A Two-layer Authentication Using Voiceprint for Voice Assistants

Abstract

Voice assistants are a ubiquitous service of contemporary daily life. Their intuitive use and 24-hour-a-day convenience make them popular and have more users. However, the security of voice assistants does not increase as much as the rising amount of users and increasing abilities. The lack of an authentication mechanism gives attackers an opportunity to exploit voice assistants to control and get personal information from linked services. The goal of this thesis is to provide an authentication method that protects voice assistants from attacks without degrading their usability. We utilize Microsoft cognitive speaker recognition API and Google speech API to implement an Android application to examine the approach. The result indicates that the voice authentication method can resist replay attacks and it is easy to use and learn for users.