Videostream surveillance and face recognition system

The article looks into the practical aspects of developing a system of face identification and verification in a videostream. The practical part of this paper is made up of the pilot surveillance system capable of videostream face recognition, and the possibilities of scaling and putting it in operation quickly. For its implementation, advanced (so-termed "state of the art") machine-learning algorithms were used; and for the creation of the necessary pilot framework, the authors used containerization and data storage tools capable of scaling and grid computing.