Comparison of Complex Emotion Detection Based on the Basic Emotion Model and the Complex Emotion Model

In this paper, a basic emotion recognition model is established based on the basic facial expression data set of RAF-DB, and a method is proposed to select two types of basic expressions corresponding to two maximum confidence levels to determine complex emotions. At the same time, a complex emotion recognition model was established based on the compound facial expression data set of RAF-DB, and a comparison of the recognition accuracy of complex emotions based on the basic emotion recognition model and the complex emotion recognition model was completed. The results show that the basic emotion model has higher accuracy in identifying complex emotions than the complex emotion model. In addition, the relationship between the detection accuracy of complex emotion and the spatial distance between two kinds of basic emotions in the three-dimensional PAD (Pleasure, Arousal and Dominance) space of complex emotion is explored, which provides a method for monitoring the learning state of students in the network training based on complex emotion recognition.