Research On The Differences Of Emotional Speech Characteristics Of Introverted And Introverted Individuals And Machine Recognition Of Emotional Speech

Emotion is people’s psychological state and the resulting coping behavior when facing the objective situation,which is very important to the survival and development of individuals.Therefore,how to effectively identify the emotional state and control and regulate the emotion is of great significance to the development of individuals.With the development of computer science and technology and the deepening application of artificial intelligence and human-computer interaction,the research on emotion recognition came into being.Emotion recognition is based on emotion calculation.Different emotional states will lead to changes in physiological characteristics.Emotion recognition calculates individual emotional states according to these changes.Emotion recognition has a wide range of applications,such as psychological counseling,emotion detection,mental health monitoring and so on.Emotion recognition is also the basis of real-time emotion regulation.The purpose of this study is to establish an emotion data sample data set containing personality characteristics and emotional characteristics,explore the differences of speech characteristics of different personality individuals under different emotional priming,and combine machine learning technology and methods to improve the accuracy of emotion recognition based on speech characteristics.The experiment recruited 80 college students as subjects（40 males and 40females）.The experiment adopted a 2x4 two factor experimental design.The independent variables were personality type and emotional priming type,and the dependent variables were speech characteristics.Among them,personality was divided into two levels（introversion high score and introversion low score）,and emotional priming status was divided into four levels（neutral emotional priming,happy emotional priming,sad emotional priming and angry emotional priming）.Emotional priming is carried out through video.The types of emotional priming are neutral,happy,sad and angry.Collect the voice information of the subjects immediately after the emotion starts,and then complete the emotion self-assessment scale.Reprocess the collected speech information and extract the speech feature parameters.Analyze the phonetic differences of different personality individuals under different emotional priming.Speech emotion recognition through machine learning.The experimental results show that there are differences in speech feature parameters among individuals with different personality types.There are significant differences in fundamental frequency mean,fundamental frequency dispersion and energy mean between individuals with low extroversion and extroversion and individuals with high extroversion and extroversion.In different emotional priming States,individual speech has differences in speech feature parameters.Individuals in different emotional priming states have significant differences in fundamental frequency mean,fundamental frequency dispersion and energy dispersion parameters.There are significant differences in the speech characteristics expressed by low introversion and high introversion groups under different emotional priming States,that is,there is a significant interaction between personality types and emotional priming types.Adding personality type factors to emotional speech machine recognition can improve the accuracy of emotional recognition.