The "independent components" of natural scenes are edge filters

Vision Research
A J Bell, T J Sejnowski

Abstract

It has previously been suggested that neurons with line and edge selectivities found in primary visual cortex of cats and monkeys form a sparse, distributed representation of natural scenes, and it has been reasoned that such responses should emerge from an unsupervised learning algorithm that attempts to find a factorial code of independent visual features. We show here that a new unsupervised learning algorithm based on information maximization, a nonlinear "infomax" network, when applied to an ensemble of natural scenes produces sets of visual filters that are localized and oriented. Some of these filters are Gabor-like and resemble those produced by the sparseness-maximization network. In addition, the outputs of these filters are as independent as possible, since this infomax network performs Independent Components Analysis or ICA, for sparse (super-gaussian) component distributions. We compare the resulting ICA filters and their associated basis functions, with other decorrelating filters produced by Principal Components Analysis (PCA) and zero-phase whitening filters (ZCA). The ICA filters have more sparsely distributed (kurtotic) outputs on natural scenes. They also resemble the receptive fields of simple cells in visua...Continue Reading

References

Jan 1, 1992·Biological cybernetics·J H van Hateren
Jan 1, 1990·Biological cybernetics·P Földiák
Dec 1, 1987·Journal of the Optical Society of America. A, Optics and Image Science·D J Field
Jul 1, 1985·Journal of the Optical Society of America. A, Optics and Image Science·J G Daugman
Mar 1, 1968·The Journal of Physiology·D H Hubel, T N Wiesel
Feb 29, 1980·Proceedings of the Royal Society of London. Series B, Containing Papers of a Biological Character·D Marr, E Hildreth
Nov 1, 1995·Neural Computation·A J Bell, T J Sejnowski
Aug 2, 1994·Proceedings of the National Academy of Sciences of the United States of America·C C Law, L N Cooper
May 1, 1996·Network : Computation in Neural Systems·A J Bell, T J Sejnowski
May 1, 1996·Network : Computation in Neural Systems·G F Harpur, R W Prager
May 1, 1996·Network : Computation in Neural Systems·B A Olshausen, D J Field
Jan 1, 1997·IEEE Transactions on Neural Networks·J KarhunenJ Joutsensalo
Jan 1, 1996·IEEE Transactions on Neural Networks·A Hirose, R Eckmiller

❮ Previous
Next ❯

Citations

Dec 15, 2004·Biological cybernetics·Li Zhao
Sep 29, 2005·Biological cybernetics·Kareem I Batarseh
Oct 21, 2006·Journal of Computational Neuroscience·Martin Rehn, Friedrich T Sommer
Aug 29, 2009·Journal of Computational Neuroscience·Stuart D WickHermann Riecke
Nov 13, 2008·Cognitive Neurodynamics·Sheng Li, Si Wu
Jul 10, 2004·Journal of Physiology, Paris·Antonio Turiel, Néstor Parga
Jul 10, 2004·Journal of Physiology, Paris·Peggy SerièsYves Frégnac
Jun 1, 2004·Neurocomputing·Odelia SchwartzTerrence J Sejnowski
Jan 19, 2000·Journal of Neuroscience Methods·M LaubachM A Nicolelis
Feb 13, 2001·Trends in Neurosciences·G D BrownT J Sejnowski
Nov 23, 2001·Neural Networks : the Official Journal of the International Neural Network Society·K Okajima
Apr 4, 2003·Neural Networks : the Official Journal of the International Neural Network Society·Simone Fiori
Jul 10, 2003·Neural Networks : the Official Journal of the International Neural Network Society·Liming Zhang, Jianfeng Mei
Mar 29, 2003·Neural Networks : the Official Journal of the International Neural Network Society·Aapo Hyvärinen, Petteri Pajunen
Mar 29, 2003·Neural Networks : the Official Journal of the International Neural Network Society·K Doya
May 15, 2003·Current Opinion in Neurobiology·Eero P Simoncelli
May 4, 2005·Trends in Cognitive Sciences·James V Stone
Mar 25, 2009·Visual Neuroscience·Constantin A Rothkopf, Dana H Ballard
Mar 6, 2002·The European Journal of Neuroscience·Wolfgang EinhäuserKonrad P Körding
Aug 12, 2009·Proceedings of the National Academy of Sciences of the United States of America·Marc Schönwiesner, Robert J Zatorre
Oct 7, 2010·Proceedings of the National Academy of Sciences of the United States of America·Gasper TkacikVijay Balasubramanian
Apr 5, 2012·Proceedings of the National Academy of Sciences of the United States of America·Jason M SamondsTai Sing Lee
Mar 14, 1998·Proceedings of the National Academy of Sciences of the United States of America·M J McKeownT J Sejnowski
May 9, 1998·Proceedings of the National Academy of Sciences of the United States of America·E M Callaway
Aug 1, 2009·Visual Cognition·Christopher KananGarrison W Cottrell
Sep 1, 2011·Cognitive Neuroscience·Floris P de Lange, Harold Bekkering
Jun 23, 1999·Network : Computation in Neural Systems·M G Thomson
Mar 1, 2000·Network : Computation in Neural Systems·P Reinagel, A M Zador
Oct 3, 2000·Network : Computation in Neural Systems·P O Hoyer, A Hyvärinen
Aug 27, 2003·Network : Computation in Neural Systems·Antonio Torralba, Aude Oliva
Aug 27, 2003·Network : Computation in Neural Systems·Yan Karklin, Michael S Lewicki
Aug 27, 2003·Network : Computation in Neural Systems·Jarmo Hurri, Aapo Hyvärinen
Aug 27, 2003·Network : Computation in Neural Systems·Roberto Valerio, Rafael Navarro
Dec 23, 1998·Network : Computation in Neural Systems·J P NadalN Parga
Dec 23, 1998·Network : Computation in Neural Systems·R P Rao, D H Ballard
Apr 30, 1999·Network : Computation in Neural Systems·C J Webber
May 14, 2009·Journal of Neural Engineering·A B Ajiboye, R F Weir
May 28, 2013·Brain Connectivity·Alexandre R FrancoAndrew R Mayer
Apr 2, 1998·Proceedings. Biological Sciences·J H van Hateren, A van der Schaaf
Feb 12, 2000·Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences·A J Bell
Mar 21, 2000·Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences·K J Friston
May 25, 2002·Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences·Wilson S Geisler, Randy L Diehl
Mar 11, 2003·Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences·Ulrich Hillenbrand, J Leo van Hemmen

❮ Previous
Next ❯

Related Concepts

Related Feeds

BRAIN Initiative Cell Census Network (BICCN)

The BRAIN Initiative Cell Census Network aims to identify and provide experimental access to the different brain cell types to determine their roles in health and disease. Discover the latest research from researchers in the BRAIN Initiative Cell Census Network here.