<![CDATA[Student attendance recording is an important aspect in supporting discipline and administrative order in academic environments. Manual attendance methods still have several limitations, such as potential fraud and inefficient recapitulation processes. This study aims to develop a real-time face recognition-based student attendance system by implementing the YOLOv8 algorithm for face detection and ArcFace for identity recognition, complemented with an anti-spoofing feature to prevent fraudulent attempts. The system is designed to detect faces directly, recognize registered student identities, and record attendance automatically. The main contribution of this study lies in the integration of YOLOv8-based face detection, ArcFace-based face recognition, and an anti-spoofing mechanism into a single unified real-time attendance system. Experimental results show that the system successfully recognizes all registered students with a 100% success rate. The YOLOv8 anti-spoofing model demonstrates excellent performance in distinguishing real and fake faces, achieving an mAP@0.5 value of 0.995 and an F1-score close to 1. The system is also able to record attendance time in real time according to the actual time and present attendance data systematically. Based on these results, the developed real-time face recognition attendance system is accurate, secure, and feasible to be implemented as an attendance solution in academic environments]]>
Copyrights © 2026
