Learning models of object structure

Joseph Schlecht, Jacobus J Barnard

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

6 Citations (Scopus)

Abstract

We present an approach for learning stochastic geometric models of object categories from single view images. We focus here on models expressible as a spatially contiguous assemblage of blocks. Model topologies are learned across groups of images, and one or more such topologies is linked to an object category (e.g. chairs). Fitting learned topologies to an image can be used to identify the object class, as well as detail its geometry. The latter goes beyond labeling objects, as it provides the geometric structure of particular instances. We learn the models using joint statistical inference over category parameters, camera parameters, and instance parameters. These produce an image likelihood through a statistical imaging model. We use trans-dimensional sampling to explore topology hypotheses, and alternate between Metropolis-Hastings and stochastic dynamics to explore instance parameters. Experiments on images of furniture objects such as tables and chairs suggest that this is an effective approach for learning models that encode simple representations of category geometry and the statistics thereof, and support inferring both category and geometry on held out single view images.

Original languageEnglish (US)
Title of host publicationAdvances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference
Pages1615-1623
Number of pages9
StatePublished - 2009
Event23rd Annual Conference on Neural Information Processing Systems, NIPS 2009 - Vancouver, BC, Canada
Duration: Dec 7 2009Dec 10 2009

Other

Other23rd Annual Conference on Neural Information Processing Systems, NIPS 2009
CountryCanada
CityVancouver, BC
Period12/7/0912/10/09

Fingerprint

Topology
Geometry
Stochastic models
Labeling
Cameras
Statistics
Sampling
Imaging techniques
Experiments

ASJC Scopus subject areas

  • Information Systems

Cite this

Schlecht, J., & Barnard, J. J. (2009). Learning models of object structure. In Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference (pp. 1615-1623)

Learning models of object structure. / Schlecht, Joseph; Barnard, Jacobus J.

Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. p. 1615-1623.

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

Schlecht, J & Barnard, JJ 2009, Learning models of object structure. in Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. pp. 1615-1623, 23rd Annual Conference on Neural Information Processing Systems, NIPS 2009, Vancouver, BC, Canada, 12/7/09.
Schlecht J, Barnard JJ. Learning models of object structure. In Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. p. 1615-1623
Schlecht, Joseph ; Barnard, Jacobus J. / Learning models of object structure. Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. pp. 1615-1623
@inproceedings{389ae36ac67c4274a44152584b4296ab,
title = "Learning models of object structure",
abstract = "We present an approach for learning stochastic geometric models of object categories from single view images. We focus here on models expressible as a spatially contiguous assemblage of blocks. Model topologies are learned across groups of images, and one or more such topologies is linked to an object category (e.g. chairs). Fitting learned topologies to an image can be used to identify the object class, as well as detail its geometry. The latter goes beyond labeling objects, as it provides the geometric structure of particular instances. We learn the models using joint statistical inference over category parameters, camera parameters, and instance parameters. These produce an image likelihood through a statistical imaging model. We use trans-dimensional sampling to explore topology hypotheses, and alternate between Metropolis-Hastings and stochastic dynamics to explore instance parameters. Experiments on images of furniture objects such as tables and chairs suggest that this is an effective approach for learning models that encode simple representations of category geometry and the statistics thereof, and support inferring both category and geometry on held out single view images.",
author = "Joseph Schlecht and Barnard, {Jacobus J}",
year = "2009",
language = "English (US)",
isbn = "9781615679119",
pages = "1615--1623",
booktitle = "Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference",

}

TY - GEN

T1 - Learning models of object structure

AU - Schlecht, Joseph

AU - Barnard, Jacobus J

PY - 2009

Y1 - 2009

N2 - We present an approach for learning stochastic geometric models of object categories from single view images. We focus here on models expressible as a spatially contiguous assemblage of blocks. Model topologies are learned across groups of images, and one or more such topologies is linked to an object category (e.g. chairs). Fitting learned topologies to an image can be used to identify the object class, as well as detail its geometry. The latter goes beyond labeling objects, as it provides the geometric structure of particular instances. We learn the models using joint statistical inference over category parameters, camera parameters, and instance parameters. These produce an image likelihood through a statistical imaging model. We use trans-dimensional sampling to explore topology hypotheses, and alternate between Metropolis-Hastings and stochastic dynamics to explore instance parameters. Experiments on images of furniture objects such as tables and chairs suggest that this is an effective approach for learning models that encode simple representations of category geometry and the statistics thereof, and support inferring both category and geometry on held out single view images.

AB - We present an approach for learning stochastic geometric models of object categories from single view images. We focus here on models expressible as a spatially contiguous assemblage of blocks. Model topologies are learned across groups of images, and one or more such topologies is linked to an object category (e.g. chairs). Fitting learned topologies to an image can be used to identify the object class, as well as detail its geometry. The latter goes beyond labeling objects, as it provides the geometric structure of particular instances. We learn the models using joint statistical inference over category parameters, camera parameters, and instance parameters. These produce an image likelihood through a statistical imaging model. We use trans-dimensional sampling to explore topology hypotheses, and alternate between Metropolis-Hastings and stochastic dynamics to explore instance parameters. Experiments on images of furniture objects such as tables and chairs suggest that this is an effective approach for learning models that encode simple representations of category geometry and the statistics thereof, and support inferring both category and geometry on held out single view images.

UR - http://www.scopus.com/inward/record.url?scp=84858727290&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84858727290&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781615679119

SP - 1615

EP - 1623

BT - Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference

ER -