Opens in a new tab

AES E-Library

← Back to search

Engineering Brief

Single-Channel Sound Source Separation Using NMF with Sparseness Constraints

Authors: Geng, Shijia; Leider, Colby N.

AES Convention 139 · Paper 231 · October 2015

Abstract

While challenging, sound source separation is a task that has many practical applications in audio signal processing. In this paper three sound files with two sources in each were separated using the non-negative matrix factorization (NMF) approach, with and without sparseness constraints. The results showed that adding sparseness constraints had no effect when separating drums and bass guitar but had better performances when separating piano and drums, and piano and bass guitar.

Details

Published in
AES Convention 139
AES Convention
139
Paper number
231
Publication date
October 6, 2015
Affiliation
University of Miami, Coral Gables, FL, USA (See document for exact affiliation information.)
Type
Engineering Brief