Students' classroom Emotion Analysis Based on Intelligent Recognition

Abstract

In order to further accurately obtain the classroom learning status of online students, a multimodal emotion recognition method was proposed to improve the accuracy of students' classroom emotion recognition. Convolutional neural network was used as speech recognition and facial expression recognition methods. The decision-level fusion method based on sum rule was used as multimodal information fusion method. The simulation results showed that the average recognition rate of the speech emotion recognition method proposed in this paper increases with the increase of training samples. When the learning rate is 0.001, the average recognition rate of the model is the highest, reaching 80.1%. The facial expression recognition method proposed in this paper is the most accurate for the recognition of approval and distracted expressions. Compared with other models, the multimodal emotion recognition method based on sum rule and decision-level fusion proposed in this paper has the highest recognition rate, reaching 87.99%. The above simulation results verified the superiority of the design of this paper and the necessity of improvement, and had certain practical reference value.