Learning Obstacle Representations for Neural Motion Planning

Robin STRUDEL, Ricardo Garcia Pinel, Justin Carpentier, Jean-Paul Laumond, Ivan Laptev, Cordelia Schmid
Proceedings of the 2020 Conference on Robot Learning, PMLR 155:355-364, 2021.

Abstract

Motion planning and obstacle avoidance is a key challenge in robotics applications. While previous work succeeds to provide excellent solutions for known environments, sensor-based motion planning in new and dynamic environments remains to be difficult. In this work we address sensor-based motion planning from a learning perspective. Motivated by recent advances in visual recognition, we argue the importance of learning appropriate representations for motion planning. We propose a new obstacle representation based on the PointNet architecture and learn it jointly with policies for obstacle avoidance. We experimentally evaluate our approach for rigid body motion planning in challenging environments and demonstrate significant improvements of the state of the art in terms of accuracy and efficiency.

Cite this Paper


BibTeX
@InProceedings{pmlr-v155-strudel21a, title = {Learning Obstacle Representations for Neural Motion Planning}, author = {STRUDEL, Robin and Pinel, Ricardo Garcia and Carpentier, Justin and Laumond, Jean-Paul and Laptev, Ivan and Schmid, Cordelia}, booktitle = {Proceedings of the 2020 Conference on Robot Learning}, pages = {355--364}, year = {2021}, editor = {Kober, Jens and Ramos, Fabio and Tomlin, Claire}, volume = {155}, series = {Proceedings of Machine Learning Research}, month = {16--18 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v155/strudel21a/strudel21a.pdf}, url = {https://proceedings.mlr.press/v155/strudel21a.html}, abstract = {Motion planning and obstacle avoidance is a key challenge in robotics applications. While previous work succeeds to provide excellent solutions for known environments, sensor-based motion planning in new and dynamic environments remains to be difficult. In this work we address sensor-based motion planning from a learning perspective. Motivated by recent advances in visual recognition, we argue the importance of learning appropriate representations for motion planning. We propose a new obstacle representation based on the PointNet architecture and learn it jointly with policies for obstacle avoidance. We experimentally evaluate our approach for rigid body motion planning in challenging environments and demonstrate significant improvements of the state of the art in terms of accuracy and efficiency.} }
Endnote
%0 Conference Paper %T Learning Obstacle Representations for Neural Motion Planning %A Robin STRUDEL %A Ricardo Garcia Pinel %A Justin Carpentier %A Jean-Paul Laumond %A Ivan Laptev %A Cordelia Schmid %B Proceedings of the 2020 Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2021 %E Jens Kober %E Fabio Ramos %E Claire Tomlin %F pmlr-v155-strudel21a %I PMLR %P 355--364 %U https://proceedings.mlr.press/v155/strudel21a.html %V 155 %X Motion planning and obstacle avoidance is a key challenge in robotics applications. While previous work succeeds to provide excellent solutions for known environments, sensor-based motion planning in new and dynamic environments remains to be difficult. In this work we address sensor-based motion planning from a learning perspective. Motivated by recent advances in visual recognition, we argue the importance of learning appropriate representations for motion planning. We propose a new obstacle representation based on the PointNet architecture and learn it jointly with policies for obstacle avoidance. We experimentally evaluate our approach for rigid body motion planning in challenging environments and demonstrate significant improvements of the state of the art in terms of accuracy and efficiency.
APA
STRUDEL, R., Pinel, R.G., Carpentier, J., Laumond, J., Laptev, I. & Schmid, C.. (2021). Learning Obstacle Representations for Neural Motion Planning. Proceedings of the 2020 Conference on Robot Learning, in Proceedings of Machine Learning Research 155:355-364 Available from https://proceedings.mlr.press/v155/strudel21a.html.

Related Material