Speech Emotion Recognition: Node.js and Web Speech API

Abstract

Speech Emotion Recognition (SER) is a field of study that seeks to extract the emotion from speech signals. It is one of the most challenging tasks in the speech signal analysis domain. In this paper, we are going to present a web application using Web speech API, AFINN, Node.js. In this model the Web speech API available in Google which makes the web apps to handle voice data and it also helps us to convert the speech to text and make a request to the Node.js server with the text, Node.js is a server which helps us to evaluate the text using AFINN's list to return the score, where AFINN is a word list based approach for sentiment analysis. It evaluates the part where emotion is present and after the evaluation of the text the browser displays a different emoji depending on the score. We are building this model so that the user/call center employee can recognize customer's emotions from speech and can improve their service and converse with more people.