Hierarchical modeling of local image features through L p-nested symmetric distributions

Fabian Sinz, Eero Simoncelli, Matthias Bethge

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

Abstract

We introduce a new family of distributions, called L p-nested symmetric distributions, whose densities are expressed in terms of a hierarchical cascade of L p-norms. This class generalizes the family of spherically and L p-spherically symmetric distributions which have recently been successfully used for natural image modeling. Similar to those distributions it allows for a nonlinear mechanism to reduce the dependencies between its variables. With suitable choices of the parameters and norms, this family includes the Independent Subspace Analysis (ISA) model as a special case, which has been proposed as a means of deriving filters that mimic complex cells found in mammalian primary visual cortex. L p-nested distributions are relatively easy to estimate and allow us to explore the variety of models between ISA and the L p-spherically symmetric models. By fitting the generalized L p-nested model to 8 × 8 image patches, we show that the subspaces obtained from ISA are in fact more dependent than the individual filter coefficients within a subspace. When first applying contrast gain control as preprocessing, however, there are no dependencies left that could be exploited by ISA. This suggests that complex cell modeling can only be useful for redundancy reduction in larger image patches.

Original languageEnglish (US)
Title of host publicationAdvances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference
Pages1696-1704
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

Gain control
Redundancy

ASJC Scopus subject areas

  • Information Systems

Cite this

Sinz, F., Simoncelli, E., & Bethge, M. (2009). Hierarchical modeling of local image features through L p-nested symmetric distributions. In Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference (pp. 1696-1704)

Hierarchical modeling of local image features through L p-nested symmetric distributions. / Sinz, Fabian; Simoncelli, Eero; Bethge, Matthias.

Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. p. 1696-1704.

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

Sinz, F, Simoncelli, E & Bethge, M 2009, Hierarchical modeling of local image features through L p-nested symmetric distributions. in Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. pp. 1696-1704, 23rd Annual Conference on Neural Information Processing Systems, NIPS 2009, Vancouver, BC, Canada, 12/7/09.
Sinz F, Simoncelli E, Bethge M. Hierarchical modeling of local image features through L p-nested symmetric distributions. In Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. p. 1696-1704
Sinz, Fabian ; Simoncelli, Eero ; Bethge, Matthias. / Hierarchical modeling of local image features through L p-nested symmetric distributions. Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference. 2009. pp. 1696-1704
@inproceedings{6993a38df4e641208c8ce090b540ba49,
title = "Hierarchical modeling of local image features through L p-nested symmetric distributions",
abstract = "We introduce a new family of distributions, called L p-nested symmetric distributions, whose densities are expressed in terms of a hierarchical cascade of L p-norms. This class generalizes the family of spherically and L p-spherically symmetric distributions which have recently been successfully used for natural image modeling. Similar to those distributions it allows for a nonlinear mechanism to reduce the dependencies between its variables. With suitable choices of the parameters and norms, this family includes the Independent Subspace Analysis (ISA) model as a special case, which has been proposed as a means of deriving filters that mimic complex cells found in mammalian primary visual cortex. L p-nested distributions are relatively easy to estimate and allow us to explore the variety of models between ISA and the L p-spherically symmetric models. By fitting the generalized L p-nested model to 8 × 8 image patches, we show that the subspaces obtained from ISA are in fact more dependent than the individual filter coefficients within a subspace. When first applying contrast gain control as preprocessing, however, there are no dependencies left that could be exploited by ISA. This suggests that complex cell modeling can only be useful for redundancy reduction in larger image patches.",
author = "Fabian Sinz and Eero Simoncelli and Matthias Bethge",
year = "2009",
language = "English (US)",
isbn = "9781615679119",
pages = "1696--1704",
booktitle = "Advances in Neural Information Processing Systems 22 - Proceedings of the 2009 Conference",

}

TY - GEN

T1 - Hierarchical modeling of local image features through L p-nested symmetric distributions

AU - Sinz, Fabian

AU - Simoncelli, Eero

AU - Bethge, Matthias

PY - 2009

Y1 - 2009

N2 - We introduce a new family of distributions, called L p-nested symmetric distributions, whose densities are expressed in terms of a hierarchical cascade of L p-norms. This class generalizes the family of spherically and L p-spherically symmetric distributions which have recently been successfully used for natural image modeling. Similar to those distributions it allows for a nonlinear mechanism to reduce the dependencies between its variables. With suitable choices of the parameters and norms, this family includes the Independent Subspace Analysis (ISA) model as a special case, which has been proposed as a means of deriving filters that mimic complex cells found in mammalian primary visual cortex. L p-nested distributions are relatively easy to estimate and allow us to explore the variety of models between ISA and the L p-spherically symmetric models. By fitting the generalized L p-nested model to 8 × 8 image patches, we show that the subspaces obtained from ISA are in fact more dependent than the individual filter coefficients within a subspace. When first applying contrast gain control as preprocessing, however, there are no dependencies left that could be exploited by ISA. This suggests that complex cell modeling can only be useful for redundancy reduction in larger image patches.

AB - We introduce a new family of distributions, called L p-nested symmetric distributions, whose densities are expressed in terms of a hierarchical cascade of L p-norms. This class generalizes the family of spherically and L p-spherically symmetric distributions which have recently been successfully used for natural image modeling. Similar to those distributions it allows for a nonlinear mechanism to reduce the dependencies between its variables. With suitable choices of the parameters and norms, this family includes the Independent Subspace Analysis (ISA) model as a special case, which has been proposed as a means of deriving filters that mimic complex cells found in mammalian primary visual cortex. L p-nested distributions are relatively easy to estimate and allow us to explore the variety of models between ISA and the L p-spherically symmetric models. By fitting the generalized L p-nested model to 8 × 8 image patches, we show that the subspaces obtained from ISA are in fact more dependent than the individual filter coefficients within a subspace. When first applying contrast gain control as preprocessing, however, there are no dependencies left that could be exploited by ISA. This suggests that complex cell modeling can only be useful for redundancy reduction in larger image patches.

UR - http://www.scopus.com/inward/record.url?scp=84858738800&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84858738800&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84858738800

SN - 9781615679119

SP - 1696

EP - 1704

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

ER -