Stereo integration, mean Field theory and psychophysics

Alan L. Yuille, Davi Geiger, Heinrich H. Bülthoff

Research output: Contribution to journalArticle

Abstract

We describe a theoretical formulation for stereo in terms of the Bayesian approach to vision and relates it to psychophysical experiments. The formulation enables us to integrate the depth information from different types of matching primitives, or from different vision modules. We solve the correspondence problem using compatibility constraints between features and prior assumptions on the interpolated surfaces that result from the matching. We use techniques from statistical physics to show how our theory relates to previous work. Finally we show that, by a suitable choice of prior assumptions about surfaces, the theory is consistent with some psychophysical experiments which investigate the relative importance of different matching primitives.

Original languageEnglish (US)
Pages (from-to)423-442
Number of pages20
JournalNetwork: Computation in Neural Systems
Volume2
Issue number4
DOIs
StatePublished - 1991

Fingerprint

Psychophysics
Bayes Theorem
Physics

ASJC Scopus subject areas

  • Neuroscience (miscellaneous)
  • Neuroscience(all)

Cite this

Stereo integration, mean Field theory and psychophysics. / Yuille, Alan L.; Geiger, Davi; Bülthoff, Heinrich H.

In: Network: Computation in Neural Systems, Vol. 2, No. 4, 1991, p. 423-442.

Research output: Contribution to journalArticle

Yuille, Alan L. ; Geiger, Davi ; Bülthoff, Heinrich H. / Stereo integration, mean Field theory and psychophysics. In: Network: Computation in Neural Systems. 1991 ; Vol. 2, No. 4. pp. 423-442.
@article{6ecab24c9c1644c38d1ec97efc19c423,
title = "Stereo integration, mean Field theory and psychophysics",
abstract = "We describe a theoretical formulation for stereo in terms of the Bayesian approach to vision and relates it to psychophysical experiments. The formulation enables us to integrate the depth information from different types of matching primitives, or from different vision modules. We solve the correspondence problem using compatibility constraints between features and prior assumptions on the interpolated surfaces that result from the matching. We use techniques from statistical physics to show how our theory relates to previous work. Finally we show that, by a suitable choice of prior assumptions about surfaces, the theory is consistent with some psychophysical experiments which investigate the relative importance of different matching primitives.",
author = "Yuille, {Alan L.} and Davi Geiger and B{\"u}lthoff, {Heinrich H.}",
year = "1991",
doi = "10.1088/0954-898X_2_4_006",
language = "English (US)",
volume = "2",
pages = "423--442",
journal = "Network: Computation in Neural Systems",
issn = "0954-898X",
publisher = "Informa Healthcare",
number = "4",

}

TY - JOUR

T1 - Stereo integration, mean Field theory and psychophysics

AU - Yuille, Alan L.

AU - Geiger, Davi

AU - Bülthoff, Heinrich H.

PY - 1991

Y1 - 1991

N2 - We describe a theoretical formulation for stereo in terms of the Bayesian approach to vision and relates it to psychophysical experiments. The formulation enables us to integrate the depth information from different types of matching primitives, or from different vision modules. We solve the correspondence problem using compatibility constraints between features and prior assumptions on the interpolated surfaces that result from the matching. We use techniques from statistical physics to show how our theory relates to previous work. Finally we show that, by a suitable choice of prior assumptions about surfaces, the theory is consistent with some psychophysical experiments which investigate the relative importance of different matching primitives.

AB - We describe a theoretical formulation for stereo in terms of the Bayesian approach to vision and relates it to psychophysical experiments. The formulation enables us to integrate the depth information from different types of matching primitives, or from different vision modules. We solve the correspondence problem using compatibility constraints between features and prior assumptions on the interpolated surfaces that result from the matching. We use techniques from statistical physics to show how our theory relates to previous work. Finally we show that, by a suitable choice of prior assumptions about surfaces, the theory is consistent with some psychophysical experiments which investigate the relative importance of different matching primitives.

UR - http://www.scopus.com/inward/record.url?scp=0010533260&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0010533260&partnerID=8YFLogxK

U2 - 10.1088/0954-898X_2_4_006

DO - 10.1088/0954-898X_2_4_006

M3 - Article

VL - 2

SP - 423

EP - 442

JO - Network: Computation in Neural Systems

JF - Network: Computation in Neural Systems

SN - 0954-898X

IS - 4

ER -