基于学习情绪面部特征识别的课堂教学智慧评价方法

Innovation of teaching evaluation method is an indispensable component of education modernization reform. To solve some problems encountered in most evaluation methods with stage examination, questionnaire survey, man observation and so on for classroom teaching effectiveness, an intelligent evaluation method of classroom teaching was proposed based on the learning-emotional facial feature recognition and deep learning technology, overcoming the limitations of the feedback delay of teaching situations, operation complexity, consuming human resources, and vulnerability to subjective factors. Firstly, a dataset of learning-emotional facial features for students was established, and different facial features were arranged to be recognized with deep network. According to the survey questionnaire results, a quantitative evaluation strategy of teaching effectiveness was subsequently developed based on the learning-emotional facial features, achieving objective and real-time feedback of classroom teaching effectiveness. Experimental results show the superior performance of the deep neural network “You Only Look Once (YOLO)” in this work than other comparative models on the students’ facial feature recognition, and demonstrate the capability of fast and high-precise recognition for students’ facial features.