Abstract
Attendance is an important criterion for passing courses at Universitas Dinamika Bangsa Jambi. According to the academic regulations of Universitas Dinamika Bangsa Jambi, the minimum attendance requirement for course completion is 75%. The attendance process at the university utilizes an academic information system (SIAKAD) where students log in using a username and password, then scan an attendance barcode or input a unique code. Students often engage in proxy attendance practices, where they are marked present in the system despite being absent in reality. This study discusses the prevention of proxy attendance by employing a human detection system based on YOLOv11, capable of counting the number of students present in the classroom at Universitas Dinamika Bangsa Jambi. The research method involves the design, implementation, and evaluation of the system. This study adopts a deep learning approach using supervised learning methods for model training. The model is trained on a labeled dataset from Roboflow and implemented using the YOLOv11 algorithm. Based on the research results, the authors conclude that the human detection system is effective in counting the number of students in the classroom. However, the system still requires further development to detect criteria or features that can distinguish the detected individuals' status, specifically between students and lecturers.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call