YOLOv5能够快速、准确地检测目标,但检测精度和召回率较低,难以适用于复杂场景.基于此,文章提出一种基于改进YOLOv5的校园课堂实时监测人员方法.引入挤压激励(Squeeze and Excitation,SE)注意力机制,采用非最大抑制(Non Maximum Suppression,NMS)改进,优化Mosaic增强数据对YOLOv5改进.实验结果表明,该方法的精准度、召回率、平均精度均值均优于对比模型,能够满足实际场景对校园课堂的检测要求.
Research on Real-time Detection Personnel for Campus Classrooms Based on Improved YOLOv5
YOLOv5 can quickly and accurately detect targets,but its detection accuracy and recall rate are low,making it difficult to apply to complex scenes.Based on this,the article proposes a real-time monitoring personnel method for campus classrooms based on improved YOLOv5.Introducing Squeeze and Excitation(SE)attention mechanism,improving YOLOv5 with Non Maximum Suppression(NMS)and optimizing Mosaic enhanced data.The experimental results show that the accuracy,recall,and mean Average Precision of this method are superior to the comparison model,and can meet the detection requirements of campus classrooms in practical scenarios.
万方数据
维普
