Speech Emotion Recognition using MLP Classifier

Abstract

As human beings speech is natural way to express ourselves. Humans depend so much on it. Emotions play a important role in communication . Detection and analysis of emotion is very important in today’s digital world.Emotion detection is a challenging task. There is not a general agreement on how to measure or categorize them. Speech Emotion Recognition process and classify speech signals to detect emotions embedded in them. Speech Emotion Recognition system can be used in various areas.The application area are like interactive voice based-assistant , caller agent conversation analysis,security and other fields. This System attempts to detect emotions in audio file passed by analysing the acoustic features. System uses MLP Classifier to classify the emotions from the given wave signal. RAVDESS dataset will be used .The features to be extracted from the audio input provided will be attracted by these five parameters which are as follows, MFCC, Contrast, Mel Spectrograph Frequency, Chroma and Tonnetz.