Robots Learning (Through) Interactions

Project 4

The acquisition and self-improvement of novel motor skills is among the most important problems in robotics.

I will discuss various learning techniques we developed that enable robots to have complex interactions with their environment and humans. Complexity arises from dealing with highdimensional input data, non-linear dynamics in general and contacts in particular, multiple reference frames, and variability in objects, environments, and tasks. A human teacher is always involved in the learning process, either directly (providing data) or indirectly (designing the optimization criterion), which raises the question: How to best make use of the interactions with the human teacher to render the learning process efficient and effective?

TU Delft:
  • Jens Kober

Archived event

This event has already happened.

Documents & links

  • Robots Learning (Through) Interactions
    Jens Kober
    Summary

    The acquisition and self-improvement of novel motor skills is among the most important problems in robotics.

    Jens discussed various learning techniques his group developed that enable robots to have complex interactions with their environment and humans. Complexity arises from dealing with high-dimensional input data, non-linear dynamics in general and contacts in particular, multiple reference frames, and variability in objects, environments, and tasks. A human teacher is always involved in the learning process, either directly (providing data) or indirectly (designing the optimization criterion), which raises the question: How to best make use of the interactions with the human teacher to render the learning process efficient and effective?

    Download