Jove
Visualize
Contact Us
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

Related Experiment Videos

Modeling global and focal hyperarticulation during human-computer error resolution

S Oviatt1, G A Levow, E Moreton

  • 1Department of Computer Science, Oregon Graduate Institute of Science and Technology, Portland 97291, USA. oviatt@cse.ogi.edu

The Journal of the Acoustical Society of America
|November 20, 1998
PubMed
Summary

When resolving speech recognition errors, people speak more deliberately, increasing pauses and articulation clarity. This hyperarticulation aids error correction in human-computer interactions.

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

Impact of early exposure to a cafeteria diet on prefrontal cortex monoamines and novel object recognition in adolescent rats.

Behavioural brain research·2019
Same author

Linguistic adaptations during spoken and multimodal error resolution.

Language and speech·2000
Same author

Relationship of mineral metabolism and long-term calcium and cholecalciferol supplementation to blood pressure in normotensive men.

The American journal of clinical nutrition·1990
Same author

[Cholesterol and its beta and alpha heparinoid fractions in normal and atheromatous subjects].

Revue de l'atherosclerose et des arteriopathies peripheriques·1969
Same author

[Metabolic and environmental factors in the etiopathogenesis of myocardial infarct in young persons].

Revista argentina de endocrinologia y metabolismo·1968
Same author

[Total lipids. Comparative values of the Kunkel reaction and the gravimetric method].

Revue de l'atherosclerose et des arteriopathies peripheriques·1967

Area of Science:

  • Speech Communication
  • Human-Computer Interaction
  • Linguistics

Background:

  • Users often hyperarticulate when correcting speech recognition errors.
  • Hyperarticulation, a clarified speech style, has paradoxically been linked to increased recognition errors.
  • Understanding speech adaptations during error resolution is crucial for improving interactive systems.

Purpose of the Study:

  • To analyze acoustic, prosodic, and phonological speech adaptations during human-computer error resolution.
  • To investigate speech changes during both global and focal utterance repairs.
  • To validate and generalize the computer-elicited hyperarticulate adaptation model (CHAM).

Main Methods:

  • Utilized a semi-automatic simulation with novel error-generation for speech analysis.

Related Experiment Videos

  • Compared speech samples immediately before and after system recognition errors.
  • Analyzed matched original-repeat utterance pairs for linguistic adaptations during repairs.

    • Main Results:

    • Hyperarticulate speech adaptations were primarily durational, marked by increased pauses.
    • Speech became more deliberate with enhanced articulatory clarity.
    • Focal error repairs used durational, pitch, and amplitude cues for prominence marking.

    Conclusions:

    • Speech adaptations during error resolution are consistent with the CHAM model.
    • Findings suggest strategies for enhancing error handling in spoken language systems.
    • Improved system design can leverage user speech adaptations for better interaction outcomes.