Optimal Defenses Against Data Reconstruction Attacks

Yuxiao Chen, Gamze Gursoy, Qi Lei

Keywords: Federated learning, Data reconstruction, Bayesian C-R lower bound

TL;DR: A lower bound for reconstruction loss in federated learning is proposed, and is used to optimize gradient pruning and gradient noise.

Abstract:

Federated Learning (FL) is designed to prevent
data leakage through collaborative model train-
ing without centralized data storage. However,
it is vulnerable to reconstruction attacks that re-
cover original training data from shared gradients.
To optimize the trade-off between data leakage
and utility loss, we first derive a theoretical lower
bound of reconstruction error (among all attack-
ers) for the two standard methods: adding noise,
and gradient pruning. We then customize these
two defenses to be parameter- and model-specific
and achieve the optimal trade-off between our ob-
tained reconstruction lower bound and model util-
ity. Experimental results validate that our methods
outperform Gradient Noise and Gradient Pruning
by protecting the training data better while also
achieving better utility.

Submission Number: 28