Online unsupervised terrain classification for a compliant tensegrity robot using a mixture of echo state networks
- Author
- Jeroen Burms (UGent) , Ken Caluwaerts (UGent) and Joni Dambre (UGent)
- Organization
- Abstract
- Truly autonomous robots require the capacity to recognise their surroundings by interpreting their sensorimotor stream. We present an online learning algorithm for training a mixture of echo state network experts that can segment a compliant robot's sensorimotor stream. Our method follows a probabilistic approach, using a hidden Markov model to model the switching dynamics between the experts. The algorithm's performance is evaluated on an unsupervised terrain classification problem using a compliant, underactuated, six-strut tensegrity robot. The results show that our model captures the influence of terrain-robot interactions on the robot's complex dynamics and correctly segments the sensorimotor stream. We demonstrate that the activity pattern of the experts can be used to train a highly compliant robot to distinguish between different environments using only noisy internal sensors
- Keywords
- NEURONS, LOCOMOTION, MODELS, TIME
Downloads
-
(...).pdf
- full text
- |
- UGent only
- |
- |
- 1.65 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-6960123
- MLA
- Burms, Jeroen, et al. “Online Unsupervised Terrain Classification for a Compliant Tensegrity Robot Using a Mixture of Echo State Networks.” IEEE International Conference on Robotics and Automation ICRA, IEEE, 2015, pp. 4252–57, doi:10.1109/ICRA.2015.7139785.
- APA
- Burms, J., Caluwaerts, K., & Dambre, J. (2015). Online unsupervised terrain classification for a compliant tensegrity robot using a mixture of echo state networks. IEEE International Conference on Robotics and Automation ICRA, 4252–4257. https://doi.org/10.1109/ICRA.2015.7139785
- Chicago author-date
- Burms, Jeroen, Ken Caluwaerts, and Joni Dambre. 2015. “Online Unsupervised Terrain Classification for a Compliant Tensegrity Robot Using a Mixture of Echo State Networks.” In IEEE International Conference on Robotics and Automation ICRA, 4252–57. IEEE. https://doi.org/10.1109/ICRA.2015.7139785.
- Chicago author-date (all authors)
- Burms, Jeroen, Ken Caluwaerts, and Joni Dambre. 2015. “Online Unsupervised Terrain Classification for a Compliant Tensegrity Robot Using a Mixture of Echo State Networks.” In IEEE International Conference on Robotics and Automation ICRA, 4252–4257. IEEE. doi:10.1109/ICRA.2015.7139785.
- Vancouver
- 1.Burms J, Caluwaerts K, Dambre J. Online unsupervised terrain classification for a compliant tensegrity robot using a mixture of echo state networks. In: IEEE International Conference on Robotics and Automation ICRA. IEEE; 2015. p. 4252–7.
- IEEE
- [1]J. Burms, K. Caluwaerts, and J. Dambre, “Online unsupervised terrain classification for a compliant tensegrity robot using a mixture of echo state networks,” in IEEE International Conference on Robotics and Automation ICRA, Seattle, USA, 2015, pp. 4252–4257.
@inproceedings{6960123,
abstract = {{Truly autonomous robots require the capacity to recognise their surroundings by interpreting their sensorimotor stream. We present an online learning algorithm for training a mixture of echo state network experts that can segment a compliant robot's sensorimotor stream. Our method follows a probabilistic approach, using a hidden Markov model to model the switching dynamics between the experts. The algorithm's performance is evaluated on an unsupervised terrain classification problem using a compliant, underactuated, six-strut tensegrity robot. The results show that our model captures the influence of terrain-robot interactions on the robot's complex dynamics and correctly segments the sensorimotor stream. We demonstrate that the activity pattern of the experts can be used to train a highly compliant robot to distinguish between different environments using only noisy internal sensors}},
author = {{Burms, Jeroen and Caluwaerts, Ken and Dambre, Joni}},
booktitle = {{IEEE International Conference on Robotics and Automation ICRA}},
isbn = {{9781479969234}},
issn = {{1050-4729}},
keywords = {{NEURONS,LOCOMOTION,MODELS,TIME}},
language = {{eng}},
location = {{Seattle, USA}},
pages = {{4252--4257}},
publisher = {{IEEE}},
title = {{Online unsupervised terrain classification for a compliant tensegrity robot using a mixture of echo state networks}},
url = {{http://doi.org/10.1109/ICRA.2015.7139785}},
year = {{2015}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: