Automatic speech and singing classification in ambulatory recordings for normal and disordered voices

Andrew J. Ortiz; Laura E. Toles; Katherine L. Marks; Silvia Capobianco; Daryush D. Mehta; Robert E. Hillman; Jarrad H. Van Stan

doi:10.1121/1.5115804

Automatic speech and singing classification in ambulatory recordings for normal and disordered voices

Andrew J. Ortiz, Laura E. Toles, Katherine L. Marks, Silvia Capobianco, Daryush D. Mehta, Robert E. Hillman, Jarrad H. Van Stan

Research output: Contribution to journal › Article › peer-review

14 Scopus citations

Abstract

Ambulatory voice monitoring is a promising tool for investigating phonotraumatic vocal hyperfunction (PVH), associated with the development of vocal fold lesions. Since many patients with PVH are professional vocalists, a classifier was developed to better understand phonatory mechanisms during speech and singing. Twenty singers with PVH and 20 matched healthy controls were monitored with a neck-surface accelerometer-based ambulatory voice monitor. An expert-labeled ground truth data set was used to train a logistic regression on 15 subject-pairs with fundamental frequency and autocorrelation peak amplitude as input features. Overall classification accuracy of 94.2% was achieved on the held-out test set.

Original language	English (US)
Pages (from-to)	EL22-EL27
Journal	Journal of the Acoustical Society of America
Volume	146
Issue number	1
DOIs	https://doi.org/10.1121/1.5115804
State	Published - Jul 1 2019
Externally published	Yes

ASJC Scopus subject areas

Arts and Humanities (miscellaneous)
Acoustics and Ultrasonics

Access to Document

10.1121/1.5115804

Cite this

@article{2202341e46c84a39a622c5a57c22547a,

title = "Automatic speech and singing classification in ambulatory recordings for normal and disordered voices",

abstract = "Ambulatory voice monitoring is a promising tool for investigating phonotraumatic vocal hyperfunction (PVH), associated with the development of vocal fold lesions. Since many patients with PVH are professional vocalists, a classifier was developed to better understand phonatory mechanisms during speech and singing. Twenty singers with PVH and 20 matched healthy controls were monitored with a neck-surface accelerometer-based ambulatory voice monitor. An expert-labeled ground truth data set was used to train a logistic regression on 15 subject-pairs with fundamental frequency and autocorrelation peak amplitude as input features. Overall classification accuracy of 94.2% was achieved on the held-out test set.",

author = "Ortiz, {Andrew J.} and Toles, {Laura E.} and Marks, {Katherine L.} and Silvia Capobianco and Mehta, {Daryush D.} and Hillman, {Robert E.} and {Van Stan}, {Jarrad H.}",

note = "Publisher Copyright: {\textcopyright} 2019 Acoustical Society of America.",

year = "2019",

month = jul,

day = "1",

doi = "10.1121/1.5115804",

language = "English (US)",

volume = "146",

pages = "EL22--EL27",

journal = "Journal of the Acoustical Society of America",

issn = "0001-4966",

publisher = "Acoustical Society of America",

number = "1",

}

TY - JOUR

T1 - Automatic speech and singing classification in ambulatory recordings for normal and disordered voices

AU - Ortiz, Andrew J.

AU - Toles, Laura E.

AU - Marks, Katherine L.

AU - Capobianco, Silvia

AU - Mehta, Daryush D.

AU - Hillman, Robert E.

AU - Van Stan, Jarrad H.

PY - 2019/7/1

Y1 - 2019/7/1

N2 - Ambulatory voice monitoring is a promising tool for investigating phonotraumatic vocal hyperfunction (PVH), associated with the development of vocal fold lesions. Since many patients with PVH are professional vocalists, a classifier was developed to better understand phonatory mechanisms during speech and singing. Twenty singers with PVH and 20 matched healthy controls were monitored with a neck-surface accelerometer-based ambulatory voice monitor. An expert-labeled ground truth data set was used to train a logistic regression on 15 subject-pairs with fundamental frequency and autocorrelation peak amplitude as input features. Overall classification accuracy of 94.2% was achieved on the held-out test set.

AB - Ambulatory voice monitoring is a promising tool for investigating phonotraumatic vocal hyperfunction (PVH), associated with the development of vocal fold lesions. Since many patients with PVH are professional vocalists, a classifier was developed to better understand phonatory mechanisms during speech and singing. Twenty singers with PVH and 20 matched healthy controls were monitored with a neck-surface accelerometer-based ambulatory voice monitor. An expert-labeled ground truth data set was used to train a logistic regression on 15 subject-pairs with fundamental frequency and autocorrelation peak amplitude as input features. Overall classification accuracy of 94.2% was achieved on the held-out test set.

UR - http://www.scopus.com/inward/record.url?scp=85069055888&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85069055888&partnerID=8YFLogxK

U2 - 10.1121/1.5115804

DO - 10.1121/1.5115804

M3 - Article

C2 - 31370647

AN - SCOPUS:85069055888

SN - 0001-4966

VL - 146

SP - EL22-EL27

JO - Journal of the Acoustical Society of America

JF - Journal of the Acoustical Society of America

IS - 1

ER -

Automatic speech and singing classification in ambulatory recordings for normal and disordered voices

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this