LivePresence Attendance: Encrypted, Anti-Spoof and Liveness-Verified System

Abstract The increasing demand for secure, contactless, and automated attendance systems has led to significant research in biometric authentication technologies. However, traditional attendance systems suffer from limitations such as proxy attendance, spoofing attacks, and lack of secure audit mechanisms. This paper presents LivePresence Attendance, a real-time face recognition-based attendance system integrated with liveness detection, anti-spoofing mechanisms, and encrypted logging. The system employs a multi-stage verification pipeline consisting of face detection, behavioral liveness verification using blink and head movement analysis, identity recognition through embedding comparison, and secure database storage. A challenge-based liveness mechanism ensures that only physically present users can authenticate, effectively preventing spoofing attacks using printed images, digital screens, or replayed videos. Experimental evaluation under real-world conditions demonstrates a recognition accuracy of 92.8%, liveness verification success rate of 94.1%, and an average system latency of 320 ms, making the system suitable for real-time deployment. The system is implemented using Python, OpenCV, MediaPipe, and SQLite, with AES-based encryption ensuring data confidentiality. The proposed system provides a secure, scalable, and practical solution for attendance management in academic and organizational environments. Keywords: Face Recognition, Liveness Detection, Anti-Spoofing, Secure Systems, AES Encryption, Computer Vision