Music Genre Classification using CNN

Dataset Heirarchy

gtzan_dataset
	\blues
	\classical
	\country
	\disco
	\hiphop
	\jazz
	\metal
	\pop
	\reggae
	\rock

Run order of files:

python3 data_augmentation

Increases the size of dataset from 1000 to 10000.

python3 wav_convert.py

Converts music files to the needed .wav format and store into ./converted.

python3 wav_extract_fft.py

Extracts fft values and saves numpy files to ./fft_extracted.

python3 cnn_fft.py fft_extracted/

Applies CNN to train and classify; gives prediction accuracy, saves the model.

python3 test.py /home/path/to/your/song/file

Loads saved model to predict genre of new file.