Speech emotion recognition using machine learning

Abstract

Humans connect to each other through language. Verbal words play an important role in communication. The project works on determining an emotion behind verbal words. Speech Emotion Recognition is a system where we determine emotions from live audio. People from all around the globe use speech to convey their emotion irrespective of their background. Emotion recognition from human speech is challenging as there are many factors which play important role in formation of an emotion. It is one of the growing fields in interaction of machine and human. Majorly sub-domains of artificial intelligence are used in the task of prediction. Machine learning is used in the project. Machine learning (ML) uses a dataset and algorithm to predict or detect any future possibility. In this project we propose the application of Artificial Neural Network to determine emotion. Artificial Neural Network is based on how biological brain work. It has neurons which are connected to each other and are called nodes. The Classifier used in this project is Multilayer perception (MLP), Decision tree classifier, support vector machine (SVM), random forest classifier. Speech Emotion recognition using machine learning have certain steps to attain result. Firstly we need a dataset to train the program.