Gender Identification and Population Detection in a Room Using YOLOv8

Abstract

In today's digital era, lectures and higher education have experienced rapid development, especially in the
use of technology. Technology has opened up various opportunities to improve efficiency and effectiveness in various
aspects of education. One aspect that needs to be improved is the presence of students in lectures. Through this thesis,
the author aims to develop a tool that can overcome the problem of recapitulation of student attendance using image
processing technology and the YOLOv8 algorithm, then adding male and female face training data to identify
gender. By combining the latest technology and this innovative approach, it is hoped that an efficient and accurate
solution can be created to record student attendance in lectures. In this study, the implementation of population count
and gender identification within a room using YOLOv8 achieved a precision value of 100%, a recall value of 100%,
an accuracy value of 100%, and an F1-score value of 100%.