This project is part of the Master in Music and Sound Computing at the Universitat Pompeu Fabra in Barcelona.
This repository shows a jupyter notebook implementing what is called "Audio content based analysis", in which I create a playlist based on certain audio related features. This means that I am making playlist using the "Collaborative Filtering" method.
The implementation of this project is clear. We have two parts: the jupyter notebook where all the features are extracted and the python script where an interface is implemented using streamlit in which playlists are generated from a dataset based on these features, combining them so that playlists of different styles and moods can be generated.