Skip to content

KunalVaidya99/Music-Genre-Classification

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

78 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Music-Genre-Classification

Music Genre Classification using GTZAN Dataset.Here is the Link to Dataset GTZAN Dataset

I Made a Web App using Streamlit For the Convolutional Neural Network (CNN) Model and Deployed it on Amazon EC2 Streamlit App

The Dataset contains 10 Genres of music :

Genres

  • Blues
  • Classical
  • Country
  • Disco
  • Hip-Hop
  • Jazz
  • Metal
  • Pop
  • Reggae
  • Rock

Dataset

  • Each Genre contains 100 audio files each of duration 30s.
  • Divided 30s audio files into 10 files of 3s each.
  • This Data Augmentation really helps model learn better and it performes better on the test set.

Models

I have explored the problem using two approaches

  1. Using Deep Neural Network (DNN) Model which made use of features such as MFCC's,spectral centroids, extracted features are in features_3sec.csv
  2. Using Convolutional Neural Network (CNN) Model which made use of Mel Spectrogram of the Audio Files.

Results

Training Loss Plot

So this the training and test accuracy obtained after 250 epochs. A simple DNN model is also showing quite good test accuracy due to data augmentation