Research On Real-time Intervention In Negative Emotional Contagion In The Classroom Based On Computer Vision

Emotion is one of the important factors that affect students' learning.Research shows that negative emotions may reduce students' learning efficiency and engagement,or even break away teaching activities.And the negative emotions are similar to infectious diseases,which can infect the surrounding students through facial expressions,face-to-face communication and other ways,so that more students are in a negative emotional state.However,there are few methods for intervention in negative emotional contagion,which are mostly focused on the human intervention of teachers and cannot achieve real-time intervention.It's an urgent problem to intervene negative emotional contagion in the classroom.In order to solve the problem of negative emotional contagion in the classroom,this paper proposed a real-time intervention method based on computer vision.The main work is divided into four parts: Image super-resolution,academic emotion recognition,negative emotional contagion model and the system for real-time intervention in negative emotional contagion.In the classroom,a high-definition camera is used to collect students' emotional images,and image super-resolution technology is used to improve the quality of the low-resolution images of back row students,the algorithm of academic emotion recognition is used to recognize students' academic emotions in real time,and then the negative emotional contagion model is used to analyze the source of negative emotional contagion,and finally intervention is implemented to prevent negative emotional contagion.The specific research work is as follows:a)Research on image super-resolution method based on WDSR.When the camera collects the images of the students in the back row,there will be distortion,which will affect the effect of emotion recognition in subsequent studies.First,the training data set for face reconstruction of students is constructed.Then,according to the work needs of this paper,the super-resolution model of image is designed based on WDSR model.b)Research on the method of Academic Emotion Recognition Based on convolutional neural network.First of all,by analyzing the existing emotion recognition work,the concept of Academic Emotion score is proposed to describe students' emotion in a fine-grained way.Secondly,based on the Fer2013 emotion data set,the training data set suitable for classroom emotion score recognition is constructed by re labeling and preprocessing the data set.Finally,referring to the existing neural network model,a convolution neural network is constructed to recognize Academic Emotion score,and the optimal model is obtained by adjusting the network results and training parameters through experiments.c)The model of negative emotion infection in classroom.Based on the analysis of the current situation and shortcomings of the research on the emotional infection model,this paper puts forward a negative emotional infection model suitable for classroom scenes to realize the real-time location of the source of negative emotional infection.d)Real time intervention of negative emotion infection system in classroom.Integrate the above research work,build a real-time intervention classroom negative emotion infection system based on computer vision,through real-time monitoring of students' Academic Emotion score,positioning negative emotion infection,and timely implementation of intervention.In this study,the effectiveness of the relevant algorithm model is verified,and the real-time intervention of classroom negative emotional infection system is applied to the actual classroom to test its effect.The experimental results show that the image super-resolution model can improve the image resolution under the premise of twice the size of the enlarged image,and the recognition accuracy of the emotion recognition algorithm reaches 99.4%.And the real-time intervention system of negative emotion infection can effectively reduce the number of negative emotion students and block the negative emotion infection.This paper combines artificial intelligence technology with educational scene,which provides technical support for intelligent education on the one hand,and provides guarantee for students to learn and improve better on the other hand.At the same time,it provides a new perspective for the study of negative emotion infection in the classroom.