Research On Emotion Of Digital Learning Interface In Intelligent Learning Environment

The existing intelligent learning environment pays more attention to the adaptive adjustment of knowledge content,learning path and other aspects,and less attention is paid to the personalized support of learners' emotion.In the learning process,learners lack effective emotional interaction with teachers or smart devices,and the problem of emotion loss still exists.In addition to providing learners with adaptive knowledge content,intelligent learning environment should also provide intelligent services that meet learners' emotional state,and can realize the adaptive adjustment of emotional state,so as to ensure that learners maintain a relatively positive emotional stage,so as to improve the learning effect.Therefore,the realization of emotional interaction between intelligent learning environment and learners is of vital practical significance to the occurrence of intelligent learning,and has become an important task to be solved urgently in the current construction of intelligent learning environment.In an ideal intelligent learning environment,learners acquire knowledge mainly by watching interfaces and pictures(i.e.digital learning pictures)rich in learning content displayed by smart devices.Meanwhile,digital learning pictures cause changes in learners' emotions due to their underlying emotional characteristics.The emotion of digital learning picture is usually conveyed by visual features,while the learner's emotion is the emotional state displayed by the learner when viewing the picture.However,how the two affect each other still needs to be further studied.For example,does the emotion of digital learning picture cause emotional changes in learners? How does it affect? And so on.Therefore,this research study by the digital images and emotional learning as the breakthrough point,to explore digital learning image emotional characteristics and the interaction between learners emotional relationships,in order to strengthen the wisdom of the learner's personalized learning environment emotional support,learning environment for wisdom emotional interaction theory and practice for reference.Fast and accurate recognition of emotions in digital learning images is an inevitable condition for subsequent studies.This study adopts deep convolutional neural network in the field of machine learning for automatic recognition of emotions in images.Before that,it is necessary to determine classification and emotion description models and carry out sample data collection of digital learning images.Therefore,through reviewing a large number of literatures,this study designed the classification and coding system of digital learning pictures according to certain purposes and principles,and divided the emotions of digital learning pictures into 14 types: warm,cheerful,lively,funny,exaggerated,humorous,funny,desolate,dull,dull,messy,unreal,thrilling,and scary.After that,the network crawler and screenshots were programmed to obtain the sample data required by convolutional neural network training in batches,and 126,297 digital learning images were finally obtained.Then,17,433 representative images were selected from 126,297 images to be manually labeled according to 14 emotions,so as to form the training data set and provide them to the convolutional neural network model.Using the convolutional neural network to identify the emotion of the picture can avoid the subjectivity of artificial recognition.In this study,a nine-layer convolutional neural network with four convolutional layers,four pooling layers and one fully connected layer was designed to identify the emotion of the digital learning picture.The convolutional neural network USES the digital learning picture marked with emotion to carry out the training,and the final test shows that the network has a good recognition effect.Then,the trained recognition model was used to analyze the emotion of all the collected images,and the emotion distribution level was analyzed in the dimensions of subjects,learning segments and teaching methods.In addition,through the way of facial expression recognition for the learners' emotional accurate classification smoothly for the future study are equally important,this study through the literature study expression can be divided into normal,happy,angry,sad,7 classes focus,daydreaming,etc,then on the basis of collecting a large number of facial expression data and construct of learners' emotional dynamic convolution neural network identification.The test results showed that both face recognition accuracy and expression recognition accuracy reached a high level,which proved that the model could quickly and accurately judge face regions and identify learners' emotional states.In this study,a total of 41 students majoring in digital media in a university in jinan were selected as experimental subjects to explore the influence of digital learning picture emotion on learners' emotion.Experiments designed using digital learning images in the process of teaching,and using the self-made synchronization acquisition in the process of lecture tool for learners in pairs expression image study with digital image,finally get learners 177515 face images,digital learning image 177515,two kinds of image one-to-one correspondence group stored in the database.After that,emotion recognition of digital learning images and learner emotion recognition of each group of images were carried out respectively.After generating the result data table,the influence of emotion of different digital learning images on learner emotion was discussed through correlation coefficient.The experimental results show that digital learning pictures with warm,cheerful,lively,funny,humorous and interesting emotions can arouse learners' positive emotions to a certain extent,while dull,dull and other emotions can arouse learners' negative emotions.Although the influence of exaggeration,dreariness,chaos and illusory types on learners' emotion is complicated.Finally,based on the above experimental conclusions,combined with the visual art characteristics and emotional characteristics of the picture,this study puts forward enlightenment and Suggestions on the design of digital learning picture,in order to enrich the theory and practice of emotional interaction in the intelligent learning environment.