Design and Implementation of Identity Verification Software Based on Deep Learning

Abstract

Identity verification, a noncontact biometric identification technology, has important scientific significance in theoretical research and shows great practical value in national security, public safety, and finance. In view of this situation, this paper designs an identity verification software based on deep learning, which has been successfully applied to real-world applications. The central idea of the software can be summarized as follows: First, the lightweight multi-task cascaded convolutional network (MTCNN), which can learn correlations between face detection and alignment, is employed for face detection. The software then conducts face recognition with MobileFaceNet which is an efficient and lightweight neural network, reducing the hardware cost. The test results show that the software meets the design requirements and can complete the corresponding identity confirmation function.