摘要
随着智慧课堂和教育大数据挖掘的普及,创建一种智慧课堂中学生学习数据的自动检测和分析方法成为可能.基于YOLOv5模型实现两类学习数据:班级同学计数、学习行为识别.该模型对低头写字、低头看书、抬头听课、转头、举手、站立、小组讨论七种学生课堂行为进行识别,以此辅助教师判断学生学习情况并做出教学决策.研究表明,检测结果精确度达到97.92%.
Abstract
With the popularization of smart classrooms and educational data mining,our project aims to create an automatic detection and analysis method for student learning data in a smart classroom.Based on the YOLOv5 model,two types of learning data are implemented:classmate counting and learning behavior recognition.This model recognizes seven types of student class-room behaviors,which assists teachers in judging student learning situations and making pedagogical decisions,including writing with head down,reading with head down,listening with head up,turning,raising hand,standing,and group discussion.Research shows that the detection accuracy of the model reaches 97.921%.
基金项目
四川省大学生创新创业训练计划(X202210656224)
NETL
NSTL
万方数据