Federated Dataset Dictionary Learning for Multi-Source Domain Adaptation
This addresses domain adaptation in privacy-sensitive federated settings, though it appears incremental as it builds upon existing Dataset Dictionary Learning with new protocols.
The paper tackles the problem of federated domain adaptation where clients have distributional shifts and some have unlabeled data, proposing FedDaDiL which uses dictionary learning of empirical distributions to generate labeled data on target domains, achieving successful results on Caltech-Office, TEP, and CWRU benchmarks.
In this article, we propose an approach for federated domain adaptation, a setting where distributional shift exists among clients and some have unlabeled data. The proposed framework, FedDaDiL, tackles the resulting challenge through dictionary learning of empirical distributions. In our setting, clients' distributions represent particular domains, and FedDaDiL collectively trains a federated dictionary of empirical distributions. In particular, we build upon the Dataset Dictionary Learning framework by designing collaborative communication protocols and aggregation operations. The chosen protocols keep clients' data private, thus enhancing overall privacy compared to its centralized counterpart. We empirically demonstrate that our approach successfully generates labeled data on the target domain with extensive experiments on (i) Caltech-Office, (ii) TEP, and (iii) CWRU benchmarks. Furthermore, we compare our method to its centralized counterpart and other benchmarks in federated domain adaptation.