With the widespread development of biometric systems, concerns about security and privacy are increasing. An active area of research is template protection technology, which aims to protect registered biometric data. We focus on a homomorphic encryption approach, which enables building a “cryptographically-secure” system. In DPM 2013, Yasuda et al. proposed an efficient template protection system, using the homomorphic encryption scheme proposed by Brakerski and Vaikuntanathan. In this work, we improve and fortify their system to withstand impersonation attacks such as replay and spoofing attacks. We introduce a challenge-response authentication mechanism in their system and design a practical distributed architecture where computation and authentication are segregated. Our comprehensive system would be useful to build a large-scale and secure biometric system such as secure remote authentication over public networks.
Mandal, A., Roy, A., & Yasuda, M. (2016). Comprehensive and improved secure biometric system using homomorphic encryption. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9481, pp. 183–198). Springer Verlag. https://doi.org/10.1007/978-3-319-29883-2_12