Federated Domain Generalization with Data-free On-server Matching Gradient
Trong-Binh Nguyen *, Minh-Duong Nguyen *, Jinsun Park †, Quoc-Viet Pham, Won Joo Hwang † · 2025
Read Federated Domain Generalization with Data-free On-server Matching Gradient by Trong-Binh Nguyen *, Minh-Duong Nguyen *, Jinsun Park †, Quoc-Viet Pham, Won Joo Hwang † (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-14653.txt.
Description
Trong-Binh Nguyen and co-authors teach federated models to work in places they have never seen. Domain generalization trains a model on several source settings so it still performs in new ones, usually by learning representations that ignore the setting. In the federated version, each setting's data sits on a different client and cannot be pooled.
Their method, FedOMG, holds the alignment on the server using gradients alone. Each client's local gradient carries information about its model and data; the server searches for a gradient direction that all domains agree on by maximizing inner products between them. Because it rides on gradients that federated training already exchanges, it adds no communication cost. It also plugs into existing federated methods rather than replacing them. Across seven standard benchmarks the approach beat recent baselines, the authors report.
Catalog record
| CREATOR | Trong-Binh Nguyen *, Minh-Duong Nguyen *, Jinsun Park †, Quoc-Viet Pham, Won Joo Hwang † |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 83 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-14653