Multibiometric Secure System Based on Deep Learning
This work addresses security and privacy issues in biometric systems for applications like access control, though it appears incremental as it builds on existing fusion and error-correction methods.
The paper tackles the problem of secure biometric authentication by proposing a deep learning-based multibiometric system that fuses multiple biometrics into a cancelable template, achieving state-of-the-art matching performance with enhanced security and cancelability.
In this paper, we propose a secure multibiometric system that uses deep neural networks and error-correction coding. We present a feature-level fusion framework to generate a secure multibiometric template from each user's multiple biometrics. Two fusion architectures, fully connected architecture and bilinear architecture, are implemented to develop a robust multibiometric shared representation. The shared representation is used to generate a cancelable biometric template that involves the selection of a different set of reliable and discriminative features for each user. This cancelable template is a binary vector and is passed through an appropriate error-correcting decoder to find a closest codeword and this codeword is hashed to generate the final secure template. The efficacy of the proposed approach is shown using a multimodal database where we achieve state-of-the-art matching performance, along with cancelability and security.