A Secure and Compact Multimodal Biometric Authentication Scheme using Deep Hashing

Unimodal biometrics-based authentication systems are trending and applied to various real-time applications since biometrics is difficult to forge than traditional password-based access systems. These systems are also having the challenges like protecting biometrics from identity theft and database compromise. Multimodal biometric systems have several advantages, including lower error rates, higher accuracy, and larger population coverage. However, multimodal systems have an increased demand for integrity and privacy because they must store multiple biometric traits associated with each user. Even though multimodal biometric system provides reasonable advantages, but still, there is a significant requirement for secure multimodal biometric template protection schemes to protect the multimodal biometric templates. In this paper, a deep hashing framework for feature-level fusion that generates an advanced secure multimodal template from each user's finger and finger-vein biometrics is proposed. The matching performance will get improved due to the fusion of multiple biometrics. Furthermore, the proposed approach also provides cancelability and unlinkability of the templates along with improved privacy of the biometric data to protect from the different attacks. It is for integrating multimodal fusion, deep hashing, and biometric security, with an emphasis on structural data from modalities like fingerprint and finger-vein. VGG-19 is used for feature extraction and experiments were conducted using standard datasets of fingerprint and finger vein images. This system achieved an accuracy of 95%.