Probabilistic inference methods in robotics-filling the gap between high-level reasoning and low-level motion control

Posted in Science on September 04, 2008


Probabilistic inference methods in robotics-filling the gap between high-level reasoning and low-level motion control

Lecture slides:

  • Probabilistic inference methods in robotics
  • Robotics
  • Outline
  • Background: Probabilistic Inference & Planning
  • Planning as Likelihood Maximization
  • Extensions to POMDPs
  • Outline
  • Bayesian view on classical control
  • Bayesian view
  • Same also works for dynamic control...
  • Bayesian view
  • Motion control
  • Example 1
  • Example 2
  • Summary
  • ML approaches for all subproblems?
  • Thanks

Author: Marc Toussaint, Tu Berlin

Watch Video

Tags: Science, Lectures, Computer Science, Machine Learning, VideoLectures.Net, Bayesian Learning, Robotics, Statistical Learning