FuSeFL: Fully Secure and Scalable Cross-Silo Federated Learning

Sahar Ghoflsaz Ghinani, Elaheh Sadredini · 2025

Read FuSeFL: Fully Secure and Scalable Cross-Silo Federated Learning by Sahar Ghoflsaz Ghinani, Elaheh Sadredini (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-2507-13591.txt.

Description

Sahar Ghoflsaz Ghinani and Elaheh Sadredini present FuSeFL in this 2025 paper. The scheme lets groups train a shared model without pooling their raw data. That setup suits fields where data are private. Past secure methods often cost much time and memory. Many also fail to keep the shared model itself secret.

FuSeFL splits training across pairs of clients and uses a light form of secure joint computing. The server only links pairs, routes data, and merges updates. That design cuts server strain and keeps data, model, and updates hidden. In their tests, it blocked attacks that try to rebuild or probe data. It also trained far faster and used half the server memory of their baseline.

Catalog record

CREATORSahar Ghoflsaz Ghinani, Elaheh Sadredini
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE62 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-13591

Browse the full library · About the Tower of Babel