Speech Privacy Leakage from Shared Gradients in Distributed Learning

TL;DR

This paper investigates how shared gradients in distributed speech learning models can leak private information, demonstrating the potential to recover speech content and speaker identity without direct data access.

Contribution

It is the first to analyze speech privacy leakage from gradients in distributed learning, showing the feasibility of inferring speech content and speaker identity.

Findings

Shared gradients can reveal speech content and speaker identity.

Experiments show high similarity between original and recovered speech signals.

Privacy leakage occurs even without direct data access.

Abstract

Distributed machine learning paradigms, such as federated learning, have been recently adopted in many privacy-critical applications for speech analysis. However, such frameworks are vulnerable to privacy leakage attacks from shared gradients. Despite extensive efforts in the image domain, the exploration of speech privacy leakage from gradients is quite limited. In this paper, we explore methods for recovering private speech/speaker information from the shared gradients in distributed learning settings. We conduct experiments on a keyword spotting model with two different types of speech features to quantify the amount of leaked information by measuring the similarity between the original and recovered speech signals. We further demonstrate the feasibility of inferring various levels of side-channel information, including speech content and speaker identity, under the distributed learning framework without accessing the user's data.