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

Bayesian integration in force estimation.

Konrad P Körding1, Shih-pi Ku, Daniel M Wolpert

  • 1Institute of Neurology, Sobell Department of Movement Neuroscience, University College London, London WC1N 3BG, UK. konrad@koerding.de.

Journal of Neurophysiology
|June 11, 2004
PubMed
Summary

This study shows that the brain uses Bayesian integration to combine sensory input with prior experience for optimal force estimation. Humans can learn and adapt these priors, suggesting Bayesian models guide motor control.

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

The Simons Collaboration on Ecological Neuroscience: Studying how the brain interacts with the world.

Neuron·2026
Same author

Physical contact reveals a hidden layer of cortical architecture.

bioRxiv : the preprint server for biology·2026
Same author

The organization of multiple motor memories.

Current opinion in neurobiology·2026
Same author

Modeling attention and binding in the brain through bidirectional recurrent gating.

Nature communications·2026
Same author

Mid-superior temporal sulcus encodes spatial context and behavioral state in freely moving macaques.

bioRxiv : the preprint server for biology·2026
Same author

Adaptive integration of model-based and model-free strategies in human reinforcement learning of reachable space.

bioRxiv : the preprint server for biology·2026

Area of Science:

  • Neuroscience
  • Motor Control
  • Computational Neuroscience

Background:

  • Accurate force estimation is crucial for interacting with objects.
  • Sensory information alone is insufficient due to environmental uncertainties.
  • Prior experience (knowledge of previously encountered objects) aids in force estimation.

Purpose of the Study:

  • To investigate if the central nervous system (CNS) employs Bayesian integration for force estimation.
  • To determine if humans combine sensory data with prior information to estimate forces.
  • To explore the adaptability of prior distributions in force estimation tasks.

Main Methods:

  • Developed a novel sensorimotor estimation task.
  • Controlled the distribution of forces to establish a prior experience.

Related Experiment Videos

  • Analyzed subject responses to assess integration of sensory and prior information.

    • Main Results:

    • Subjects successfully integrated sensory information with their prior experience to estimate forces.
    • Participants demonstrated the ability to learn and adapt to different prior force distributions.
    • Behavioral data supports the hypothesis of Bayesian integration in force estimation.

    Conclusions:

    • The CNS appears to utilize Bayesian models for estimating required forces.
    • This suggests a sophisticated mechanism for combining real-time sensory data with learned priors.
    • Findings provide insights into the neural basis of sensorimotor control and adaptation.