Learning Obstacle Representations for Neural Motion Planning - Archive ouverte HAL Access content directly
Conference Papers Year : 2021

Learning Obstacle Representations for Neural Motion Planning

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 [1] 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.
Fichier principal
Vignette du fichier
2008.11174.pdf (1.26 Mo) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-02944348 , version 1 (21-09-2020)

Identifiers

Cite

Robin Strudel, Ricardo Garcia, Justin Carpentier, Jean-Paul Laumond, Ivan Laptev, et al.. Learning Obstacle Representations for Neural Motion Planning. CoRL 2020 - Conference on Robot Learning, Nov 2020, Cambridge MA / Virtual, United States. pp.355-364. ⟨hal-02944348⟩
190 View
177 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More