An Online Classroom Atmosphere Assessment System for Evaluating Teaching Quality

The classroom atmosphere is the most direct reflection of the teaching quality. Rapid receipt of the response from students about their understanding during a lecture can help them learn more efficiently. In this paper, we propose an online classroom atmosphere assessment system based on deep learning technology, which can significantly increase the interactivity of teaching and learning. Firstly, separate the video stream to get the picture frame sequence. Then, extract the valid data by the classroom analysis module, Mask R-CNN model marks the face region, and the eigenvalue of the face is obtained through the XCeption framework. After that, the SVM model is used for analyzing the emotions of the students' faces, and the weight model is used to parse the current students' learning State. Finally, use the assessment strategy to get the classroom atmosphere information. The teacher can grasp the classroom atmosphere in real time, to facilitate the adjustment of the teaching strategy in time. Empirical experiment results show that the proposed system has acceptable performance. Also, the system improves the teaching quality, and students can then acquire a much better learning experience and satisfaction.