A two-step NMF based algorithm for single channel speech separation

Shuo Wang; Wenjun Wu

A two-step NMF based algorithm for single channel speech separation

Shuo Wang
, Wenjun Wu

School of Artificial Intelligence(Institute of Artificial Intelligence)

Beihang University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Nonnegative Matrix Factorization (NMF) has become an increasingly popular method in the field of non-stationary speech denoising. However most NMF-based algorithms assume prior knowledge about the background noise, which is often not available in time-varying and mobile environments. In this paper, we propose a two-step NMF based speech-noise separation algorithm to address this issue. This algorithm takes the outcome of the first NMF separation as the dataset to train the basis vectors for the background noise, which will be used for the second-step NMF separation with fixed speech and noise basis vectors. Experimental results show that the proposed algorithm could achieve better results than other NMF algorithms for speech-noise separation.

Original language	English
Title of host publication	13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012
Pages	1987-1990
Number of pages	4
State	Published - 2012
Event	13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012 - Portland, OR, United States Duration: 9 Sep 2012 → 13 Sep 2012

Publication series

Name	13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012
Volume	3

Conference

Conference	13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012
Country/Territory	United States
City	Portland, OR
Period	9/09/12 → 13/09/12

Keywords

Nonnegative matrix factorization
Single channel speech separation
Voiced/unvoiced sound classification
Wiener filter

Cite this

@inproceedings{288061035a934e6f932955646754cc6d,

title = "A two-step NMF based algorithm for single channel speech separation",

abstract = "Nonnegative Matrix Factorization (NMF) has become an increasingly popular method in the field of non-stationary speech denoising. However most NMF-based algorithms assume prior knowledge about the background noise, which is often not available in time-varying and mobile environments. In this paper, we propose a two-step NMF based speech-noise separation algorithm to address this issue. This algorithm takes the outcome of the first NMF separation as the dataset to train the basis vectors for the background noise, which will be used for the second-step NMF separation with fixed speech and noise basis vectors. Experimental results show that the proposed algorithm could achieve better results than other NMF algorithms for speech-noise separation.",

keywords = "Nonnegative matrix factorization, Single channel speech separation, Voiced/unvoiced sound classification, Wiener filter",

author = "Shuo Wang and Wenjun Wu",

year = "2012",

language = "英语",

isbn = "9781622767595",

series = "13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012",

pages = "1987--1990",

booktitle = "13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012",

note = "13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012 ; Conference date: 09-09-2012 Through 13-09-2012",

}

Wang, S & Wu, W 2012, A two-step NMF based algorithm for single channel speech separation. in 13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012. 13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012, vol. 3, pp. 1987-1990, 13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012, Portland, OR, United States, 9/09/12.

A two-step NMF based algorithm for single channel speech separation. / Wang, Shuo; Wu, Wenjun.
13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012. 2012. p. 1987-1990 (13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012; Vol. 3).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - A two-step NMF based algorithm for single channel speech separation

AU - Wang, Shuo

AU - Wu, Wenjun

PY - 2012

Y1 - 2012

N2 - Nonnegative Matrix Factorization (NMF) has become an increasingly popular method in the field of non-stationary speech denoising. However most NMF-based algorithms assume prior knowledge about the background noise, which is often not available in time-varying and mobile environments. In this paper, we propose a two-step NMF based speech-noise separation algorithm to address this issue. This algorithm takes the outcome of the first NMF separation as the dataset to train the basis vectors for the background noise, which will be used for the second-step NMF separation with fixed speech and noise basis vectors. Experimental results show that the proposed algorithm could achieve better results than other NMF algorithms for speech-noise separation.

AB - Nonnegative Matrix Factorization (NMF) has become an increasingly popular method in the field of non-stationary speech denoising. However most NMF-based algorithms assume prior knowledge about the background noise, which is often not available in time-varying and mobile environments. In this paper, we propose a two-step NMF based speech-noise separation algorithm to address this issue. This algorithm takes the outcome of the first NMF separation as the dataset to train the basis vectors for the background noise, which will be used for the second-step NMF separation with fixed speech and noise basis vectors. Experimental results show that the proposed algorithm could achieve better results than other NMF algorithms for speech-noise separation.

KW - Nonnegative matrix factorization

KW - Single channel speech separation

KW - Voiced/unvoiced sound classification

KW - Wiener filter

UR - https://www.scopus.com/pages/publications/84878592865

M3 - 会议稿件

AN - SCOPUS:84878592865

SN - 9781622767595

T3 - 13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012

SP - 1987

EP - 1990

BT - 13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012

T2 - 13th Annual Conference of the International Speech Communication Association 2012, INTERSPEECH 2012

Y2 - 9 September 2012 through 13 September 2012

ER -

A two-step NMF based algorithm for single channel speech separation

Abstract

Publication series

Conference

Keywords

Other files and links

Fingerprint

Cite this