Assignment of multiplicative mixtures in natural images

Odelia Schwartz, Terrence J. Sejnowski, Peter Dayan

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

2 Citations (Scopus)

Abstract

In the analysis of natural images, Gaussian scale mixtures (GSM) have been used to account for the statistics of filter responses, and to inspire hierarchical cortical representational learning schemes. GSMs pose a critical assignment problem, working out which filter responses were generated by a common multiplicative factor. We present a new approach to solving this assignment problem through a probabilistic extension to the basic GSM, and show how to perform inference in the model using Gibbs sampling. We demonstrate the efficacy of the approach on both synthetic and image data.

Original languageEnglish (US)
Title of host publicationAdvances in Neural Information Processing Systems
PublisherNeural information processing systems foundation
ISBN (Print)0262195348, 9780262195348
StatePublished - 2005
Externally publishedYes
Event18th Annual Conference on Neural Information Processing Systems, NIPS 2004 - Vancouver, BC, Canada
Duration: Dec 13 2004Dec 16 2004

Other

Other18th Annual Conference on Neural Information Processing Systems, NIPS 2004
CountryCanada
CityVancouver, BC
Period12/13/0412/16/04

Fingerprint

Global system for mobile communications
Statistics
Sampling

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Schwartz, O., Sejnowski, T. J., & Dayan, P. (2005). Assignment of multiplicative mixtures in natural images. In Advances in Neural Information Processing Systems Neural information processing systems foundation.

Assignment of multiplicative mixtures in natural images. / Schwartz, Odelia; Sejnowski, Terrence J.; Dayan, Peter.

Advances in Neural Information Processing Systems. Neural information processing systems foundation, 2005.

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

Schwartz, O, Sejnowski, TJ & Dayan, P 2005, Assignment of multiplicative mixtures in natural images. in Advances in Neural Information Processing Systems. Neural information processing systems foundation, 18th Annual Conference on Neural Information Processing Systems, NIPS 2004, Vancouver, BC, Canada, 12/13/04.
Schwartz O, Sejnowski TJ, Dayan P. Assignment of multiplicative mixtures in natural images. In Advances in Neural Information Processing Systems. Neural information processing systems foundation. 2005
Schwartz, Odelia ; Sejnowski, Terrence J. ; Dayan, Peter. / Assignment of multiplicative mixtures in natural images. Advances in Neural Information Processing Systems. Neural information processing systems foundation, 2005.
@inproceedings{22f48fc9c4654c1c8b0c81ab4dc0ab0a,
title = "Assignment of multiplicative mixtures in natural images",
abstract = "In the analysis of natural images, Gaussian scale mixtures (GSM) have been used to account for the statistics of filter responses, and to inspire hierarchical cortical representational learning schemes. GSMs pose a critical assignment problem, working out which filter responses were generated by a common multiplicative factor. We present a new approach to solving this assignment problem through a probabilistic extension to the basic GSM, and show how to perform inference in the model using Gibbs sampling. We demonstrate the efficacy of the approach on both synthetic and image data.",
author = "Odelia Schwartz and Sejnowski, {Terrence J.} and Peter Dayan",
year = "2005",
language = "English (US)",
isbn = "0262195348",
booktitle = "Advances in Neural Information Processing Systems",
publisher = "Neural information processing systems foundation",

}

TY - GEN

T1 - Assignment of multiplicative mixtures in natural images

AU - Schwartz, Odelia

AU - Sejnowski, Terrence J.

AU - Dayan, Peter

PY - 2005

Y1 - 2005

N2 - In the analysis of natural images, Gaussian scale mixtures (GSM) have been used to account for the statistics of filter responses, and to inspire hierarchical cortical representational learning schemes. GSMs pose a critical assignment problem, working out which filter responses were generated by a common multiplicative factor. We present a new approach to solving this assignment problem through a probabilistic extension to the basic GSM, and show how to perform inference in the model using Gibbs sampling. We demonstrate the efficacy of the approach on both synthetic and image data.

AB - In the analysis of natural images, Gaussian scale mixtures (GSM) have been used to account for the statistics of filter responses, and to inspire hierarchical cortical representational learning schemes. GSMs pose a critical assignment problem, working out which filter responses were generated by a common multiplicative factor. We present a new approach to solving this assignment problem through a probabilistic extension to the basic GSM, and show how to perform inference in the model using Gibbs sampling. We demonstrate the efficacy of the approach on both synthetic and image data.

UR - http://www.scopus.com/inward/record.url?scp=33750584146&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33750584146&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33750584146

SN - 0262195348

SN - 9780262195348

BT - Advances in Neural Information Processing Systems

PB - Neural information processing systems foundation

ER -