Facial Recognition with Neural Networks for Access Control to restricted areas

Abstract

Access control to restricted areas is facing issues with its identity verification systems for Chief Executive Officers (CEOs) based on access cards, which are vulnerable to cloning, identity theft, and cyberattacks like Whaling. To address these concerns, a facial recognition system was developed using transfer learning and the face-recognition library. The methodology for implementing the facial recognition system was based on Scrum and Kanban, and Python was used as the programming language. As a result, the system successfully extracts key facial features and compares them with those stored in a database, while also notifying of failed attempts by unauthorized personnel. In conclusion, the implementation of an intelligent facial recognition system has proven to be an effective solution for access control in restricted areas, verifying the identity of CEOs stored in a database. © 2024, Associacao Iberica de Sistemas e Tecnologias de Informacao. All rights reserved.