This project uses Hidden Markov Models as a machine learning training algorithm to detect and classify sample input audio files into genres of Rock, Classical, Blues and country. Signal processing concepts including logfbank and mel filterbakn energies were used. The data used comes from GTZAN dataset.
A paper on this project can be found in Music Genre Detection using Hidden Markov Models (https://github.com/meghana-kantharaj/Music_genre_recognition/blob/master/Music%20Genre%20detection%20using%20Hidden%20Markov%20Models.docx)
A presentation describing the steps and execution of this project can be found here - https://www.slideshare.net/MeghanaKantharaj/music-genre-detection-using-hidden-markov-models