Dynamic Cloth Manipulation with Deep Reinforcement Learning
Authors:
- 1 May 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 4630-4636
Abstract
In this paper we present a Deep Reinforcement Learning approach to solve dynamic cloth manipulation tasks. Differing from the case of rigid objects, we stress that the followed trajectory (including speed and acceleration) has a decisive influence on the final state of cloth, which can greatly vary even if the positions reached by the grasped points are the same. We explore how goal positions for non-grasped points can be attained through learning adequate trajectories for the grasped points. Our approach uses few demonstrations to improve control policy learning, and a sparse reward approach to avoid engineering complex reward functions. Since perception of textiles is challenging, we also study different state representations to assess the minimum observation space required for learning to succeed. Finally, we compare different combinations of control policy encodings, demonstrations, and sparse reward learning techniques, and show that our proposed approach can learn dynamic cloth manipulation in an efficient way, i.e., using a reduced observation space, a few demonstrations, and a sparse reward.Keywords
All Related Versions
This publication has 27 references indexed in Scilit:
- Learning from multiple demonstrations using trajectory-aware non-rigid registration with applications to deformable object manipulation2015
- A friction-model-based framework for Reinforcement Learning of robotic tasks in non-rigid environments2015
- Learning force-based manipulation of deformable objects from multiple demonstrations2015
- FINDDD: A fast 3D descriptor to characterize textiles for robot manipulation2013
- Survey on model-based manipulation planning of deformable objectsRobotics and Computer-Integrated Manufacturing, 2012
- SOFA: A Multi-Model Framework for Interactive Physical Simulation2012
- A geometric approach to robotic laundry foldingThe International Journal of Robotics Research, 2011
- Motion planning for dynamic folding of a cloth with two high-speed robot hands and two high-speed sliders2011
- Bringing clothing into desired configurations with limited perception2011
- Reinforcement learning of motor skills with policy gradientsNeural Networks, 2008