MUSDB18 - a corpus for music separation - Archive ouverte HAL Access content directly
Other Publications Year : 2017

MUSDB18 - a corpus for music separation

(1) , (2) , (2) , (3) , (4)
1
2
3
4

Abstract

The sigsep musdb18 data set consists of a total of 150 full-track songs of different styles and includes both the stereo mixtures and the original sources, divided between a training subset and a test subset. Its purpose is to serve as a reference database for the design and the evaluation of source separation algorithms. The objective of such signal processing methods is to estimate one or more sources from a set of mixtures, e.g. for karaoke applications. It has been used as the official dataset in the professionally-produced music recordings task for SiSEC 2018, which is the international campaign for the evaluation of source separation algorithms.
Fichier principal
Vignette du fichier
musdb.pdf (141.8 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-02190845 , version 1 (30-07-2019)

Licence

Copyright

Identifiers

Cite

Zafar Rafii, Antoine Liutkus, Fabian-Robert Stöter, Stylianos Ioannis Mimilakis, Rachel Bittner. MUSDB18 - a corpus for music separation. 2017, ⟨10.5281/zenodo.1117371⟩. ⟨hal-02190845⟩
1014 View
857 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More