Hierarchical Visuomotor Control of Humanoids

Josh Merel, Arun Ahuja, Vu Pham, Saran Tunyasuvunakool, Siqi Liu, Dhruva Tirumala, Nicolas Heess, Greg Wayne

ICLR 2019

/iclr/2019/merel2019iclr-hierarchical/

Abstract

We aim to build complex humanoid agents that integrate perception, motor control, and memory. In this work, we partly factor this problem into low-level motor control from proprioception and high-level coordination of the low-level skills informed by vision. We develop an architecture capable of surprisingly flexible, task-directed motor control of a relatively high-DoF humanoid body by combining pre-training of low-level motor controllers with a high-level, task-focused controller that switches among low-level sub-policies. The resulting system is able to control a physically-simulated humanoid body to solve tasks that require coupling visual perception from an unstabilized egocentric RGB camera during locomotion in the environment. Supplementary video link: https://youtu.be/fBoir7PNxPk

PDF ICLR Semantic Scholar

Cite

Text

Merel et al. "Hierarchical Visuomotor Control of Humanoids." International Conference on Learning Representations, 2019.

Markdown

[Merel et al. "Hierarchical Visuomotor Control of Humanoids." International Conference on Learning Representations, 2019.](https://mlanthology.org/iclr/2019/merel2019iclr-hierarchical/)

BibTeX

@inproceedings{merel2019iclr-hierarchical,
  title     = {{Hierarchical Visuomotor Control of Humanoids}},
  author    = {Merel, Josh and Ahuja, Arun and Pham, Vu and Tunyasuvunakool, Saran and Liu, Siqi and Tirumala, Dhruva and Heess, Nicolas and Wayne, Greg},
  booktitle = {International Conference on Learning Representations},
  year      = {2019},
  url       = {https://mlanthology.org/iclr/2019/merel2019iclr-hierarchical/}
}