Auditing Privacy Defenses in Federated Learning via Generative Gradient Leakage
This work addresses privacy risks for users in distributed learning systems by exposing vulnerabilities in current defenses, though it is incremental as it builds on prior gradient leakage methods.
The authors tackled the problem of privacy leakage in Federated Learning despite existing defenses, by introducing Generative Gradient Leakage (GGL), which uses GAN priors and gradient-free optimization to reconstruct high-quality images from degraded gradients, achieving superior reconstruction compared to gradient-based methods.
Federated Learning (FL) framework brings privacy benefits to distributed learning systems by allowing multiple clients to participate in a learning task under the coordination of a central server without exchanging their private data. However, recent studies have revealed that private information can still be leaked through shared gradient information. To further protect user's privacy, several defense mechanisms have been proposed to prevent privacy leakage via gradient information degradation methods, such as using additive noise or gradient compression before sharing it with the server. In this work, we validate that the private training data can still be leaked under certain defense settings with a new type of leakage, i.e., Generative Gradient Leakage (GGL). Unlike existing methods that only rely on gradient information to reconstruct data, our method leverages the latent space of generative adversarial networks (GAN) learned from public image datasets as a prior to compensate for the informational loss during gradient degradation. To address the nonlinearity caused by the gradient operator and the GAN model, we explore various gradient-free optimization methods (e.g., evolution strategies and Bayesian optimization) and empirically show their superiority in reconstructing high-quality images from gradients compared to gradient-based optimizers. We hope the proposed method can serve as a tool for empirically measuring the amount of privacy leakage to facilitate the design of more robust defense mechanisms.