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

Accurate visible speech synthesis based on concatenating variable length motion capture data.

Jiyong Ma1, Ron Cole, Bryan Pellom

  • 1Center for Spoken Language Research, University of Colorado at Boulder, CO 80309-0594, USA. jiyong@cslr.colorado.edu

IEEE Transactions on Visualization and Computer Graphics
|March 3, 2006
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

Reducing Nuisance Telemetry Alarms in the Adult Inpatient Setting.

Journal of doctoral nursing practice·2025
Same author

Effects of liver X receptor in O3-induced airway inflammation and remodeling in mice.

Journal of thoracic disease·2024
Same author

Identification of an Exosome-relevant SNHG6-hsa-miR-429- CHRDL1/CCNA2 Axis for Lung Adenocarcinoma Prognosis Evaluation.

Current medicinal chemistry·2024
Same author

Implementation of an Extubation Readiness Guideline for Preterm Infants.

Advances in neonatal care : official journal of the National Association of Neonatal Nurses·2024
Same author

ITGA9-AS1 up-regulates ITGA9 by targeting miR-4765 and recruiting HNRNPU to affect the proliferation and apoptosis of non-small cell lung cancer cells.

Cellular and molecular biology (Noisy-le-Grand, France)·2024
Same author

A benchmarking framework for the accurate and cost-effective detection of clinically-relevant structural variants for cancer target identification and diagnosis.

Journal of translational medicine·2024

This study introduces a new method for creating realistic visible speech animations by optimizing speech units from motion capture data. The approach accurately maps facial movements for improved speech synthesis and reading education tools.

Area of Science:

  • Computer Vision
  • Speech Synthesis
  • Machine Learning

Background:

  • Visible speech synthesis is crucial for applications like education and accessibility.
  • Existing methods often struggle with naturalness and accurate facial motion mapping.
  • Modeling coarticulation effects in visible speech is a significant challenge.

Purpose of the Study:

  • To develop a novel approach for accurate visible speech synthesis.
  • To create a machine learning technique for mapping facial motions between different faces.
  • To implement an end-to-end visible speech animation system for educational purposes.

Main Methods:

  • Utilizing a large corpus of motion capture data for variable-length speech units.
  • Employing a machine learning model to map facial motions from source to target faces.

Related Experiment Videos

  • Developing search and adaptation algorithms for optimal lip motion synthesis.
  • Creating a complete visible speech animation system.

    • Main Results:

    • The developed system successfully synthesizes accurate visible speech.
    • The approach effectively models long-distance coarticulation effects.
    • Subjective and objective evaluations confirm the system's accuracy and power.
    • The system is deployed in classrooms for reading education.

    Conclusions:

    • The proposed method offers an accurate and powerful solution for visible speech synthesis.
    • The developed system has practical applications in educational technology.
    • The approach demonstrates the potential of motion capture data for realistic speech animation.