Biometric Template Protection: Bridging the performance gap between theory and practice

摘要: Biometric recognition is an integral component of modern identity management and access control systems. Due to the strong permanent link between individuals their biometric traits, exposure enrolled users? information adversaries can seriously compromise system security user privacy. Numerous techniques have been proposed for template protection over last 20 years. While these are theoretically sound, they seldom guarantee desired noninvertibility, revocability, nonlinkability properties without significantly degrading performance. The objective this work analyze factors contributing performance divide highlight promising research directions bridge gap. design invariant representations remains a fundamental problem, despite recent attempts address issue through feature adaptation schemes. difficulty in estimating statistical distribution features not only hinders development better algorithms but also diminishes ability quantify noninvertibility existing algorithms. Finally, achieving use external secrets (e.g., passwords) continues be challenging proposition. Further on above issues required cross chasm theory practice ?template protection.