<![CDATA[Manual attendance systems in student organizations are highly vulnerable to data manipulation, recording errors, and slow compilation. This study aims to design and implement a secure, accurate, and real-time digital attendance system using face recognition and passive liveness detection based on the YOLOv11 deep learning algorithm, integrated with the Laravel web framework. Following the structured Waterfall model, this research focuses on a population of 50 active members of Tapak Suci at Universitas Muhammadiyah Kudus, with a test sample of 20 individuals selected through purposive sampling. The system architecture is organized into five functional layers: Input, Processing (comprising YOLOv11, ArcFace, and Cosine Similarity), Application, Data, and Output. The system evaluation demonstrates excellent biometrics performance, achieving an overall system accuracy of 95.0%, a False Acceptance Rate (FAR) of 1.5%, a False Rejection Rate (FRR) of 5.0%, and an average processing latency of 1.2 seconds. Furthermore, the passive liveness detection module successfully mitigated up to 97.5% of presentation spoofing attacks, including high-resolution printed photos and video replays. The developed system effectively delivers an efficient, transparent, and highly secure biometric attendance solution to replace conventional paper-based methods.]]>
Copyrights © 2026
