Privacy-Preserving Preselection for Protected Biometric Identification Using Public-Key Encryption with Keyword Search

The efficiency of biometric systems, in particular efficient and accurate identification, is one the most challenging open problems biometrics today. Adding to that, data are sensitive deserving adequate protection. As a solution, this work proposes an privacy-preserving reduction computational workload identification systems using public-key encryption with keyword search (PEKS). For long-term protection data, fully homomorphic applied for template all cryptographic schemes lattice-based, they also offer post-quantum security. Throughout system, recognition accuracy unprotected system preserved. In evaluation on public face database, encrypted domain reduced down 8.4% compared exhaustive search, achieving 1062 subjects 210 milliseconds. Based these results, 1 million can be estimated at under 3 minutes off-the-shelf hardware.