A method for automatic detection of teacher's gaze area based on smart classroom system.

The quality of teaching is closely related to the ability of the teacher, which is still mostly assessed in the traditional way. With the development of artificial intelligence, smart classroom systems are beginning to appear, which can quantify and analyze the teaching data in the classroom to produce corresponding results for teaching management or teaching evaluation. Therefore, this paper proposes an automatic method for detecting the teacher's position and gaze area based on the smart classroom system. The method mainly uses adaptive area face detection technology, face feature point matching, head pose acquisition and teacher gaze point calculation algorithms to automatically obtain the teacher's position and gaze area distribution in the classroom. Through experiments and public classroom tests, it was found that the maximum mean error in teacher position obtained by this method was 11.72cm and that the teacher's area of gaze was largely consistent with the actual change. This method can also be used to assess teaching ability and quality, and the data obtained can also be used as a learning resource for novice teachers or average students, so that they can continuously improve their teaching ability.