Physiological signals based emotional state recognition model of audience

Abstract

To study the relationship between movie plot and the physiological signals of audience,a physiological signals based emotional state recognition model of movie audience was proposed.Features were extracted from physiological signals,and sequential forward selection(SFS) method was used for feature selection purpose.The emotional state recognition model was built based on support vector machine(SVM).Experiments were conducted to evaluate the performance of the model.In the experiments,three movies with different types were employed.During watching these movies,the facial expressions and physiological signals of 11 participants were recorded,and the human judgment of the emotional states of the participants was obtained based on their facial expressions.The experimental results indicate that the proposed model can distinguish among various emotional states,and the average recognition rate is higher than 90%.