CSF425_Major_Project

Problem Statement:

To create a generative model (either a GAN or a VAE) that would generate an audio clip based on an input parameter that specifies the genre of the audio clip. For testing purposes, we shall be considering samples of the genre ”classical”, as they are the easiest to generate (as they do not have any vocals)

Here, we have used VAE as a generative model

Dataset used:

GTZAN Dataset

To run the ipynb file:

1. Go to Kaggle
2. Then, Accounts
3. Then, Create new API Token
4. Upload the downloaded kaggle.json file to the notebook

Contributors:

Harsha Vishak : 2019A7PS0079G

Aastha Rastogi : 2019A7PS0175G

Athreya Krishnan : 2018B4A70809G