Benchmarking and challenges in security and privacy for voice biometrics

View on arXiv ← Back to list

Authors: Jean-Francois Bonastre, Hector Delgado, Nicholas Evans, Tomi Kinnunen, Kong Aik Lee, Xuechen Liu, Andreas Nautsch, Paul-Gauthier Noe, Jose Patino, Md Sahidullah, Brij Mohan Lal Srivastava, Massimiliano Todisco, Natalia Tomashenko, Emmanuel Vincent, Xin Wang, Junichi Yamagishi

Published: 2021-09-01 09:41:44+00:00

AI Summary

This paper provides a high-level overview of benchmarking methodologies used in voice biometrics security and privacy research. It describes the ASVspoof challenge for spoofing countermeasures and the VoicePrivacy initiative for privacy preservation through anonymization.

Abstract

For many decades, research in speech technologies has focused upon improving reliability. With this now meeting user expectations for a range of diverse applications, speech technology is today omni-present. As result, a focus on security and privacy has now come to the fore. Here, the research effort is in its relative infancy and progress calls for greater, multidisciplinary collaboration with security, privacy, legal and ethical experts among others. Such collaboration is now underway. To help catalyse the efforts, this paper provides a high-level overview of some related research. It targets the non-speech audience and describes the benchmarking methodology that has spearheaded progress in traditional research and which now drives recent security and privacy initiatives related to voice biometrics. We describe: the ASVspoof challenge relating to the development of spoofing countermeasures; the VoicePrivacy initiative which promotes research in anonymisation for privacy preservation.

Key findings

Spoofing attacks significantly degrade ASV system reliability, but effective countermeasures exist, achieving EERs as low as 0.2%. Anonymization techniques, while aiming to prevent speaker recognition, have not yet achieved full anonymization, with re-identification still possible depending on the model adaptation strategy.

Approach

The paper focuses on benchmarking challenges, ASVspoof and VoicePrivacy, to evaluate the performance of spoofing countermeasures and anonymization techniques in audio. These challenges use standardized datasets, evaluation rules, and metrics to allow for meaningful comparison of different approaches.

Datasets

ASVspoof (various years) and VoicePrivacy datasets are mentioned, but specific datasets within those challenges are not explicitly named.

Model(s)

The paper does not detail specific models used in the challenges; it focuses on the benchmarking methodologies and results rather than the individual models submitted by participants.

Author countries

France, Spain, UK, Finland, Singapore, China, Germany, Japan

← Previous