A Study of Behavioural Analysis Assisted Classroom Teaching Techniques Based on Multi-Representational Computer Vision

Abstract

Teachers primarily engage with students through classroom observation and questioning in traditional teaching methods, which will inevitably lead to one-sidedness and a lagging of information transmission and feedback because of things like a lack of personal energy. In contrast, artificial intelligence brings computer vision recognition into the classroom, which unquestionably increases teaching efficiency. The final teaching effect of the close relationship between the effective teaching strategies of the teachers and the good classroom behavior of the students can be enhanced by the positive classroom behavior of the students. Through the use of deep learning-based target recognition techniques, classroom camera visual analysis may be utilized to gather information on the teaching behaviors of students. This includes the identification and analysis of voice, posture, facial, physiological signal, and other data, to extract and define the distinctive behaviors of students. Lastly, the study demonstrates how computer vision applied to the teaching classroom can precisely identify and analyze students' emotional states and learning status, enabling teachers to adapt their instruction to the students' emotional states and learning status. In order to enhance teaching effectiveness and student progress, it assists educators in refining and optimizing their pedagogical approaches in accordance with the students' current circumstances. Consequently, there are numerous benefits to using computer vision recognition in the classroom.