Videoverification Module for Bank Customer
Polina A. Shelupanova, Polina Yu. Davydchenko, Daria V. Ivanova, Zaxar M. Medvedev, Mikhail A. Cholokogly, Svetlana A. Chudina · 2023
Modern technologies are actively developing and are including in many fields. The spread of remote provision of most services requires increased attention to ensuring the security of user data, as there is currently an increase in cybercrime. Some companies in different fields and using different means use video identification for customer service. But the decisions of most of them have nothing to do with protection from the activities of fraudsters. And the most vulnerable to the actions of cybercriminals is the credit and financial sphere. A neural network-based solution has been proposed to protect the bank's customers from banking fraud through remote banking service. This paper describes the software product “DoubleCheck” which allows using video identification with high accuracy to confirm the identity of a client who wants to perform some banking operation. “DoubleCheck” is based on the Meso-4 neural network architecture and allows you to effectively distinguish the image of a bank customer from a DeepFake image.