Research On Face Detection And Expression Recognition And Its Application In Classroom Teaching Evaluation

Class,as the most important occasion for teachers and students to learn and communicate,has always been widely concerned and valued by schools and society.However,the traditional classroom teaching often lacks communication and communication,only attaches importance to the teaching content of teachers and ignores the feedback information of students.In addition,students' emotions in the classroom can also reflect students' acceptance of the classroom to a certain extent,thus reflecting the quality of classroom teaching.Face detection is the premise and component of face recognition.Emotion recognition is the branch and development of face recognition.They are both important branches of computer vision technology.Face detection and emotion recognition technology based on deep learning has the characteristics of high accuracy and robustness.It can also show excellent accuracy and timeliness in large scenes such as classroom.In this paper,face detection and emotion recognition technology based on in-depth learning is studied,and its precise and real-time characteristics are utilized.It is applied to the evaluation of classroom teaching to solve the problem of backward and impersonal evaluation of classroom teaching.The main work of this paper is as follows:(1)Reviewing the development of computer vision face detection and face recognition technology,analyzing and absorbing advanced methods of face detection and emotion recognition,combining with the classroom teaching scene of our school,an efficient and accurate face detection method based on in-depth learning is proposed.This method uses convolutional neural network CNN as the main network,uses Pytorch as the main framework,which is one of the most popular deep learning frameworks recently.It combines RPN,Anchor,Attention Layer and other advanced methods commonly used in target detection,and combines affine,Delaunay triangulation and other methods commonly used in image science to analyze human faces,so as to achieve more efficient than traditional methods.Accurate results.On the basis of face detection,an emotion recognition and scoring method based on Naive Bayesian classification is proposed to classify and score the positive and negative facial emotions.(2)To analyze and summarize the pain points and shortcomings of current classroom education quality evaluation,especially the lack of interaction between teachers and students and students' attention.This paper makes a detailed investigation and analysis of the students' expressions in the classroom,and refers to the previous research results,sums up several common expressions in the classroom as training labels for experimental Expression Recognition.This paper analyses and summarizes the scoring of the degree of concentration corresponding to each expression,and takes it as a reference criterion for evaluating the quality of classroom teaching,and establishes an evaluation axis of the teaching quality of "student's rising rate-student's expression-student's concentration-classroom teaching quality".(3)At the end of this paper,we use the data collected by mobile camera in the classroom to test and analyze the system,verify the feasibility and accuracy of the system,and analyze the advantages and disadvantages of the system and the areas for improvement.