Music Genre Classifier

Classify your audio files on the go
Explore the Project »

View On Github · Report Bug · Request Feature

Table of Contents

1. About The Project
   • Project Overview
   • Built With
2. Getting Started
   • Prerequisites
   • Installation
3. License
4. Contact
5. Acknowledgements

About The Project

There are many great genre classifiers in the industry but how many of them are free, and reliable, and does the job fast, and are able to actually classify music ? I bet there are very few if not none that fits the criteria.

Some key features to note :

• You can classify any audio with just few clicks in realtime 😃
• You can analyse any audio's waveforms, FFTs etc.
• This app is cross platform so you can run it on any device (except on a rock 🎸)

Project Overview

1. Classification Genres : Classify any music into 10 genres. pop metal disco blues reggae classical rock hiphop country jazz

2. Dataset Used : Used GTZAN Dataset dataset by Andrada Olteanu from Kaggle.

3. Model Accuracy : Predict any genre with 72.6% accuracy.

Built With

Music genre classifier uses these technologies to run.

• Python3
• Flask
• Electron.js
• Tensorflow
• Librosa
• Matplotlib

Getting Started

Getting started with this project is very easy. Follown Prerequisites and installation guidelines.

Prerequisites

1. Clone the repo

   git clone https://github.com/anuragsati/music-genre-classifier.git

2. Download Processed GTZAN DATASET.

3. Paste your downloaded dataset (data.json) to /assets

4. Install NPM packages

    npm install

Installation

1. Run the backend server using (Necessary):

    python3 ./backend/main.py

2. Run the project using :

    npm run start

License

Distributed under the MIT License.

Contact

Anurag sati - [email protected]

Project Link: https://github.com/anuragsati/music-genre-classifier

References

• Audio Data Analysis
• Audio signal processing
• Deep Learning for audio
• Guide To The Fourier Transform
• Max Pooling
• Batch Normalization
• Sparse_categorial_crossentropy