Towards Robust Skill Generalization: Unifying Learning from Demonstration and Motion Planning

Muhammad Asif Rana, Mustafa Mukadam, Seyed Reza Ahmadzadeh, Sonia Chernova, Byron Boots
; Proceedings of the 1st Annual Conference on Robot Learning, PMLR 78:109-118, 2017.

Abstract

In this paper, we present Combined Learning from demonstration And Motion Planning (CLAMP) as an efficient approach to skill learning and generalizable skill reproduction. CLAMP combines the strengths of Learning from Demonstration (LfD) and motion planning into a unifying framework. We carry out probabilistic inference to find trajectories which are optimal with respect to a given skill and also feasible in different scenarios. We use factor graph optimization to speed up inference. To encode optimality, we provide a new probabilistic skill model based on a stochastic dynamical system. This skill model requires minimal parameter tuning to learn, is suitable to encode skill constraints, and allows efficient inference. Preliminary experimental results showing skill generalization over initial robot state and unforeseen obstacles are presented.

Cite this Paper


BibTeX
@InProceedings{pmlr-v78-rana17a, title = {Towards Robust Skill Generalization: Unifying Learning from Demonstration and Motion Planning}, author = {Muhammad Asif Rana and Mustafa Mukadam and Seyed Reza Ahmadzadeh and Sonia Chernova and Byron Boots}, booktitle = {Proceedings of the 1st Annual Conference on Robot Learning}, pages = {109--118}, year = {2017}, editor = {Sergey Levine and Vincent Vanhoucke and Ken Goldberg}, volume = {78}, series = {Proceedings of Machine Learning Research}, month = {13--15 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v78/rana17a/rana17a.pdf}, url = {http://proceedings.mlr.press/v78/rana17a.html}, abstract = {In this paper, we present Combined Learning from demonstration And Motion Planning (CLAMP) as an efficient approach to skill learning and generalizable skill reproduction. CLAMP combines the strengths of Learning from Demonstration (LfD) and motion planning into a unifying framework. We carry out probabilistic inference to find trajectories which are optimal with respect to a given skill and also feasible in different scenarios. We use factor graph optimization to speed up inference. To encode optimality, we provide a new probabilistic skill model based on a stochastic dynamical system. This skill model requires minimal parameter tuning to learn, is suitable to encode skill constraints, and allows efficient inference. Preliminary experimental results showing skill generalization over initial robot state and unforeseen obstacles are presented.} }
Endnote
%0 Conference Paper %T Towards Robust Skill Generalization: Unifying Learning from Demonstration and Motion Planning %A Muhammad Asif Rana %A Mustafa Mukadam %A Seyed Reza Ahmadzadeh %A Sonia Chernova %A Byron Boots %B Proceedings of the 1st Annual Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2017 %E Sergey Levine %E Vincent Vanhoucke %E Ken Goldberg %F pmlr-v78-rana17a %I PMLR %J Proceedings of Machine Learning Research %P 109--118 %U http://proceedings.mlr.press %V 78 %W PMLR %X In this paper, we present Combined Learning from demonstration And Motion Planning (CLAMP) as an efficient approach to skill learning and generalizable skill reproduction. CLAMP combines the strengths of Learning from Demonstration (LfD) and motion planning into a unifying framework. We carry out probabilistic inference to find trajectories which are optimal with respect to a given skill and also feasible in different scenarios. We use factor graph optimization to speed up inference. To encode optimality, we provide a new probabilistic skill model based on a stochastic dynamical system. This skill model requires minimal parameter tuning to learn, is suitable to encode skill constraints, and allows efficient inference. Preliminary experimental results showing skill generalization over initial robot state and unforeseen obstacles are presented.
APA
Rana, M.A., Mukadam, M., Ahmadzadeh, S.R., Chernova, S. & Boots, B.. (2017). Towards Robust Skill Generalization: Unifying Learning from Demonstration and Motion Planning. Proceedings of the 1st Annual Conference on Robot Learning, in PMLR 78:109-118

Related Material