Advanced search
1 file | 970.87 KB Add to list

A differentiable physics engine for deep learning in robotics

Author
Organization
Abstract
An important field in robotics is the optimization of controllers. Currently, robots are often treated as a black box in this optimization process, which is the reason why derivative-free optimization methods such as evolutionary algorithms or reinforcement learning are omnipresent. When gradient-based methods are used, models are kept small or rely on finite difference approximations for the Jacobian. This method quickly grows expensive with increasing numbers of parameters, such as found in deep learning. We propose an implementation of a modern physics engine, which can differentiate control parameters. This engine is implemented for both CPU and GPU. Firstly, this paper shows how such an engine speeds up the optimization process, even for small problems. Furthermore, it explains why this is an alternative approach to deep Q-learning, for using deep learning in robotics. Finally, we argue that this is a big step for deep learning in robotics, as it opens up new possibilities to optimize robots, both in hardware and software.
Keywords
Differential physics engine, deep learning, backpropagation, Robotics, Simulation Technology

Downloads

  • fnbot-13-00006.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 970.87 KB

Citation

Please use this url to cite or link to this publication:

MLA
Degrave, Jonas et al. “A Differentiable Physics Engine for Deep Learning in Robotics.” Ed. Florian Röhrbein. FRONTIERS IN NEUROROBOTICS 13.6 (2019): n. pag. Print.
APA
Degrave, J., Hermans, M., Dambre, J., & wyffels, F. (2019). A differentiable physics engine for deep learning in robotics. (F. Röhrbein, Ed.)FRONTIERS IN NEUROROBOTICS, 13(6).
Chicago author-date
Degrave, Jonas, Michiel Hermans, Joni Dambre, and Francis wyffels. 2019. “A Differentiable Physics Engine for Deep Learning in Robotics.” Ed. Florian Röhrbein. Frontiers in Neurorobotics 13 (6).
Chicago author-date (all authors)
Degrave, Jonas, Michiel Hermans, Joni Dambre, and Francis wyffels. 2019. “A Differentiable Physics Engine for Deep Learning in Robotics.” Ed. Florian Röhrbein. Frontiers in Neurorobotics 13 (6).
Vancouver
1.
Degrave J, Hermans M, Dambre J, wyffels F. A differentiable physics engine for deep learning in robotics. Röhrbein F, editor. FRONTIERS IN NEUROROBOTICS. 2019;13(6).
IEEE
[1]
J. Degrave, M. Hermans, J. Dambre, and F. wyffels, “A differentiable physics engine for deep learning in robotics,” FRONTIERS IN NEUROROBOTICS, vol. 13, no. 6, 2019.
@article{8600966,
  abstract     = {An important field in robotics is the optimization of controllers. Currently, robots are often treated as a black box in this optimization process, which is the reason why derivative-free optimization methods such as evolutionary algorithms or reinforcement learning are omnipresent. When gradient-based methods are used, models are kept small or rely on finite difference approximations for the Jacobian. This method quickly grows expensive with increasing numbers of parameters, such as found in deep learning. We propose an implementation of a modern physics engine, which can differentiate control parameters. This engine is implemented for both CPU and GPU. Firstly, this paper shows how such an engine speeds up the optimization process, even for small problems. Furthermore, it explains why this is an alternative approach to deep Q-learning, for using deep learning in robotics. Finally, we argue that this is a big step for deep learning in robotics, as it opens up new possibilities to optimize robots, both in hardware and software.},
  author       = {Degrave, Jonas and Hermans, Michiel and Dambre, Joni and wyffels, Francis},
  editor       = {Röhrbein, Florian},
  issn         = {1662-5218},
  journal      = {FRONTIERS IN NEUROROBOTICS},
  keywords     = {Differential physics engine,deep learning,backpropagation,Robotics,Simulation Technology},
  language     = {eng},
  number       = {6},
  title        = {A differentiable physics engine for deep learning in robotics},
  url          = {http://dx.doi.org/10.3389/fnbot.2019.00006},
  volume       = {13},
  year         = {2019},
}

Altmetric
View in Altmetric
Web of Science
Times cited: