DensePhysNet: Learning Dense Physical Object Representations Via Multi-Step Dynamic Interactions
Name
1906.03853.pdf
Description
Accepted version
Size
4.93 MB
Format
Adobe PDF
Checksum (MD5)
1f2a0b05f2e1d7ff433f341a8a0f2660
Author(s)
Xu, Zhenjia
Wu, Jiajun
Zeng, Andy
Tenenbaum, Joshua
Song, Shuran
Date Issued
2019
Journal
Robotics: Science and Systems XV
Publisher
Robotics: Science and Systems Foundation
Citation
Xu, Zhenjia, Wu, Jiajun, Zeng, Andy, Tenenbaum, Joshua and Song, Shuran. 2019. "DensePhysNet: Learning Dense Physical Object Representations Via Multi-Step Dynamic Interactions." Robotics: Science and Systems XV.
Version
Author's final manuscript
Abstract
We study the problem of learning physical object
representations for robot manipulation. Understanding object
physics is critical for successful object manipulation, but also
challenging because physical object properties can rarely be
inferred from the object’s static appearance. In this paper, we
propose DensePhysNet, a system that actively executes a sequence
of dynamic interactions (e.g., sliding and colliding), and uses a
deep predictive model over its visual observations to learn dense,
pixel-wise representations that reflect the physical properties of
observed objects. Our experiments in both simulation and real
settings demonstrate that the learned representations carry rich
physical information, and can directly be used to decode physical
object properties such as friction and mass. The use of dense
representation enables DensePhysNet to generalize well to novel
scenes with more objects than in training. With knowledge of
object physics, the learned representation also leads to more
accurate and efficient manipulation in downstream tasks than
the state-of-the-art.
MIT Department
Center for Brains, Minds, and Machines
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.15607/RSS.2019.XV.046
