One-Factor Cancelable Biometric Template Protection Scheme for Real-Valued Features

Ruoqi Zhang, Peisong Shen, Kewei Lv, Chi Chen

Abstract: Biometric template protection is crucial in biometric recognition. Cancelable biometrics, a method for template protection, typically requires both human biometric data and a token as inputs for template generation. However, when the token is compromised, the template is vulnerable to various security attacks, leading to privacy breaches. This paper focuses on one-factor cancelable biometric scheme that relieves users from the burden of managing tokens. Specifically, we propose a new one-factor cancelable biometric scheme for protecting real-valued biometric features. It first obtains a hashed code from the biometirc feature using prime Cosine LSH functions, then generates an index sequence from the hashed code, finally, reorders a random string using the index sequence to create a cancelable template. The experiment results demonstrate the efficiency of the proposal scheme on real-valued facial, ear and fingerprint biometric features. We also verify the unlinkability, revocability and irreversibility of our scheme against various attacks.