Learning to touch objects through stage-wise deep reinforcement learning - Université Clermont Auvergne Access content directly
Conference Papers Year : 2018

Learning to touch objects through stage-wise deep reinforcement learning

Céline Teulière
Thierry Chateau

Abstract

Learning complex behaviors through reinforcement learning is particularly challenging when reward is only available upon successful completion of the full behavior. In manipulation robotics, so-called shaping rewards are often used to overcome this problem. However, these usually require human engineering or (partial) world models describing, e.g., the kinematics of the robot or high-level modules for perception. Here we propose an alternative method to learn an object palm-touching task through a weakly-supervised and stage-wise learning of simpler tasks. First, the robot learns to fixate the object with its cameras. Second, the robot learns eye-hand coordination by learning to fixate its end effector. Third, using the previously acquired skills an informative shaping reward can be computed which facilitates efficient learning of the object palm-touching task. We demonstrate in simulation that learning the full task with this shaping reward is comparable to learning with an informative supervised reward.

Domains

Automatic
Fichier principal
Vignette du fichier
fdlb_paper_iros_v1.pdf (1.63 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01915423 , version 1 (15-01-2019)

Identifiers

  • HAL Id : hal-01915423 , version 1

Cite

François de La Bourdonnaye, Céline Teulière, Jochen Triesch, Thierry Chateau. Learning to touch objects through stage-wise deep reinforcement learning. IEEE/RSJ International Conference on Intelligent Robots and Systems, Oct 2018, Madrid, Spain. ⟨hal-01915423⟩
65 View
178 Download

Share

Gmail Facebook X LinkedIn More