Restore, Separate, Restore: A Modular Framework for Music Source Restoration

Tobias Morocutti, Emmanouil Karystinaios, Gerhard Widmer
October, 2026
PDF Code

Abstract

A modular system restores a degraded mix before separation and then applies stem-specific restoration to the resulting instrument tracks.

Type
Preprint
Publication
arXiv preprint
Audio Processing Music Source Separation Restoration
Emmanouil Karystinaios
Emmanouil Karystinaios
Senior Machine Learning Engineer and Postdoctoral Researcher

I work in music information retrieval as a Senior Machine Learning Engineer at Muse Group and a part-time Postdoctoral Researcher at Johannes Kepler University Linz.