Jove
Visualize
Contact Us

Related Experiment Videos

Vowel-onset detection.

D J Hermes1

  • 1Institute for Perception Research/IPO, Eindhoven, The Netherlands.

The Journal of the Acoustical Society of America
|February 1, 1990
PubMed
Summary

This study presents an algorithm for accurately detecting vowel onsets in fluent speech using spectral analysis. Modifications significantly reduced false alarms in isolated words, achieving high temporal accuracy.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Auditory and visual similarity of pitch contours.

Journal of speech, language, and hearing research : JSLHR·1998
Same author

Measuring the perceptual similarity of pitch contours.

Journal of speech, language, and hearing research : JSLHR·1998
Same author

Timing of pitch movements and accentuation of syllables in Dutch.

The Journal of the Acoustical Society of America·1998
Same author

Prominence lent by rising and falling pitch movements: Testing two models.

The Journal of the Acoustical Society of America·1996
Same author

Vowel-onset detection by vowel-strength measurement, cochlear-nucleus simulation, and multilayer perceptrons.

The Journal of the Acoustical Society of America·1996
Same author

Teaching intonation to young deaf children with the intonation meter.

Folia phoniatrica et logopaedica : official organ of the International Association of Logopedics and Phoniatrics (IALP)·1996
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Area of Science:

  • Speech processing
  • Phonetics
  • Signal analysis

Background:

  • Vowel onsets are crucial for speech intelligibility.
  • Accurate detection of vowel onsets in fluent speech remains a challenge.
  • Existing methods may struggle with contextual variations and false alarms.

Purpose of the Study:

  • To develop and validate an algorithm for detecting vowel onsets in fluent speech.
  • To improve the accuracy and reduce false alarms compared to previous methods.
  • To assess the temporal precision of the developed algorithm.

Main Methods:

  • An algorithm based on detecting rapidly increasing resonance peaks in the amplitude spectrum was developed.
  • The algorithm was applied to isolated words, and parameters were modified to reduce false alarms.
  • Temporal accuracy was evaluated against phonetician-determined perceptual moments of vowel onset.

Main Results:

  • The algorithm successfully detects the majority of vowel onsets in fluent speech.
  • Parameter modifications significantly reduced false alarms in isolated word contexts.
  • The algorithm demonstrated temporal accuracy better than 20 milliseconds.

Conclusions:

  • The presented algorithm offers a robust method for vowel onset detection in speech.
  • The refined algorithm balances detection accuracy with a low false alarm rate.
  • This tool has potential applications in speech analysis and recognition systems.

Related Experiment Videos