SPEECH EMOTION RECOGNITION USING DEEP LEARNING

Abstract

Speech Emotion Recognition is a present topic of the research since it has wide range of application. SER is a vital part of effective human interaction in the speech processing. Speech Emotion recognition is a domain that is growing rapidly in the recent years. Unlike humans, machines deficit the potential to perceive and express emotions. But the improvisation of human-computer interaction can be done by automated SER thereby turn down the need of human mediation in recent time. The primary goal of SER is to improve man-machine interface. This paper covers Deep Learning to train the model, Librosa to classify the audio data. In deep learning CNN is used to classify the model based on frequency parameter. This paper also contains the study of various speech emotion recognition methods like, happy, sad, angry, disgust, surprise and fear.