An Efficient Random Unitary Matrix for Biometric Template Protection
Yuko Saito, Ibuki Nakamura, Sayaka Shiota, Hitoshi Kiya · 2016
This paper proposes a new way to generate random unitary matrices for biometric template protection. It is well known that the unitary transform-based template protection that is a class of cancelable biometrics systems, has some desirable properties such as being applicable to l2-norm minimization problems. However, its performance and effectiveness depend on the variety of a unitary matrix. The aim of this paper is to generate an effective random unitary matrix and evaluate the effectiveness in terms of security, recognition performance and the complexity of the recognition system. The proposed random matrix consists of a random permutation matrix and a unitary matrix in which all elements have fixed values as the discrete Fourier transform(DFT). It is also applied to face recognition experiments to demonstrate the effectiveness.