| dc.description.abstract | Student attendance recording is an essential aspect of educational administration; however, manual methods are often slow, error-prone, and susceptible to manipulation. This study developed a real-time facial recognition-based attendance monitoring system utilizing the YOLOv8 algorithm for face detection and FaceNet for facial feature extraction, generating unique numerical representations for identification. The system is designed with attendance validation using smile detection based on facial landmark analysis to ensure the reliability of recorded data, which is automatically stored in a MySQL database. Key features include automatic attendance recording, real-time notifications sent to parents via Telegram bots, live attendance monitoring through an interactive dashboard, and daily report generation in PDF format. Additionally, the system supports various camera types, such as laptop webcams and ESP32-CAM, offering flexibility in usage. Implementation results demonstrated high accuracy, achieving 95% for face detection and 100% for student identification, along with significant efficiency improvements in attendance recording. The system provides enhanced transparency to parents through live monitoring features and automated reporting. With high accuracy, device flexibility, and additional features like smile-based validation, this system offers a modern solution for educational administration. This study is expected to support digital transformation, improve attendance recording efficiency, and be widely adopted by educational institutions for more effective and transparent management of student attendance data. | en_US |
