【摘要】 针对高校教学评教的不足,采用深度学习算法分析智慧教室中由跟随录播摄像头采集到的教师上课视频。通过人脸检测和表情识别技术,分析教师各类表情所占比例随上课时间的变化,进而理解教师上课情绪的积极程度、课堂气氛活跃程度等教学情境。结果表明:该研究成果可以为教学评教提供参考,也可以用于分析学生对教学内容的感兴趣程度。更多还原

【Abstract】 Aiming at the shortage of teaching evaluation in university,we use deep learning algorithm to analyze the teaching video captured by follow-up video cameras in smart classrooms. By using face detection and expression recognition technology,we analyze the proportion of teachers’ facial expression varying with teaching time,so that we can understand the teaching situation of teachers’ positive mood and the activity of atmosphere in class. The results showthat the study can provide reference for teaching evaluation and be used to analyze students’ interest in teaching content.更多还原