Teacher Tracking in the Classroom Based on YOLO and DeepSORT

With the rapid development of computer vision technologies, these technologies have been applied in the analysis of classroom teaching videos. The classroom is a complex environment, and teachers play a leading role in this real-life scenario. The location of the teacher in the classroom can impact students' learning efficiency. To enhance teaching effectiveness, this paper proposes an automatic tracking method of teacher whin classroom based on YOLOv5 and DeepSORT. Teachers are detected in instructional videos, and their location in the classroom was automatically captured. Mapping the tracked location information through camera ranging to various areas in the classroom for statistical analysis. The average accuracy of the experiments in this paper is 85.33%. The experimental results indicate that the method is effective. The tracking method can make teachers enhance their teaching efficiency and increasing student engagement in the classroom.