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

Computational approaches to motor learning by imitation.

Stefan Schaal1, Auke Ijspeert, Aude Billard

  • 1Computer Science & Neuroscience, University of Southern California, 3641 Watt Way, Los Angeles 90089-2520, USA. sschaal@usc.edu

Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences
|April 12, 2003
PubMed
Summary

This study explores computational approaches to movement imitation, focusing on the motor control aspects. It reviews statistical and mathematical methods for tackling challenges in replicating observed actions.

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

Leaping out of the water: Aerial-aquatic locomotion with flapping wings.

Science (New York, N.Y.)·2026
Same author

Swimming with robots: investigating fish locomotion, sensing, and schooling behavior with robotic swimmers.

Nature communications·2026
Same author

Centralized brain networks controlling antennal grooming coordination.

Nature communications·2026
Same author

Demonstrate once, execute on many: Kinematic intelligence for cross-robot skill transfer.

Science robotics·2026
Same author

A detachable crawling robotic hand.

Nature communications·2026
Same author

Four-Handed Solo Laparoscopic Anastomosis with the Support of Supernumerary Robotic Arms.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025

Area of Science:

  • Robotics
  • Computational Neuroscience
  • Artificial Intelligence

Background:

  • Movement imitation involves complex perception-action loops, requiring mechanisms for movement recognition, pose estimation, and coordinate transformation.
  • Existing research often focuses on action recognition, leaving the motor control side of imitation less explored.
  • A complete imitation system necessitates understanding the integration of perceptual and motor processes.

Purpose of the Study:

  • To computationally analyze the challenges in movement imitation.
  • To review and discuss statistical and mathematical approaches for the motor control aspects of imitation.
  • To propose a taxonomy for imitation learning based on control policies.

Main Methods:

  • Review of statistical and mathematical methods for movement imitation.

Related Experiment Videos

  • Focus on the motor control side, assuming pre-processed perceptual information.
  • Formalization of motor control using control policies and performance criteria.

    • Main Results:

    • Identified key computational problems in movement imitation, including pose estimation, body correspondence, and coordinate transformation.
    • Discussed merits, disadvantages, and underlying principles of various computational approaches.
    • Proposed a framework for classifying imitation learning strategies.

    Conclusions:

    • Movement imitation is a complex problem requiring integrated perception-action systems.
    • Computational and mathematical approaches offer viable strategies for addressing specific imitation challenges.
    • Further research into control policies and performance criteria can advance imitation learning.