Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers

Narjes Bozorg; Michael T. Johnson

doi:10.1109/ISSPIT.2018.8705093

Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers

Narjes Bozorg, Michael T. Johnson

Electrical and Computer Engineering

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

3 Scopus citations

Abstract

This paper compares the performance of acoustic-to-articulatory inversion for both L1 and L2 speakers of English, as a function of the number of Gaussian Mixtures used in the inversion model. The inversion system is based on an HMM-GMM approach and is implemented on the Marquette Electromagnetic Articulography corpus of Mandarin Accented English (EMAMAE) including 20 native English speakers and 19 native Mandarin speakers of English. Results indicate that for Mandarin speakers 12 Gaussian mixtures and for L1 American English speakers 11 Gaussian mixtures give the lowest Root-Mean-Squared error (RMSE) and highest correlation between the estimated and actual articulatory pattern.

Original language	English
Title of host publication	2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018
ISBN (Electronic)	9781538675687
DOIs	https://doi.org/10.1109/ISSPIT.2018.8705093
State	Published - Jul 2 2018
Event	2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018 - Louisville, United States Duration: Dec 6 2018 → Dec 8 2018

Publication series

Name	2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018
Volume	2019-January

Conference

Conference	2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018
Country/Territory	United States
City	Louisville
Period	12/6/18 → 12/8/18

Bibliographical note

Publisher Copyright:
© 2018 IEEE.

Keywords

Acoustic-to-Articulatory Inversion
Articulatory Features
Electro-Magnetic Articulography
Gaussian Mixture Model
Hidden Markov Model

ASJC Scopus subject areas

Signal Processing
Computer Science Applications
Computer Networks and Communications

Access to Document

10.1109/ISSPIT.2018.8705093

Cite this

Bozorg, N., & Johnson, M. T. (2018). Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers. In 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018 Article 8705093 (2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018; Vol. 2019-January). https://doi.org/10.1109/ISSPIT.2018.8705093

@inproceedings{d74520f88cea4e1da3b2ba074ec6d52e,

title = "Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers",

abstract = "This paper compares the performance of acoustic-to-articulatory inversion for both L1 and L2 speakers of English, as a function of the number of Gaussian Mixtures used in the inversion model. The inversion system is based on an HMM-GMM approach and is implemented on the Marquette Electromagnetic Articulography corpus of Mandarin Accented English (EMAMAE) including 20 native English speakers and 19 native Mandarin speakers of English. Results indicate that for Mandarin speakers 12 Gaussian mixtures and for L1 American English speakers 11 Gaussian mixtures give the lowest Root-Mean-Squared error (RMSE) and highest correlation between the estimated and actual articulatory pattern.",

keywords = "Acoustic-to-Articulatory Inversion, Articulatory Features, Electro-Magnetic Articulography, Gaussian Mixture Model, Hidden Markov Model",

author = "Narjes Bozorg and Johnson, {Michael T.}",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018 ; Conference date: 06-12-2018 Through 08-12-2018",

year = "2018",

month = jul,

day = "2",

doi = "10.1109/ISSPIT.2018.8705093",

language = "English",

series = "2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018",

booktitle = "2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018",

}

Bozorg, N & Johnson, MT 2018, Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers. in 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018., 8705093, 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018, vol. 2019-January, 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018, Louisville, United States, 12/6/18. https://doi.org/10.1109/ISSPIT.2018.8705093

Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers. / Bozorg, Narjes; Johnson, Michael T.
2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018. 2018. 8705093 (2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018; Vol. 2019-January).

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

TY - GEN

T1 - Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers

AU - Bozorg, Narjes

AU - Johnson, Michael T.

PY - 2018/7/2

Y1 - 2018/7/2

N2 - This paper compares the performance of acoustic-to-articulatory inversion for both L1 and L2 speakers of English, as a function of the number of Gaussian Mixtures used in the inversion model. The inversion system is based on an HMM-GMM approach and is implemented on the Marquette Electromagnetic Articulography corpus of Mandarin Accented English (EMAMAE) including 20 native English speakers and 19 native Mandarin speakers of English. Results indicate that for Mandarin speakers 12 Gaussian mixtures and for L1 American English speakers 11 Gaussian mixtures give the lowest Root-Mean-Squared error (RMSE) and highest correlation between the estimated and actual articulatory pattern.

AB - This paper compares the performance of acoustic-to-articulatory inversion for both L1 and L2 speakers of English, as a function of the number of Gaussian Mixtures used in the inversion model. The inversion system is based on an HMM-GMM approach and is implemented on the Marquette Electromagnetic Articulography corpus of Mandarin Accented English (EMAMAE) including 20 native English speakers and 19 native Mandarin speakers of English. Results indicate that for Mandarin speakers 12 Gaussian mixtures and for L1 American English speakers 11 Gaussian mixtures give the lowest Root-Mean-Squared error (RMSE) and highest correlation between the estimated and actual articulatory pattern.

KW - Acoustic-to-Articulatory Inversion

KW - Articulatory Features

KW - Electro-Magnetic Articulography

KW - Gaussian Mixture Model

KW - Hidden Markov Model

UR - http://www.scopus.com/inward/record.url?scp=85065668587&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85065668587&partnerID=8YFLogxK

U2 - 10.1109/ISSPIT.2018.8705093

DO - 10.1109/ISSPIT.2018.8705093

M3 - Conference contribution

AN - SCOPUS:85065668587

T3 - 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018

BT - 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018

T2 - 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018

Y2 - 6 December 2018 through 8 December 2018

ER -

Bozorg N, Johnson MT. Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers. In 2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018. 2018. 8705093. (2018 IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2018). doi: 10.1109/ISSPIT.2018.8705093

Comparing performance of acoustic-to-articulatory inversion for mandarin accented english and american english speakers

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this