Research on Student Classroom Behavior Recognition Based on Improved YOLOv5 Algorithm
In recent years,with the vigorous development of artificial intelligence technology,significant progress has been made in education reform.To better study the classroom state of students and promote the high-quality development of the college,a method based on deep learning for intelligent recognition of student classroom behavior is proposed.Due to the poor robustness and low accuracy of traditional student behavior recognition methods.This article is based on deep learning,annotating and constructing a dataset of student classroom behavior.Based on YOLOv5,CBMA attention mechanism is introduced to effectively extract student classroom behavior features from both channel and spatial dimensions,enhancing the robustness of the model.The experiment shows that compared with the YOLOv5 model,the model with attention mechanism significantly improves the accuracy of classroom behavior recognition for three types of students:reading,raising hands,and standing.
Deep learningBehavior recognitionYOLOv5Attention mechanism
NETL
NSTL
万方数据
