Statistical models of linear and non-linear contextual interactions in early visual processing

Ruben Coen Cagli, Peter Dayan, Odelia Schwartz

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

7 Citations (Scopus)

Abstract

A central hypothesis about early visual processing is that it represents inputs in a coordinate system matched to the statistics of natural scenes. Simple versions of this lead to Gabor-like receptive fields and divisive gain modulation from local surrounds; these have led to influential neural and psychological models of visual processing. However, these accounts are based on an incomplete view of the visual context surrounding each point. Here, we consider an approximate model of linear and non-linear correlations between the responses of spatially distributed Gabor-like receptive fields, which, when trained on an ensemble of natural scenes, unifies a range of spatial context effects. The full model accounts for neural surround data in primary visual cortex (V1), provides a statistical foundation for perceptual phenomena associated with Li's (2002) hypothesis that V1 builds a saliency map, and fits data on the tilt illusion.

Original languageEnglish (US)
Title of host publicationAdvances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference
Pages369-377
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

Processing
Modulation
Statistics
Statistical Models

ASJC Scopus subject areas

  • Information Systems

Cite this

Coen Cagli, R., Dayan, P., & Schwartz, O. (2009). Statistical models of linear and non-linear contextual interactions in early visual processing. In Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference (pp. 369-377)

Statistical models of linear and non-linear contextual interactions in early visual processing. / Coen Cagli, Ruben; Dayan, Peter; Schwartz, Odelia.

Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. p. 369-377.

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

Coen Cagli, R, Dayan, P & Schwartz, O 2009, Statistical models of linear and non-linear contextual interactions in early visual processing. in Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. pp. 369-377, 23rd Annual Conference on Neural Information Processing Systems, NIPS 2009, Vancouver, BC, Canada, 12/7/09.
Coen Cagli R, Dayan P, Schwartz O. Statistical models of linear and non-linear contextual interactions in early visual processing. In Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. p. 369-377
Coen Cagli, Ruben ; Dayan, Peter ; Schwartz, Odelia. / Statistical models of linear and non-linear contextual interactions in early visual processing. Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. pp. 369-377
@inproceedings{cd2c64bba16d44b09ffb631a8bb0ab06,
title = "Statistical models of linear and non-linear contextual interactions in early visual processing",
abstract = "A central hypothesis about early visual processing is that it represents inputs in a coordinate system matched to the statistics of natural scenes. Simple versions of this lead to Gabor-like receptive fields and divisive gain modulation from local surrounds; these have led to influential neural and psychological models of visual processing. However, these accounts are based on an incomplete view of the visual context surrounding each point. Here, we consider an approximate model of linear and non-linear correlations between the responses of spatially distributed Gabor-like receptive fields, which, when trained on an ensemble of natural scenes, unifies a range of spatial context effects. The full model accounts for neural surround data in primary visual cortex (V1), provides a statistical foundation for perceptual phenomena associated with Li's (2002) hypothesis that V1 builds a saliency map, and fits data on the tilt illusion.",
author = "{Coen Cagli}, Ruben and Peter Dayan and Odelia Schwartz",
year = "2009",
language = "English (US)",
isbn = "9781615679119",
pages = "369--377",
booktitle = "Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference",

}

TY - GEN

T1 - Statistical models of linear and non-linear contextual interactions in early visual processing

AU - Coen Cagli, Ruben

AU - Dayan, Peter

AU - Schwartz, Odelia

PY - 2009

Y1 - 2009

N2 - A central hypothesis about early visual processing is that it represents inputs in a coordinate system matched to the statistics of natural scenes. Simple versions of this lead to Gabor-like receptive fields and divisive gain modulation from local surrounds; these have led to influential neural and psychological models of visual processing. However, these accounts are based on an incomplete view of the visual context surrounding each point. Here, we consider an approximate model of linear and non-linear correlations between the responses of spatially distributed Gabor-like receptive fields, which, when trained on an ensemble of natural scenes, unifies a range of spatial context effects. The full model accounts for neural surround data in primary visual cortex (V1), provides a statistical foundation for perceptual phenomena associated with Li's (2002) hypothesis that V1 builds a saliency map, and fits data on the tilt illusion.

AB - A central hypothesis about early visual processing is that it represents inputs in a coordinate system matched to the statistics of natural scenes. Simple versions of this lead to Gabor-like receptive fields and divisive gain modulation from local surrounds; these have led to influential neural and psychological models of visual processing. However, these accounts are based on an incomplete view of the visual context surrounding each point. Here, we consider an approximate model of linear and non-linear correlations between the responses of spatially distributed Gabor-like receptive fields, which, when trained on an ensemble of natural scenes, unifies a range of spatial context effects. The full model accounts for neural surround data in primary visual cortex (V1), provides a statistical foundation for perceptual phenomena associated with Li's (2002) hypothesis that V1 builds a saliency map, and fits data on the tilt illusion.

UR - http://www.scopus.com/inward/record.url?scp=84858711457&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84858711457&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84858711457

SN - 9781615679119

SP - 369

EP - 377

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

ER -