Learning Obstacle Representations for Neural Motion Planning
Authors:
- Robin StrudelRSRobin Strudel
- Ricardo GarciaRGRicardo Garcia
- Justin CarpentierJCJustin Carpentier
- Ivan LaptevILIvan Laptev
- Cordelia SchmidCSCordelia Schmid
Preprint
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 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 train 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.Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Harnessing Reinforcement Learning for Neural Motion Planning2019
- Robot Motion Planning in Learned Latent SpacesIEEE Robotics and Automation Letters, 2019
- The YCB object and Model set: Towards common benchmarks for manipulation research2015
- Kineo CAM: a success story of motion planning algorithmsIEEE Robotics & Automation Magazine, 2006
- Computational approaches to motor learning by imitationPhilosophical Transactions of the Royal Society B: Biological Sciences, 2003
- An efficient neural network approach to dynamic robot motion planningNeural Networks, 2000
- The dynamic window approach to collision avoidanceIEEE Robotics & Automation Magazine, 1997
- Probabilistic roadmaps for path planning in high-dimensional configuration spacesIEEE Transactions on Robotics and Automation, 1996
- Neural Network Dynamics for Path Planning and Obstacle AvoidanceNeural Networks, 1995