This project is a take at the music genre recognition using machine learning, aimed at Indian music genres. The dataset used for the project was collected from various sources, and made into 30s .wav clips for training. They were converted into feature vectors uisng Fast Fourier Transform (FFT) method. Finally the model was trained using ANN, so that it could differentiate between various genres like Bolly Pop, Sufi, Ghazal, Carnatic and Hindustani Classical. Another script will be added which will make this project user interactive, i.e it will use the system's microphone as an input for genre recognition.
Author: Aayush Bhaskar.