Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization

Daniele Giacobello, Mads Græsbøll Christensen, Manohar Murthi, Søren Holdt Jensen, Marc Moonen

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

13 Citations (Scopus)

Abstract

Linear prediction of speech based on 1-norm minimization has already proved to be an interesting alternative to 2-norm minimization. In particular, choosing the 1-norm as a convex relaxation of the 0-norm, the corresponding linear prediction model offers a sparser residual better suited for coding applications. In this paper, we propose a new speech modeling technique based on reweighted 1-norm minimization. The purpose of the reweighted scheme is to overcome the mismatch between 0-norm minimization and 1-norm minimization while keeping the problem solvable with convex estimation tools. Experimental results prove the effectiveness of the reweighted 1-norm minimization, offering better coding properties compared to 1-norm minimization.

Original languageEnglish
Title of host publicationICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Pages4650-4653
Number of pages4
DOIs
StatePublished - Nov 8 2010
Event2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010 - Dallas, TX, United States
Duration: Mar 14 2010Mar 19 2010

Other

Other2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010
CountryUnited States
CityDallas, TX
Period3/14/103/19/10

Keywords

  • 1-norm minimization
  • Linear prediction
  • Speech analysis
  • Speech coding

ASJC Scopus subject areas

  • Signal Processing
  • Software
  • Electrical and Electronic Engineering

Cite this

Giacobello, D., Christensen, M. G., Murthi, M., Jensen, S. H., & Moonen, M. (2010). Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings (pp. 4650-4653). [5495198] https://doi.org/10.1109/ICASSP.2010.5495198

Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization. / Giacobello, Daniele; Christensen, Mads Græsbøll; Murthi, Manohar; Jensen, Søren Holdt; Moonen, Marc.

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2010. p. 4650-4653 5495198.

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

Giacobello, D, Christensen, MG, Murthi, M, Jensen, SH & Moonen, M 2010, Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization. in ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings., 5495198, pp. 4650-4653, 2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010, Dallas, TX, United States, 3/14/10. https://doi.org/10.1109/ICASSP.2010.5495198
Giacobello D, Christensen MG, Murthi M, Jensen SH, Moonen M. Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2010. p. 4650-4653. 5495198 https://doi.org/10.1109/ICASSP.2010.5495198
Giacobello, Daniele ; Christensen, Mads Græsbøll ; Murthi, Manohar ; Jensen, Søren Holdt ; Moonen, Marc. / Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2010. pp. 4650-4653
@inproceedings{4548d258363d4e0b88972c36ada2e544,
title = "Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization",
abstract = "Linear prediction of speech based on 1-norm minimization has already proved to be an interesting alternative to 2-norm minimization. In particular, choosing the 1-norm as a convex relaxation of the 0-norm, the corresponding linear prediction model offers a sparser residual better suited for coding applications. In this paper, we propose a new speech modeling technique based on reweighted 1-norm minimization. The purpose of the reweighted scheme is to overcome the mismatch between 0-norm minimization and 1-norm minimization while keeping the problem solvable with convex estimation tools. Experimental results prove the effectiveness of the reweighted 1-norm minimization, offering better coding properties compared to 1-norm minimization.",
keywords = "1-norm minimization, Linear prediction, Speech analysis, Speech coding",
author = "Daniele Giacobello and Christensen, {Mads Gr{\ae}sb{\o}ll} and Manohar Murthi and Jensen, {S{\o}ren Holdt} and Marc Moonen",
year = "2010",
month = "11",
day = "8",
doi = "10.1109/ICASSP.2010.5495198",
language = "English",
isbn = "9781424442966",
pages = "4650--4653",
booktitle = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

}

TY - GEN

T1 - Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization

AU - Giacobello, Daniele

AU - Christensen, Mads Græsbøll

AU - Murthi, Manohar

AU - Jensen, Søren Holdt

AU - Moonen, Marc

PY - 2010/11/8

Y1 - 2010/11/8

N2 - Linear prediction of speech based on 1-norm minimization has already proved to be an interesting alternative to 2-norm minimization. In particular, choosing the 1-norm as a convex relaxation of the 0-norm, the corresponding linear prediction model offers a sparser residual better suited for coding applications. In this paper, we propose a new speech modeling technique based on reweighted 1-norm minimization. The purpose of the reweighted scheme is to overcome the mismatch between 0-norm minimization and 1-norm minimization while keeping the problem solvable with convex estimation tools. Experimental results prove the effectiveness of the reweighted 1-norm minimization, offering better coding properties compared to 1-norm minimization.

AB - Linear prediction of speech based on 1-norm minimization has already proved to be an interesting alternative to 2-norm minimization. In particular, choosing the 1-norm as a convex relaxation of the 0-norm, the corresponding linear prediction model offers a sparser residual better suited for coding applications. In this paper, we propose a new speech modeling technique based on reweighted 1-norm minimization. The purpose of the reweighted scheme is to overcome the mismatch between 0-norm minimization and 1-norm minimization while keeping the problem solvable with convex estimation tools. Experimental results prove the effectiveness of the reweighted 1-norm minimization, offering better coding properties compared to 1-norm minimization.

KW - 1-norm minimization

KW - Linear prediction

KW - Speech analysis

KW - Speech coding

UR - http://www.scopus.com/inward/record.url?scp=78049385869&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78049385869&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2010.5495198

DO - 10.1109/ICASSP.2010.5495198

M3 - Conference contribution

SN - 9781424442966

SP - 4650

EP - 4653

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

ER -