Smart Real-time Attendance System for Nigerian Universities

This study proposes a Smart Real-Time Attendance System using face recognition technology to address challenges in traditional attendance systems in Nigerian universities. These challenges include proxy attendance, manual errors, and administrative inefficiencies. The system employs Convolutional Neural Networks (CNNs) and the ArcFace algorithm for facial feature extraction and identity verification. Key development tools included InsightFace, OpenCV, and Streamlit, with Visual Studio Code as the IDE. The system ensures high accuracy, with 94% face detection, 98% face recognition, and 96% overall attendance prediction accuracy. It automates essential tasks like attendance percentage calculation and report generation, ensuring compliance with the National Universities Commission (NUC) 75% attendance requirement for exam eligibility. Ethical compliance was a core design concern, including informed consent, data encryption, access control, and fairness across facial profiles. This system significantly reduces impersonation, administrative workload, and enhances operational efficiency, making it a scalable and secure solution for attendance management. Its deployment is recommended for improving academic monitoring and policy enforcement in Nigerian universities.