Music source separation conditioned on 3D point clouds

Francesc Lluís, Vasileios Chatziioannou, Alex Hofmann

Recently, significant progress has been made in audio source separation by the application of deep learning techniques. Current methods that combine both audio and visual informati…