A review of privacy-preserving biometric identification and authentication protocols

Biometrics now play a crucial role in user identification and authentication. However, storing biometric features in plaintext and conducting authentication without protection pose a risk of privacy leakage. To address this issue, privacy-preserving biometric identification and authentication protocols have been proposed. These protocols leverage techniques such as homomorphic encryption (HE) and secure multi-party computation (MPC) to prevent involving parties from knowing other parties' biometrics when performing authentications or identifications between different parties. In this paper, we present a thorough survey of privacy-preserving biometric protocols, classifying them into seven distinct models based on their application scenarios. In each scenario, we delve into its security requirements and potential threats, underscoring the importance of comprehending varied application scenarios in the design of practical biometric protection schemes. We also summarize current research gaps and potential future research directions, offering insights for advancing the field.