Learning physical intuition of block towers by example

Adam Lerer, Sam Gross, Robert Fergus

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Wooden blocks are a common toy for infants, allowing them to develop motor skills and gain intuition about the physical behavior of the world. In this paper, we explore the ability of deep feedforward models to learn such intuitive physics. Using a 3D game engine, we create small towers of wooden blocks whose stability is randomized and render them collapsing (or remaining upright). This data allows us to train large convolutional network models which can accurately predict the outcome, as well as estimating the block trajectories. The models are also able to generalize in two important ways: (i) to new physical scenarios, e.g. towers with an additional block and (ii) to images of real wooden blocks, where it obtains a performance comparable to human subjects.

Original languageEnglish (US)
Title of host publication33rd International Conference on Machine Learning, ICML 2016
PublisherInternational Machine Learning Society (IMLS)
Pages648-656
Number of pages9
Volume1
ISBN (Electronic)9781510829008
StatePublished - 2016
Event33rd International Conference on Machine Learning, ICML 2016 - New York City, United States
Duration: Jun 19 2016Jun 24 2016

Other

Other33rd International Conference on Machine Learning, ICML 2016
CountryUnited States
CityNew York City
Period6/19/166/24/16

Fingerprint

Towers
Physics
Trajectories
Engines

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software
  • Computer Networks and Communications

Cite this

Lerer, A., Gross, S., & Fergus, R. (2016). Learning physical intuition of block towers by example. In 33rd International Conference on Machine Learning, ICML 2016 (Vol. 1, pp. 648-656). International Machine Learning Society (IMLS).

Learning physical intuition of block towers by example. / Lerer, Adam; Gross, Sam; Fergus, Robert.

33rd International Conference on Machine Learning, ICML 2016. Vol. 1 International Machine Learning Society (IMLS), 2016. p. 648-656.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Lerer, A, Gross, S & Fergus, R 2016, Learning physical intuition of block towers by example. in 33rd International Conference on Machine Learning, ICML 2016. vol. 1, International Machine Learning Society (IMLS), pp. 648-656, 33rd International Conference on Machine Learning, ICML 2016, New York City, United States, 6/19/16.
Lerer A, Gross S, Fergus R. Learning physical intuition of block towers by example. In 33rd International Conference on Machine Learning, ICML 2016. Vol. 1. International Machine Learning Society (IMLS). 2016. p. 648-656
Lerer, Adam ; Gross, Sam ; Fergus, Robert. / Learning physical intuition of block towers by example. 33rd International Conference on Machine Learning, ICML 2016. Vol. 1 International Machine Learning Society (IMLS), 2016. pp. 648-656
@inproceedings{dd4df395f9fb427d8429973dafe52d83,
title = "Learning physical intuition of block towers by example",
abstract = "Wooden blocks are a common toy for infants, allowing them to develop motor skills and gain intuition about the physical behavior of the world. In this paper, we explore the ability of deep feedforward models to learn such intuitive physics. Using a 3D game engine, we create small towers of wooden blocks whose stability is randomized and render them collapsing (or remaining upright). This data allows us to train large convolutional network models which can accurately predict the outcome, as well as estimating the block trajectories. The models are also able to generalize in two important ways: (i) to new physical scenarios, e.g. towers with an additional block and (ii) to images of real wooden blocks, where it obtains a performance comparable to human subjects.",
author = "Adam Lerer and Sam Gross and Robert Fergus",
year = "2016",
language = "English (US)",
volume = "1",
pages = "648--656",
booktitle = "33rd International Conference on Machine Learning, ICML 2016",
publisher = "International Machine Learning Society (IMLS)",

}

TY - GEN

T1 - Learning physical intuition of block towers by example

AU - Lerer, Adam

AU - Gross, Sam

AU - Fergus, Robert

PY - 2016

Y1 - 2016

N2 - Wooden blocks are a common toy for infants, allowing them to develop motor skills and gain intuition about the physical behavior of the world. In this paper, we explore the ability of deep feedforward models to learn such intuitive physics. Using a 3D game engine, we create small towers of wooden blocks whose stability is randomized and render them collapsing (or remaining upright). This data allows us to train large convolutional network models which can accurately predict the outcome, as well as estimating the block trajectories. The models are also able to generalize in two important ways: (i) to new physical scenarios, e.g. towers with an additional block and (ii) to images of real wooden blocks, where it obtains a performance comparable to human subjects.

AB - Wooden blocks are a common toy for infants, allowing them to develop motor skills and gain intuition about the physical behavior of the world. In this paper, we explore the ability of deep feedforward models to learn such intuitive physics. Using a 3D game engine, we create small towers of wooden blocks whose stability is randomized and render them collapsing (or remaining upright). This data allows us to train large convolutional network models which can accurately predict the outcome, as well as estimating the block trajectories. The models are also able to generalize in two important ways: (i) to new physical scenarios, e.g. towers with an additional block and (ii) to images of real wooden blocks, where it obtains a performance comparable to human subjects.

UR - http://www.scopus.com/inward/record.url?scp=84997703799&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84997703799&partnerID=8YFLogxK

M3 - Conference contribution

VL - 1

SP - 648

EP - 656

BT - 33rd International Conference on Machine Learning, ICML 2016

PB - International Machine Learning Society (IMLS)

ER -