Analysis And Evaluation Of Classroom Concentration Based On Facial Expression And Head Pose Recognition

In student-oriented class teaching,it's particularly important for the optimization of the teaching process to know the situation about students' class learning status and knowledge acquisition.However,due to the large number of students in class,the traditional method of teacher observation and evaluation only cannot grasp and feedback the student's learning situation in a timely and comprehensive manner.New technologies such as computer vision and artificial intelligence,which are gradually emerging and developing,can improve this problem to a certain extent,and provide the possibility for intelligent application in education.This article aims to use the technology of intelligent detection and recognition of class expressions and head postures based on the deep learning,and combine class concentration and personalized learning theory to realize automatic analysis and evaluation of class concentration evaluation,in order to provide effective assistance for teachers to carry out academic situation analysis,improve teaching methods and achieve personalized education.The main research contents of this article read as follows:(1)Adopt TensorFlow+python-based learning framework,to detect and locate the faces of students in class in real-time,and realize the classification and recognition the 7 types of facial expressions appearing in class.At the same time,by using OpenFace2.0 toolkit,detect and recognize real-time head postures of students in class,and obtain deflection angles that appear in class.(2)Based on the students' facial expressions and head postures,a method to evaluate the class concentration was designed.According to the 7 different expressions in class and the varying amplitude of head tilt and deflection angle,the method firstly ranks the degree of concentration and assigns the weight,then the matrix algorithm of fuzzy comprehensive evaluation is used to assess the score of class concentration,thereby fuzzy problems like concentration can be handled accurately.(3)The visual platform,which uses Django as the website design framework,Python as the main development language design,impletment the attention evaluation system visual platform.The concentration score obtained in the fuzzy comprehensive evaluation is displayed to students and teachers in the form of a line chart,and specific feedback suggestions are given based on it.At the same time,a video review function is supported to make it easy for teachers and students to analyze the cause of the problem and make timely improvement.(4)Collect the class video data which take the students of YunNan Nomal University as the research objects,and analyse and evaluate students' individual and whole concentration in different periods of time.By designing the class concentration assessment scale for teachers to perform manual assessment,the correlation between the results and the systematic assessment of class concentration scores assessed by the system developed in this paper verifies the validity of the data results.