Music dereverberation using harmonic structure source model and wiener filter

Naoki Yasuraoka, Takuya Yoshioka, Tomohiro Nakatani, Atsushi Nakamura, Hiroshi G. Okuno

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)

Abstract

This paper proposes a dereverberation method for musical audio signals. Existing dereverberation methods are designed for speech signals and are not necessarily effective for suppressing long and dense reverberation in musical audio signals because: 1) an all-pole model and a non-parametric model, which are used to represent source spectra, do not match musical tones, and 2) the conventional inverse-filter-based dereverberation is not effective for suppressing long and dense reverberation. To overcome the two problems, an appropriate dereverberation approach for musical audio signals is established. The first problem is resolved by using a harmonic Gaussian mixture model (GMM) to accurately model the harmonic structure of a source spectrum. The second problem is resolved by performing dereverberation with a Wiener filter based on both an estimated inverse filter and an estimated source spectrum model. Experimental results reveal the effectiveness of the proposed dereverberation method using these two solutions.

Original languageEnglish
Title of host publicationICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Pages53-56
Number of pages4
DOIs
Publication statusPublished - 2010
Externally publishedYes
Event2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010 - Dallas, TX
Duration: 2010 Mar 142010 Mar 19

Other

Other2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010
CityDallas, TX
Period10/3/1410/3/19

Fingerprint

Reverberation
Poles

Keywords

  • Dereverberation
  • Harmonic GMM
  • Music signal processing
  • Wiener filter

ASJC Scopus subject areas

  • Signal Processing
  • Software
  • Electrical and Electronic Engineering

Cite this

Yasuraoka, N., Yoshioka, T., Nakatani, T., Nakamura, A., & Okuno, H. G. (2010). Music dereverberation using harmonic structure source model and wiener filter. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings (pp. 53-56). [5496223] https://doi.org/10.1109/ICASSP.2010.5496223

Music dereverberation using harmonic structure source model and wiener filter. / Yasuraoka, Naoki; Yoshioka, Takuya; Nakatani, Tomohiro; Nakamura, Atsushi; Okuno, Hiroshi G.

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2010. p. 53-56 5496223.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Yasuraoka, N, Yoshioka, T, Nakatani, T, Nakamura, A & Okuno, HG 2010, Music dereverberation using harmonic structure source model and wiener filter. in ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings., 5496223, pp. 53-56, 2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010, Dallas, TX, 10/3/14. https://doi.org/10.1109/ICASSP.2010.5496223
Yasuraoka N, Yoshioka T, Nakatani T, Nakamura A, Okuno HG. Music dereverberation using harmonic structure source model and wiener filter. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2010. p. 53-56. 5496223 https://doi.org/10.1109/ICASSP.2010.5496223
Yasuraoka, Naoki ; Yoshioka, Takuya ; Nakatani, Tomohiro ; Nakamura, Atsushi ; Okuno, Hiroshi G. / Music dereverberation using harmonic structure source model and wiener filter. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2010. pp. 53-56
@inproceedings{a8f2b23cf3714dfda09395a848197f8d,
title = "Music dereverberation using harmonic structure source model and wiener filter",
abstract = "This paper proposes a dereverberation method for musical audio signals. Existing dereverberation methods are designed for speech signals and are not necessarily effective for suppressing long and dense reverberation in musical audio signals because: 1) an all-pole model and a non-parametric model, which are used to represent source spectra, do not match musical tones, and 2) the conventional inverse-filter-based dereverberation is not effective for suppressing long and dense reverberation. To overcome the two problems, an appropriate dereverberation approach for musical audio signals is established. The first problem is resolved by using a harmonic Gaussian mixture model (GMM) to accurately model the harmonic structure of a source spectrum. The second problem is resolved by performing dereverberation with a Wiener filter based on both an estimated inverse filter and an estimated source spectrum model. Experimental results reveal the effectiveness of the proposed dereverberation method using these two solutions.",
keywords = "Dereverberation, Harmonic GMM, Music signal processing, Wiener filter",
author = "Naoki Yasuraoka and Takuya Yoshioka and Tomohiro Nakatani and Atsushi Nakamura and Okuno, {Hiroshi G.}",
year = "2010",
doi = "10.1109/ICASSP.2010.5496223",
language = "English",
isbn = "9781424442966",
pages = "53--56",
booktitle = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

}

TY - GEN

T1 - Music dereverberation using harmonic structure source model and wiener filter

AU - Yasuraoka, Naoki

AU - Yoshioka, Takuya

AU - Nakatani, Tomohiro

AU - Nakamura, Atsushi

AU - Okuno, Hiroshi G.

PY - 2010

Y1 - 2010

N2 - This paper proposes a dereverberation method for musical audio signals. Existing dereverberation methods are designed for speech signals and are not necessarily effective for suppressing long and dense reverberation in musical audio signals because: 1) an all-pole model and a non-parametric model, which are used to represent source spectra, do not match musical tones, and 2) the conventional inverse-filter-based dereverberation is not effective for suppressing long and dense reverberation. To overcome the two problems, an appropriate dereverberation approach for musical audio signals is established. The first problem is resolved by using a harmonic Gaussian mixture model (GMM) to accurately model the harmonic structure of a source spectrum. The second problem is resolved by performing dereverberation with a Wiener filter based on both an estimated inverse filter and an estimated source spectrum model. Experimental results reveal the effectiveness of the proposed dereverberation method using these two solutions.

AB - This paper proposes a dereverberation method for musical audio signals. Existing dereverberation methods are designed for speech signals and are not necessarily effective for suppressing long and dense reverberation in musical audio signals because: 1) an all-pole model and a non-parametric model, which are used to represent source spectra, do not match musical tones, and 2) the conventional inverse-filter-based dereverberation is not effective for suppressing long and dense reverberation. To overcome the two problems, an appropriate dereverberation approach for musical audio signals is established. The first problem is resolved by using a harmonic Gaussian mixture model (GMM) to accurately model the harmonic structure of a source spectrum. The second problem is resolved by performing dereverberation with a Wiener filter based on both an estimated inverse filter and an estimated source spectrum model. Experimental results reveal the effectiveness of the proposed dereverberation method using these two solutions.

KW - Dereverberation

KW - Harmonic GMM

KW - Music signal processing

KW - Wiener filter

UR - http://www.scopus.com/inward/record.url?scp=78049401020&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78049401020&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2010.5496223

DO - 10.1109/ICASSP.2010.5496223

M3 - Conference contribution

SN - 9781424442966

SP - 53

EP - 56

BT - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

ER -