Faceauth: Sistema de Autenticação Facial IoT

Abstract

The increasing use of Internet of Things (IoT) devices in critical applications requires authentication methods that are both robust and accessible, considering the processing and storage limitations of these devices. The objective of this work is to develop a facial authentication system integrated into an IoT architecture. The methodology includes the use of the ESP32-CAM for facial image capture, Supabase for data management and authentication, and Next.js as the front-end interface. The facial authentication system achieved an average accuracy of 91.46%, with 95% precision at distances of 30 cm and 60 cm, dropping to 85% at 1.30 m. The average response time was 7.31 seconds for registered users and 10.4 seconds for unregistered users. These results suggest that the facial authentication system is viable for practical applications in IoT environments, indicating its ability to operate even under the constraints of resource-limited devices.