Implementation of Face Recognition, Attendance Detection, and Geolocation using TensorFlow Lite and Google ML Kit in a Mobile Attendance Application

Abstract

Attendance applications are increasingly needed across various sectors, including education, government, and offices. However, challenges such as identity fraud and location manipulation remain unresolved. This study aims to develop a mobile attendance system that integrates face recognition, liveness detection, geolocation, and permission features, along with push notifications via user email. Additional features such as profile updates and attendance history are also included in the mobile application in real-time. The system ensures valid attendance by utilizing TensorFlow Lite for real-time face recognition and liveness detection, and Google ML Kit to enable geolocation features for user location verification. The Waterfall method was employed in this study, covering analysis, design, implementation, testing, and finalization phases. The results of this study demonstrate the ease and convenience of using the attendance system, with successful integration of face recognition, liveness detection, and geolocation into the mobile attendance application.