Open AccessPrivacy‐preserving speaker verification system based on binary I‐vectors
Author(s) -
Mtibaa Aymen,
PetrovskaDelacrétaz Dijana,
Boudy Jérôme,
Ben Hamida Ahmed
Publication year - 2021
Publication title -
iet biometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.434
H-Index - 28
eISSN - 2047-4946
pISSN - 2047-4938
DOI - 10.1049/bme2.12013
Subject(s) - computer science , biometrics , speaker verification , key (lock) , shuffling , binary number , scheme (mathematics) , cloud computing , computer security , artificial neural network , speaker recognition , speech recognition , artificial intelligence , operating system , mathematical analysis , arithmetic , mathematics , programming language
Speaker verification is a key technology in many services and applications like smartphones and intelligent digital assistants. These applications usually require users to transmit their recordings, features, or models from their voices over untrusted public networks which stored and processed them on cloud‐based infrastructure. Furthermore, the voice signal contains a great deal of the speaker's personal and private information which raises several privacy issues. Therefore, it is necessary to develop speaker verification systems that protect the user's voice against such threats. Herein, the cancellable biometric systems have been introduced as a privacy‐preserving solution. A cancellable method for speaker verification systems is proposed using speaker i‐vector embeddings. This method includes two stages: (i) i‐vector binarisation and (ii) the protection of the binary i‐vector with a shuffling scheme derived from a user‐specific key. Privacy evaluation of this method according to the standards of biometric information protection (ISO/IEC 24745) shows that the proposed cancellable speaker verification system achieves the revocability, unlinkability, and irreversibility requirements. Moreover, the cancellable system improves biometric performance compared with the unprotected system and makes it resistant to different attack scenarios. Additionally, we demonstrate that this method can also operate to protect deep neural network speaker embeddings such as x‐vectors.