Selecting disorder-specific features for speech pathology fingerprinting

Visar Berisha, Steven Sandoval, Rene Utianski, Julie Liss, Andreas Spanias

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

6 Citations (Scopus)

Abstract

The general aim of this work is to learn a unique statistical signature for the state of a particular speech pathology. We pose this as a speaker identification problem for dysarthric individuals. To that end, we propose a novel algorithm for feature selection that aims to minimize the effects of speaker-specific features (e.g., fundamental frequency) and maximize the effects of pathology-specific features (e.g., vocal tract distortions and speech rhythm). We derive a cost function for optimizing feature selection that simultaneously trades off between these two competing criteria. Furthermore, we develop an efficient algorithm that optimizes this cost function and test the algorithm on a set of 34 dysarthric and 13 healthy speakers. Results show that the proposed method yields a set of features related to the speech disorder and not an individual's speaking style. When compared to other feature-selection algorithms, the proposed approach results in an improvement in a disorder fingerprinting task by selecting features that are specific to the disorder.

Original languageEnglish (US)
Title of host publication2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings
Pages7562-7566
Number of pages5
DOIs
StatePublished - Oct 18 2013
Externally publishedYes
Event2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Vancouver, BC, Canada
Duration: May 26 2013May 31 2013

Publication series

NameICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
ISSN (Print)1520-6149

Conference

Conference2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013
CountryCanada
CityVancouver, BC
Period5/26/135/31/13

Fingerprint

Pathology
Feature extraction
Cost functions

Keywords

  • dysarthria
  • feature selection
  • machine learning
  • speech pathology

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Berisha, V., Sandoval, S., Utianski, R., Liss, J., & Spanias, A. (2013). Selecting disorder-specific features for speech pathology fingerprinting. In 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings (pp. 7562-7566). [6639133] (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). https://doi.org/10.1109/ICASSP.2013.6639133

Selecting disorder-specific features for speech pathology fingerprinting. / Berisha, Visar; Sandoval, Steven; Utianski, Rene; Liss, Julie; Spanias, Andreas.

2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings. 2013. p. 7562-7566 6639133 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).

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

Berisha, V, Sandoval, S, Utianski, R, Liss, J & Spanias, A 2013, Selecting disorder-specific features for speech pathology fingerprinting. in 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings., 6639133, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, pp. 7562-7566, 2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013, Vancouver, BC, Canada, 5/26/13. https://doi.org/10.1109/ICASSP.2013.6639133
Berisha V, Sandoval S, Utianski R, Liss J, Spanias A. Selecting disorder-specific features for speech pathology fingerprinting. In 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings. 2013. p. 7562-7566. 6639133. (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). https://doi.org/10.1109/ICASSP.2013.6639133
Berisha, Visar ; Sandoval, Steven ; Utianski, Rene ; Liss, Julie ; Spanias, Andreas. / Selecting disorder-specific features for speech pathology fingerprinting. 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings. 2013. pp. 7562-7566 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).
@inproceedings{5d43a32c89074a8aa55bdcc5f34a3445,
title = "Selecting disorder-specific features for speech pathology fingerprinting",
abstract = "The general aim of this work is to learn a unique statistical signature for the state of a particular speech pathology. We pose this as a speaker identification problem for dysarthric individuals. To that end, we propose a novel algorithm for feature selection that aims to minimize the effects of speaker-specific features (e.g., fundamental frequency) and maximize the effects of pathology-specific features (e.g., vocal tract distortions and speech rhythm). We derive a cost function for optimizing feature selection that simultaneously trades off between these two competing criteria. Furthermore, we develop an efficient algorithm that optimizes this cost function and test the algorithm on a set of 34 dysarthric and 13 healthy speakers. Results show that the proposed method yields a set of features related to the speech disorder and not an individual's speaking style. When compared to other feature-selection algorithms, the proposed approach results in an improvement in a disorder fingerprinting task by selecting features that are specific to the disorder.",
keywords = "dysarthria, feature selection, machine learning, speech pathology",
author = "Visar Berisha and Steven Sandoval and Rene Utianski and Julie Liss and Andreas Spanias",
year = "2013",
month = "10",
day = "18",
doi = "10.1109/ICASSP.2013.6639133",
language = "English (US)",
isbn = "9781479903566",
series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",
pages = "7562--7566",
booktitle = "2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings",

}

TY - GEN

T1 - Selecting disorder-specific features for speech pathology fingerprinting

AU - Berisha, Visar

AU - Sandoval, Steven

AU - Utianski, Rene

AU - Liss, Julie

AU - Spanias, Andreas

PY - 2013/10/18

Y1 - 2013/10/18

N2 - The general aim of this work is to learn a unique statistical signature for the state of a particular speech pathology. We pose this as a speaker identification problem for dysarthric individuals. To that end, we propose a novel algorithm for feature selection that aims to minimize the effects of speaker-specific features (e.g., fundamental frequency) and maximize the effects of pathology-specific features (e.g., vocal tract distortions and speech rhythm). We derive a cost function for optimizing feature selection that simultaneously trades off between these two competing criteria. Furthermore, we develop an efficient algorithm that optimizes this cost function and test the algorithm on a set of 34 dysarthric and 13 healthy speakers. Results show that the proposed method yields a set of features related to the speech disorder and not an individual's speaking style. When compared to other feature-selection algorithms, the proposed approach results in an improvement in a disorder fingerprinting task by selecting features that are specific to the disorder.

AB - The general aim of this work is to learn a unique statistical signature for the state of a particular speech pathology. We pose this as a speaker identification problem for dysarthric individuals. To that end, we propose a novel algorithm for feature selection that aims to minimize the effects of speaker-specific features (e.g., fundamental frequency) and maximize the effects of pathology-specific features (e.g., vocal tract distortions and speech rhythm). We derive a cost function for optimizing feature selection that simultaneously trades off between these two competing criteria. Furthermore, we develop an efficient algorithm that optimizes this cost function and test the algorithm on a set of 34 dysarthric and 13 healthy speakers. Results show that the proposed method yields a set of features related to the speech disorder and not an individual's speaking style. When compared to other feature-selection algorithms, the proposed approach results in an improvement in a disorder fingerprinting task by selecting features that are specific to the disorder.

KW - dysarthria

KW - feature selection

KW - machine learning

KW - speech pathology

UR - http://www.scopus.com/inward/record.url?scp=84890535325&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890535325&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2013.6639133

DO - 10.1109/ICASSP.2013.6639133

M3 - Conference contribution

AN - SCOPUS:84890535325

SN - 9781479903566

T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

SP - 7562

EP - 7566

BT - 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings

ER -