cross-Domain Representation Variance in Federated Prototype Learning with Heterogeneous Data Domains

Open in new window