Research On Key Technologies Of Intelligent Teaching Management System Based On Micro-expression

The study of students' emotional changes and learning status in the distance teaching process has always been a research hotspot in the fields of psychology and computer vision.As an unrestrained and spontaneous facial expression,micro-expressions can better reflect the true thoughts of students,which is of great significance to the popularization of distance education and the improvement of the quality of distance education.However,when micro-expressions occur,the fact that their duration is very short and the involved muscle exercise intensity is also very low,making this problem difficult in the research field.This paper is proposed a micro-expression detection and recognition method based on local key region features to solve the problems in the actual application of the existing micro-expression detection and recognition metho'd based on global region features,and uses the proposed method to design and develop a micro-expression-based intelligent teaching,management system to effectively detect and recognize students' classroom emotional information in the distance teaching process.The main research work is as follows:(1)At present,the existing micro-expression detection and recognition methods based on global region features will reduce the ability of the described features to describe micro-expressions.In order to solve its existing problems,according to the coordinates of the face feature points detected by the subjective shape model and the motion features of the facial action coding system when the micro expression occurs Divide the 6 local key areas involved in the micro expression.On this basis,for the micro-expression detection part,the local binary pattern features of the divided local key areas are extracted respectively,and then' the feature difference analysis is performed on the proposed features to realize the detection of micro-expressions.For the micro-expression recognition part,first extracted the local binary pattern on three orthogonal plane features of the divided local key areas,and then connect all the features of the local key regions to form a feature vector,and finally support vector machines are used to classify the feature vectors to realize recognition of micro-expressions.(2)A method comparison experiment is carried out on the spontaneous micro-expression database CASME II.The experimental results show that the micro-expression detection and recognition method based on local key region features proposed is significantly better than micro-expression detection and recognition methods based on global region features,which fully illustrates that the motion characteristics of local key areas are involved,and it makes the proposed method more practical.(3)Being classified the classroom learning status into three dimensions of liking,devotion,and understanding based on domestic and foreign classroom psychology,combined with the seven basic micro-expressions defined in the micro-expression database CASME ?,the classroom emotions are evaluated in a comprehensive way.Based on the proposed method and emotion system,in order to make the system more complete,expanded the system functions to micro-expression automatic detection,and identification of micro expressions,automatic identification of picture expressions,automatic identification of video expressions,and counting of people.Finally,an intelligent teaching management system based on micro expressions is implemented.