Abstract :

The rapid growth in academic and organizational environments demands a reliable and automated system for attendance monitoring, replacing outdated manual and biometric-based processes. This research presents a computer visionbased autonomous secure attendance system utilizing facial recognition for precise identification and verification. The proposed system leverages deep learning architectures for face detection, feature extraction, and classification to ensure high accuracy in varied lighting and environmental conditions. It eliminates proxy attendance and ensures data confidentiality through encryption and secure database handling. Real-time recognition enables efficient classroom or workplace monitoring without human intervention. The system incorporates liveness detection to defend against spoofing and unauthorized face replication. Extensive experimentation demonstrates improved performance over conventional systems. The proposed framework proves scalable, secure, and efficient