Experimental Researches of Biometric Authentication Using Convolutional Neural Networks and Histograms of Oriented Graphs

Abstract

User authentication technologies are an important component of modern cybersecurity. These technologies are designed to verify the correctness and validity, for example, by verifying the authenticity of entered passwords, identifiers, certain biometric features, and so on. We consider biometric authentication techniques for the human face. Two techniques are used to process biometric features in our work: convolutional neural networks and histograms of directed gradients. We conduct experimental research to study these techniques and compare the results of both in terms of authentication accuracy and the speed of processing biometric images. The obtained results will be useful in the development and implementation of cybersecurity applications that use biometric authentication