FACT: Federated Adversarial Cross Training
This addresses the challenge of domain adaptation in federated learning for scenarios where target data is unlabeled, which is incremental as it builds on existing federated and domain adaptation methods.
The paper tackles the problem of federated model adaptation to a target client without annotated data in non-i.i.d. settings, and shows that FACT outperforms state-of-the-art models on multiple benchmarks, achieving improvements in accuracy across three multi-source-single-target benchmarks and competitive results in single-source-single-target experiments.
Federated Learning (FL) facilitates distributed model development to aggregate multiple confidential data sources. The information transfer among clients can be compromised by distributional differences, i.e., by non-i.i.d. data. A particularly challenging scenario is the federated model adaptation to a target client without access to annotated data. We propose Federated Adversarial Cross Training (FACT), which uses the implicit domain differences between source clients to identify domain shifts in the target domain. In each round of FL, FACT cross initializes a pair of source clients to generate domain specialized representations which are then used as a direct adversary to learn a domain invariant data representation. We empirically show that FACT outperforms state-of-the-art federated, non-federated and source-free domain adaptation models on three popular multi-source-single-target benchmarks, and state-of-the-art Unsupervised Domain Adaptation (UDA) models on single-source-single-target experiments. We further study FACT's behavior with respect to communication restrictions and the number of participating clients.