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

Learning the hidden structure of speech.

J L Elman1, D Zipser

  • 1Department of Linguistics, University of California, San Diego, La Jolla 92093.

The Journal of the Acoustical Society of America
|April 1, 1988
PubMed
Summary
This summary is machine-generated.

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

A model of visual-spatial memory across saccades.

Vision research·2001
Same author

Neural network models of bilateral coordination.

Biological cybernetics·1999
Same author

A model of reaching dynamics in primary motor cortex.

Journal of cognitive neuroscience·1998
Same author

Analysis of direction selectivity arising from recurrent cortical interactions.

Neural computation·1998
Same author

A model that accounts for activity in primate frontal cortex during a delayed matching-to-sample task.

The Journal of neuroscience : the official journal of the Society for Neuroscience·1998
Same author

Connectionism and developmental psychology.

Journal of child psychology and psychiatry, and allied disciplines·1997
Same journal

High-resolution depth estimation for multiple wideband sources in deep sea via sparse Bayesian learninga).

The Journal of the Acoustical Society of America·2026
Same journal

Depression markers in speech: An approach based on tract variables dynamics.

The Journal of the Acoustical Society of America·2026
Same journal

The oyster toadfish (Opsanus tau) alters active and diurnal calling amid vessel noise in New York City.

The Journal of the Acoustical Society of America·2026
Same journal

Experimental noise characterisation of phase-locked tandem-rotor in edgewise flight.

The Journal of the Acoustical Society of America·2026
Same journal

The tune-text-temporal synergy: Prosodic effects of final segmental weakening in Neapolitan.

The Journal of the Acoustical Society of America·2026
Same journal

Monitoring vessel movement above critical offshore infrastructure using distributed acoustic sensing.

The Journal of the Acoustical Society of America·2026
See all related articles

Backpropagation neural networks can learn to analyze and recognize speech, achieving up to 95% accuracy in labeling sounds. These networks develop internal representations for speech analysis and pattern recognition without needing external labels.

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Speech Processing

Background:

  • Backpropagation neural networks are a powerful machine learning technique for learning functional relationships from input/output pattern pairs.
  • Speech analysis and recognition are complex tasks that traditionally require extensive feature engineering and labeled data.

Purpose of the Study:

  • To investigate the application of backpropagation neural networks for speech analysis and recognition.
  • To assess the networks' ability to learn from labeled and unlabeled speech data.
  • To explore the feature representations developed by these networks.

Main Methods:

  • Computer simulations were conducted using backpropagation neural networks.
  • Networks were trained on presegmented speech tokens for labeling tasks.

Related Experiment Videos

  • Networks were trained on segmented and unsegmented continuous speech, with and without external labels.

    • Main Results:

    • Networks achieved up to 95% accuracy in labeling presegmented speech tokens.
    • Networks successfully recognized and delineated sounds in continuous speech without external labels.
    • Developed internal representations included traditional phonetic features and novel sound distinctions.

    Conclusions:

    • Backpropagation learning is effective for analyzing and recognizing complex, natural speech data.
    • The networks can identify feature structures crucial for speech analysis and pattern recognition.
    • This approach shows promise for unsupervised speech segmentation and label learning.