Face Biometric Authentication System for ATM using Deep Learning

Abstract

Automated Teller Machines (ATMs) are frequently used in recent years. The ATM has become insecure as the number of crooks and their actions has increased dramatically. This study presents a security architecture for automated teller machines that uses a Deep Convolutional electronic facial recognition. In the financial sector, there is a persistent need to improve security. For identity verification, ATM systems now just require an access card and a PIN. Recent advances in biometric authentication systems, such as fingerprinting, retina scanning, and facial recognition, have gone a long way in resolving the ATM’s dangerous concern. The proposed work is designed with a security paradigm for automated teller machines that uses a Deep Convolutional Neural Network to integrate an access card physically and facial recognition electronically. Faces, as well as their accounts, would be secured if this technique became widely used. For remote certification, a verification portal for face identification is developed and provided to the end user to validate the authorization of an illegal user using certain full-fledged agents of artificial intelligent. It is a challenge to identify the man’s biometric traits which cannot be duplicated. Thus the system will go a long way towards solving the account security paradigm by allowing the actual account to use.