Loading...
Loading...

Go to the content (press return)

Dynamically consistent probabilistic model for robot motion learning

Author
Pardo, D.E.; Rozo, L.; Alenyà, G.; Torras, C.
Type of activity
Presentation of work at congresses
Name of edition
IROS'2012 Workshop on Learning and Interaction in Haptic Robots
Date of publication
2012
Presentation's date
2012
Book of congress proceedings
Proceedings of the IROS'2012 Workshop on Learning and Interaction in Haptic Robots
First page
1
Last page
2
Repository
http://hdl.handle.net/2117/17450 Open in new window
URL
http://www.iit.it/images/stories/advanced-robotics/Workshops/Iros_2012/Pardo-IROS2012WS.pdf Open in new window
Abstract
This work presents a probabilistic model for learning robot tasks from human demonstrations using kinesthetic teaching. The difference with respect to previous works is that a complete state of the robot is used to obtain a consistent representation of the dynamics of the task. The learning framework is based on hidden Markov models and Gaussian mixture regression, used for coding and reproducing the skills. Benefits of the proposed approach are shown in the execution of a simple self-crossing t...
Citation
Pardo, D.E. [et al.]. Dynamically consistent probabilistic model for robot motion learning. A: Workshop on Learning and Interaction in Haptic Robots. "Proceedings of the IROS'2012 Workshop on Learning and Interaction in Haptic Robots". Algarve: 2012, p. 1-2.
Keywords
Gaussian mixture regression (GMR), hidden Markov models (HMM), kinesthetic teaching, learning (artificial intelligence) manipulators robot dynamics PARAULES AUTOR: learning from demonstration
Group of research
GREC - Knowledge Engineering Research Group
ROBiri - IRI Robotics Group

Participants

  • Pardo Ayala, Diego Esteban  (author and speaker )
  • Rozo Castañeda, Leonel  (author and speaker )
  • Alenyà Ribas, Guillem  (author and speaker )
  • Torras, Carme  (author and speaker )