Face Recognition Attendance System with Anti-Spoof Technology

Abstract

Marking attendance manually takes time and often leads to errors and proxy attendance. To overcome these problems, many institutions started using biometric systems. Among all biometrics, face recognition is the most user-friendly because it works without physical contact and can identify multiple people at the same time. However, a normal face recognition system can be easily fooled using printed photos, mobile phone images, or video playback. This project presents a smart attendance system that combines face recognition with anti-spoofing technology to ensure that only real and live users are marked present. The system captures the face using a camera, checks whether the face is real or fake using liveness detection, and then matches it with the stored database. If both conditions are satisfied, attendance is recorded automatically and stored in the system. The proposed system reduces manual work, prevents fake attendance, and provides real-time attendance monitoring through a web interface. The performance of the model shows high accuracy in identifying real users and detecting spoof attempts.