A Teacher and Student Facial Expression Recognition Model Based on Classroom Teaching Videos

“Dear teacher make students believe his way”. Related literature research shows that positive emotions of teachers’ and students have a greater impact on the effectiveness of classroom teaching, and help to improve the learning efficiency of the students. After summarizing the domestic and foreign research status, this paper focuses on the realization of face recognition technology in AI evaluation of teacher teaching. The main research work is as follows. Firstly, this paper preprocesses the obtained classroom teaching video, including framing, removing invalid frames, etc.; followed, based on the Roboflow labeling tool, self-construct a teacher’s and student’s classroom teaching facial expression data set; then analyzes the characteristics of the general emotion data set; moreover, selects the general data set FER2013 as the training set; and the teacher’s facial expressions are divided into three categories: positive, negative, and neutral, which is exported as a verification set subsequently; Finally, a teacher and student expression recognition algorithm is designed based on deep convolutional network. This article mainly uses the classic VGGNet as the main body of the network, and simultaneously completes feature extraction and expression classification. Simulation experiments show that the model has a high accuracy in teacher’s and student’s FER.