Research on classroom student behavior recognition based on improved YOLOv8s
In order to effectively identify students'behavior in real class,a classroom behavior recognition method based on improved YOLOv8s model is proposed.The lightweight Coordinate Attention mechanism is integrated into the YOLOv8s backbone network to improve the feature learning ability of the model.In the feature fusion module,the weighted bidirectional feature pyra-mid network BIFPN and the heavily parameterized module Diverse Branch Block were used to improve the feature integration capa-bility of the model by using the feature fusion module Diverse.Experimental results show that the average accuracy of the improved YOLOv8s-CB model is 1.7 percentage higher than that of the original model,reaching 92.4%,indicating that the algorithm has greater advantages in real-time detection of classroom student behavior recognition tasks.
万方数据
维普
