Deep Secure Encoding: An Application to Face Recognition
This addresses biometric security for face recognition systems, offering a practical solution with incremental improvements in template protection.
The paper tackles secure face recognition by proposing Deep Secure Encoding, a framework that maps face classes to high entropy secure codes using CNNs and hashes them for biometric template protection. It achieves state-of-the-art matching performance on CMU-PIE and Extended Yale B databases, with cancelability and high security.
In this paper we present Deep Secure Encoding: a framework for secure classification using deep neural networks, and apply it to the task of biometric template protection for faces. Using deep convolutional neural networks (CNNs), we learn a robust mapping of face classes to high entropy secure codes. These secure codes are then hashed using standard hash functions like SHA-256 to generate secure face templates. The efficacy of the approach is shown on two face databases, namely, CMU-PIE and Extended Yale B, where we achieve state of the art matching performance, along with cancelability and high security with no unrealistic assumptions. Furthermore, the scheme can work in both identification and verification modes.