Monaural separation of dependent audio sources based on a generalized Wiener filter

Guilin Ma; Finn T. Agerkvist; J.B. Luther

doi:10.1109/ISSPIT.2007.4458074

Monaural separation of dependent audio sources based on a generalized Wiener filter

Guilin Ma, Finn T. Agerkvist, J.B. Luther

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

481 Downloads (Pure)

Abstract

This paper presents a two-stage approach for single- channel separation of dependent audio sources. The proposed algorithm is developed in the Bayesian framework and designed for general audio signals. In the first stage of the algorithm, the joint distribution of discrete Fourier transform (DFT) coefficients of the dependent sources is modeled by complex Gaussian mixture models in the frequency domain from samples of individual sources to capture the properties of the sources and their correlation. During the second stage, the mixture is separated through a generalized Wiener filter, which takes correlation term and local stationarity into account. The performance of the algorithm is tested on real audio signals. The results show that the proposed algorithm works very well when the dependent sources have comparable variances and linear correlation.

Original language	English
Title of host publication	Proceedings of the 7th IEEE International Symposium on Signal Processing and Information Technology
Publisher	IEEE
Publication date	2007
ISBN (Print)	978-1-4244-1835-0
DOIs	https://doi.org/10.1109/ISSPIT.2007.4458074
Publication status	Published - 2007
Event	2007 IEEE International Symposium on Signal Processing and Information Technology - Cairo, Egypt Duration: 15 Dec 2007 → 18 Dec 2007 Conference number: 7 https://ieeexplore.ieee.org/xpl/conhome/4455349/proceeding

Conference

Conference	2007 IEEE International Symposium on Signal Processing and Information Technology
Number	7
Country/Territory	Egypt
City	Cairo
Period	15/12/2007 → 18/12/2007
Internet address	https://ieeexplore.ieee.org/xpl/conhome/4455349/proceeding

Bibliographical note

Copyright: 2007 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE

Access to Document

10.1109/ISSPIT.2007.4458074

MaFinal published version, 1.58 MB

OpenUrl availability

Full text

Cite this

@inproceedings{04a2a85bb26345fead0330b9be41708f,

title = "Monaural separation of dependent audio sources based on a generalized Wiener filter",

abstract = "This paper presents a two-stage approach for single- channel separation of dependent audio sources. The proposed algorithm is developed in the Bayesian framework and designed for general audio signals. In the first stage of the algorithm, the joint distribution of discrete Fourier transform (DFT) coefficients of the dependent sources is modeled by complex Gaussian mixture models in the frequency domain from samples of individual sources to capture the properties of the sources and their correlation. During the second stage, the mixture is separated through a generalized Wiener filter, which takes correlation term and local stationarity into account. The performance of the algorithm is tested on real audio signals. The results show that the proposed algorithm works very well when the dependent sources have comparable variances and linear correlation.",

author = "Guilin Ma and Agerkvist, {Finn T.} and J.B. Luther",

note = "Copyright: 2007 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE; 2007 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2007 ; Conference date: 15-12-2007 Through 18-12-2007",

year = "2007",

doi = "10.1109/ISSPIT.2007.4458074",

language = "English",

isbn = "978-1-4244-1835-0",

booktitle = "Proceedings of the 7th IEEE International Symposium on Signal Processing and Information Technology",

publisher = "IEEE",

address = "United States",

url = "https://ieeexplore.ieee.org/xpl/conhome/4455349/proceeding",

}

Ma, G, Agerkvist, FT & Luther, JB 2007, Monaural separation of dependent audio sources based on a generalized Wiener filter. in Proceedings of the 7th IEEE International Symposium on Signal Processing and Information Technology. IEEE, 2007 IEEE International Symposium on Signal Processing and Information Technology, Cairo, Egypt, 15/12/2007. https://doi.org/10.1109/ISSPIT.2007.4458074

TY - GEN

T1 - Monaural separation of dependent audio sources based on a generalized Wiener filter

AU - Ma, Guilin

AU - Agerkvist, Finn T.

AU - Luther, J.B.

N1 - Conference code: 7

PY - 2007

Y1 - 2007

N2 - This paper presents a two-stage approach for single- channel separation of dependent audio sources. The proposed algorithm is developed in the Bayesian framework and designed for general audio signals. In the first stage of the algorithm, the joint distribution of discrete Fourier transform (DFT) coefficients of the dependent sources is modeled by complex Gaussian mixture models in the frequency domain from samples of individual sources to capture the properties of the sources and their correlation. During the second stage, the mixture is separated through a generalized Wiener filter, which takes correlation term and local stationarity into account. The performance of the algorithm is tested on real audio signals. The results show that the proposed algorithm works very well when the dependent sources have comparable variances and linear correlation.

AB - This paper presents a two-stage approach for single- channel separation of dependent audio sources. The proposed algorithm is developed in the Bayesian framework and designed for general audio signals. In the first stage of the algorithm, the joint distribution of discrete Fourier transform (DFT) coefficients of the dependent sources is modeled by complex Gaussian mixture models in the frequency domain from samples of individual sources to capture the properties of the sources and their correlation. During the second stage, the mixture is separated through a generalized Wiener filter, which takes correlation term and local stationarity into account. The performance of the algorithm is tested on real audio signals. The results show that the proposed algorithm works very well when the dependent sources have comparable variances and linear correlation.

U2 - 10.1109/ISSPIT.2007.4458074

DO - 10.1109/ISSPIT.2007.4458074

M3 - Article in proceedings

SN - 978-1-4244-1835-0

BT - Proceedings of the 7th IEEE International Symposium on Signal Processing and Information Technology

PB - IEEE

T2 - 2007 IEEE International Symposium on Signal Processing and Information Technology

Y2 - 15 December 2007 through 18 December 2007

ER -

Monaural separation of dependent audio sources based on a generalized Wiener filter

Abstract

Conference

Bibliographical note

Access to Document

OpenUrl availability

Fingerprint

Cite this