Music Source Restoration with Ensemble Separation and Targeted Reconstruction

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Deng, Xinlong, Xia, Yu, Jiang, Jie
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915870709645312
author Deng, Xinlong
Xia, Yu
Jiang, Jie
author_facet Deng, Xinlong
Xia, Yu
Jiang, Jie
contents The Inaugural Music Source Restoration (MSR) Challenge targets the recovery of original, unprocessed stems from fully mixed and mastered music. Unlike conventional music source separation, MSR requires reversing complex production processes such as equalization, compression, reverberation, and other real-world degradations. To address MSR, we propose a two-stage system. First, an ensemble of pre-trained separation models produces preliminary source estimates. Then a set of pre-trained BSRNN-based restoration models performs targeted reconstruction to refine these estimates. On the official MSR benchmark, our system surpasses the baselines on all metrics, ranking second among all submissions. The code is available at https://github.com/xinghour/Music-source-restoration-CUPAudioGroup
format Preprint
id arxiv_https___arxiv_org_abs_2603_16926
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Music Source Restoration with Ensemble Separation and Targeted Reconstruction
Deng, Xinlong
Xia, Yu
Jiang, Jie
Sound
Artificial Intelligence
Audio and Speech Processing
The Inaugural Music Source Restoration (MSR) Challenge targets the recovery of original, unprocessed stems from fully mixed and mastered music. Unlike conventional music source separation, MSR requires reversing complex production processes such as equalization, compression, reverberation, and other real-world degradations. To address MSR, we propose a two-stage system. First, an ensemble of pre-trained separation models produces preliminary source estimates. Then a set of pre-trained BSRNN-based restoration models performs targeted reconstruction to refine these estimates. On the official MSR benchmark, our system surpasses the baselines on all metrics, ranking second among all submissions. The code is available at https://github.com/xinghour/Music-source-restoration-CUPAudioGroup
title Music Source Restoration with Ensemble Separation and Targeted Reconstruction
topic Sound
Artificial Intelligence
Audio and Speech Processing
url https://arxiv.org/abs/2603.16926