Privacy-Preserving Identification Systems With Noisy Enrollment

In this paper, we study fundamental trade-offs in privacy-preserving biometric identification systems with noisy enrollment. The proposed include helper data, secret keys, and private keys. Helper data are stored a public database used for identification. Secret keys either secure or provided to the user, can be next step, e.g. authentication. Private by users, also impose enrollment channel an arbitrarily small privacy secrecy leakage rate. We characterize optimal trade-off among identification, key, rates. Depending on how produced, two cases of systems, where generated chosen respectively. By introducing it is shown that system achieves close zero rate both key settings. results show enlarged increasing This work provides framework analyzing insight design systems.