Date of Publication :15th January 2025

Abstract: Integration of facial recognition technology with Automatic Teller Machines (ATMs) has substantially improved banking security. The main approach utilized by conventional ATMs, credit card cloning, skimming, and illicit PIN access are security concerns related with card-based authentication. By means of facial recognition technology, this work aims to offer a more foolproof and hassle-free biometric authentication solution. Thanks to the proposed system's usage of advanced facial recognition algorithms for user authentication, there are no more physical cards and reduced opportunity of fraud. Some of the factors investigated in this work apply to the construction of an effective architecture for including face recognition into present ATM systems: dataset selection, algorithm performance, and real-time processing capabilities. With a validation accuracy of 96%, extensive testing showed that the system exceeded present methods dependent on actual tokens. Evaluation criteria included processing time, True Positive Rate (TPR), and False Acceptance Rate (FAR) helped one to evaluate the dependability and performance of the system. The proposed method shows better in terms of accuracy, safety, and simplicity of use than conventi onal ones. This work expands the body of knowledge already in use by presenting a fresh approach for ATM authentication that provides great dependability and customer data protection. This solution has the power to transform the banking industry by giving customers a safer, more simple, and more quick access to their accounts.

Reference :