We did not include this in our project description, but our genre classification pipeline uses a Kaggle dataset containing spectrogram features: https://www.kaggle.com/andradaolteanu/gtzan-dataset-music-genre-classification. A source we also used for genre classification was https://gist.github.com/parulnith/7f8c174e6ac099e86f0495d3d9a4c01e#file-music_genre_classification-ipynb. We used this to help choose model architecture and preprocess data.
Log in or sign up for Devpost to join the conversation.