Vowel recognition from continuous articulatory movements for speaker-dependent applications

Jun Wang, Jordan R. Green, Ashok K Samal, Thomas D Carrell

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

4 Citations (Scopus)

Abstract

A novel approach was developed to recognize vowels from continuous tongue and lip movements. Vowels were classified based on movement patterns (rather than on derived articulatory features, e.g., lip opening) using a machine learning approach. Recognition accuracy on a single-speaker dataset was 94.02% with a very short latency. Recognition accuracy was better for high vowels than for low vowels. This finding parallels previous empirical findings on tongue movements during vowels. The recognition algorithm was then used to drive an articulation-to-acoustics synthesizer. The synthesizer recognizes vowels from continuous input stream of tongue and lip movements and plays the corresponding sound samples in near real-time.

Original languageEnglish (US)
Title of host publication4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings
DOIs
StatePublished - Dec 1 2010
Event4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Gold Coast, QLD, Australia
Duration: Dec 13 2010Dec 15 2010

Publication series

Name4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings

Conference

Conference4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010
CountryAustralia
CityGold Coast, QLD
Period12/13/1012/15/10

Fingerprint

Learning systems
Acoustics
Acoustic waves

Keywords

  • Articulation
  • Machine learning
  • Recognition
  • Support vector machine

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Signal Processing

Cite this

Wang, J., Green, J. R., Samal, A. K., & Carrell, T. D. (2010). Vowel recognition from continuous articulatory movements for speaker-dependent applications. In 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings [5709716] (4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings). https://doi.org/10.1109/ICSPCS.2010.5709716

Vowel recognition from continuous articulatory movements for speaker-dependent applications. / Wang, Jun; Green, Jordan R.; Samal, Ashok K; Carrell, Thomas D.

4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings. 2010. 5709716 (4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings).

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

Wang, J, Green, JR, Samal, AK & Carrell, TD 2010, Vowel recognition from continuous articulatory movements for speaker-dependent applications. in 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings., 5709716, 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings, 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010, Gold Coast, QLD, Australia, 12/13/10. https://doi.org/10.1109/ICSPCS.2010.5709716
Wang J, Green JR, Samal AK, Carrell TD. Vowel recognition from continuous articulatory movements for speaker-dependent applications. In 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings. 2010. 5709716. (4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings). https://doi.org/10.1109/ICSPCS.2010.5709716
Wang, Jun ; Green, Jordan R. ; Samal, Ashok K ; Carrell, Thomas D. / Vowel recognition from continuous articulatory movements for speaker-dependent applications. 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings. 2010. (4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings).
@inproceedings{d2f6db3525684796bec4545ac67eb345,
title = "Vowel recognition from continuous articulatory movements for speaker-dependent applications",
abstract = "A novel approach was developed to recognize vowels from continuous tongue and lip movements. Vowels were classified based on movement patterns (rather than on derived articulatory features, e.g., lip opening) using a machine learning approach. Recognition accuracy on a single-speaker dataset was 94.02{\%} with a very short latency. Recognition accuracy was better for high vowels than for low vowels. This finding parallels previous empirical findings on tongue movements during vowels. The recognition algorithm was then used to drive an articulation-to-acoustics synthesizer. The synthesizer recognizes vowels from continuous input stream of tongue and lip movements and plays the corresponding sound samples in near real-time.",
keywords = "Articulation, Machine learning, Recognition, Support vector machine",
author = "Jun Wang and Green, {Jordan R.} and Samal, {Ashok K} and Carrell, {Thomas D}",
year = "2010",
month = "12",
day = "1",
doi = "10.1109/ICSPCS.2010.5709716",
language = "English (US)",
isbn = "9781424479078",
series = "4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings",
booktitle = "4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings",

}

TY - GEN

T1 - Vowel recognition from continuous articulatory movements for speaker-dependent applications

AU - Wang, Jun

AU - Green, Jordan R.

AU - Samal, Ashok K

AU - Carrell, Thomas D

PY - 2010/12/1

Y1 - 2010/12/1

N2 - A novel approach was developed to recognize vowels from continuous tongue and lip movements. Vowels were classified based on movement patterns (rather than on derived articulatory features, e.g., lip opening) using a machine learning approach. Recognition accuracy on a single-speaker dataset was 94.02% with a very short latency. Recognition accuracy was better for high vowels than for low vowels. This finding parallels previous empirical findings on tongue movements during vowels. The recognition algorithm was then used to drive an articulation-to-acoustics synthesizer. The synthesizer recognizes vowels from continuous input stream of tongue and lip movements and plays the corresponding sound samples in near real-time.

AB - A novel approach was developed to recognize vowels from continuous tongue and lip movements. Vowels were classified based on movement patterns (rather than on derived articulatory features, e.g., lip opening) using a machine learning approach. Recognition accuracy on a single-speaker dataset was 94.02% with a very short latency. Recognition accuracy was better for high vowels than for low vowels. This finding parallels previous empirical findings on tongue movements during vowels. The recognition algorithm was then used to drive an articulation-to-acoustics synthesizer. The synthesizer recognizes vowels from continuous input stream of tongue and lip movements and plays the corresponding sound samples in near real-time.

KW - Articulation

KW - Machine learning

KW - Recognition

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=79952491602&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79952491602&partnerID=8YFLogxK

U2 - 10.1109/ICSPCS.2010.5709716

DO - 10.1109/ICSPCS.2010.5709716

M3 - Conference contribution

SN - 9781424479078

T3 - 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings

BT - 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings

ER -