Identification of Spoken Language Using Machine Learning

Abstract

Identification of spoken language is the way to detect the specific language which is spoken by an anonymous speaker. We will also find out several techniques of machine learning for detecting spoken language. Our major task is to identify parameters and features from spoken language that can be used to separate languages. To extract features from the audio file we will use the Mel Frequency Cepstral coefficient (MFCC). So far, many methods have been used for language identification (LID). Of all the techniques, the accuracy of machine learning is the best. That's why we also used machine learning in our project for lid. Our system will train with 30,000 data. This project aims to classify Spanish, German & English languages. The main goal of this project is to find out the best algorithm for detecting specific language. We get the best accuracy from the random forest algorithm.