Reliable, secure and privacy preserving multi-biometric person authentication – RESPECT

Biometrics refers to the automated recognition of individuals based on their behavioural or biological characteristics. In spite of their numerous advantages over traditional authentication systems based on PINs or passwords (e.g., biometric characteristics cannot be lost or forgotten), biometric systems are vulnerable to external attacks and can leak privacy. Presentations attacks (PAs) - impostors who manipulate biometric systems by masquerading as other people - are serious threats to security. Privacy concerns involve the use of personal and sensitive biometric information, as classified by the GDPR, for purposes other than those intended.

Vulnerabilities to PAs and privacy leakage are unacceptable and have hindered the deployment of biometric technology in commercial applications. The biometrics community has responded with presentation attack detection (PAD) technologies and privacy preservation mechanisms (biometric template protection schemes, BTP). Even though the latest PAD technologies are largely successful in protecting biometrics systems from known forms of PA, they tend to lack generalisation to different forms of attacks. The standard approach to privacy preservation involves some form of encryption or irreversible transformations, though the most recent fully homomorphic algorithms are general computationally prohibitive.

Multi-biometric systems, explored extensively as a means of improving recognition reliability, also offer potential to improve PAD generalisation. Multi-biometric systems offer natural protection against spoofing since an impostor is less likely to succeed in fooling multiple systems simultaneously. For the same reason, previously unseen PAs are less likely to fool multi-biometric systems protected by PAD. Unfortunately, each sub-system in a multi-biometric approach to recognition has potential to leak privacy. Multi-biometric systems only compound the need for computationally prohibitive privacy preservation.

RESPECT, a Franco-German collaborative project, will explore the potential of using multi-biometrics as a means to defend against diverse PAs and improve generalisation while still preserving privacy. Central to this idea is the use of (i) biometric characteristics that can be captured easily and reliably using ubiquitous smart devices and, (ii) biometric characteristics which facilitate computationally manageable privacy preserving, homomorphic encryption.

The research will focus on characteristics readily captured with consumer-grade microphones and video cameras, specifically face, iris and voice. Further advances beyond the current state of the art involve the consideration of dynamic characteristics, namely utterance verification and lip dynamics. The core research objective will be to determine which combination of biometrics characteristics gives the best biometric authentication reliability and PAD generalisation while remaining compatible with computationally efficient privacy preserving BTP schemes.