#gradientInversion

ModConFlexModConFlex
2025-02-06

"Approximate and Weighted Data Reconstruction Attack in Federated Learning" co-authored by researcher Ziqi Wang
➡️ demonstrates that private data can be reconstructed from shared information in federated learning
➡️ emphasizing the urgent need for stronger defense mechanisms to protect user privacy.



arxiv.org/abs/2308.06822

Graphical comparison of accuracy of description of original graphics with DLG, AGIC and AWA (own) result over four different cases.

Client Info

Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst