Automated Attendance System Using Multiple Facial Recognition

<p><span>The MITTrack app is a sophisticated face recognition-based intelligent attendance system that has been developed to mechanize the conventional manual attendance process in academic institutions. It integrates contemporary deep learning methods with mobile-friendly deployment to develop a fast, real-time solution. The system implements <span>YOLOv8</span> for rapid and precise face detection from classroom pictures, allowing it to process several faces at a time. After detecting faces, <span>MobileFaceNet</span> (executed through TensorFlow Lite) creates light facial embeddings on the Android device. These facial embeddings are transmitted to a <span>Flask-based backend server</span>, where pre-registered student information is compared with <span>DeepFace</span> to authenticate identities. On successful matching, attendance is marked and securely stored in a local <span>SQLite</span> database. MITTrack is optimized for usage in offline or local networks, with accessibility without relying on the internet. With over 95% accuracy, the system has provided a secure, contactless, and scalable means of tracking attendance that makes it well-apt for usage in classrooms, training centres, and business environments.</span></p>