Dataset Dictionary Learning in a Wasserstein Space for Federated Domain Adaptation

no code implementations • 16 Jul 2024 • Eduardo Fernandes Montesuma, Fabiola Espinoza Castellon, Fred Ngolè Mboula, Aurélien Mayoue, Antoine Souloumiac, Cédric Gouy-Pailler

Multi-Source Domain Adaptation (MSDA) is a challenging scenario where multiple related and heterogeneous source datasets must be adapted to an unlabeled target dataset.

Dictionary Learning Domain Adaptation

Federated Dataset Dictionary Learning for Multi-Source Domain Adaptation

no code implementations • 14 Sep 2023 • Fabiola Espinoza Castellon, Eduardo Fernandes Montesuma, Fred Ngolè Mboula, Aurélien Mayoue, Antoine Souloumiac, Cédric Gouy-Pailler

The proposed framework, FedDaDiL, tackles the resulting challenge through dictionary learning of empirical distributions.

Dictionary Learning Domain Adaptation

Federated learning with incremental clustering for heterogeneous data

no code implementations • 17 Jun 2022 • Fabiola Espinoza Castellon, Aurelien Mayoue, Jacques-Henri Sublemontier, Cedric Gouy-Pailler

To prevent such a bottleneck, we propose FLIC (Federated Learning with Incremental Clustering), in which the server exploits the updates sent by clients during federated training instead of asking them to send their parameters simultaneously.

Clustering Federated Learning