Vision and the statistics of the visual environment

Current Opinion in Neurobiology

Volumn 13, Issue 2, 2003, Pages 144-149

  Simoncelli, Eero P ^a  

^a NONE   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

AUDITORY SYSTEM; BAYES THEOREM; COLOR DISCRIMINATION; EXPERIMENTAL MODEL; HUMAN; HYPOTHESIS; INFORMATION SCIENCE; LIGHT ABSORPTION; NERVE POTENTIAL; NOISE; PRIORITY JOURNAL; RETINA GANGLION CELL; REVIEW; VISION; VISUAL CORTEX; VISUAL INFORMATION; VISUAL SYSTEM;

EID: 0038734336     PISSN: 09594388     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0959-4388(03)00047-3     Document Type: Review

References (66)

1. Marr D: Vision: A Computational Investigation into the Human Representation and Processing of Visual Information. San Francisco: WH Freeman and Company; 1982.
2. Simoncelli E., Olshausen B. Natural image statistics and neural representation. Annu. Rev. Neurosci. 24:2001;1193-1216.
3. Barlow HB: Redundancy reduction revisited. Network 2001, 12:241-253.
4. Reinagel P., Laughlin S. Natural stimulus statistics. Network. 12:2001;237-240.
5. Knill DC, Richards W: Perception as Bayesian Inference. Edited by Knill DC, Richards W. Cambridge, UK: Cambridge University Press; 1996.
6. Rao R.P.N., Ballard D.H. Predictive coding in the visual cortex: a functional interpretation of some extraclassical receptive-field effects. Nat. Neurosci. 2:1999;79-87.
7. Kersten D, Schrater P: Pattern inference theory: a probabilistic approach to vision. In Perception and the Physical World. Edited by Heyer D, Mausfeld R. Chichester, UK: Wiley; 2002:191-228.
8. Schrater P, Kersten D: Vision, psychophysics and Bayes. In Probabilistic Models of the Brain. Edited by Rao RPN, Olshausen BA, Lewicki MS. Cambridge, MA: MIT Press; 2002:37-60.
9. Mamassian P, Landy M, Maloney LT: Bayesian modeling of visual perception. In Probabilistic Models of the Brain. Edited by Rao RPN, Olshausen BA, Lewicki MS. Cambridge, MA: MIT Press; 2002:13-36.
10. Maloney LT: Statistical decision theory and biological vision. In Perception and the Physical World. Edited by Heyer D, Mausfeld R. Chichester, UK: Wiley; 2002:145-189.
11. Cover T, Thomas J: Elements of Information Theory. New York: Wiley; 1991.
12. Barlow HB: Possible principles underlying the transformation of sensory messages. In Sensory Communication. Edited by Rosenblith WA. Cambridge, MA: MIT Press; 1961:217-234.
13. Panzeri S., Schultz S.R., Treves A., Rolls E. Correlations and encoding of information in the nervous system. Proc. R Soc. B Biol. Sci. 266:1999;1001-1012.
14. Dayan P, Abbott LF: Theoretical Neuroscience. Cambridge, MA: MIT Press; 2001.
15. Attneave F. Some informational aspects of visual perception. Psych Rev. 61:1954;183-193.
16. Srinivasan M.V., Laughlin S.B., Dubs A. Predictive coding: a fresh view of inhibition in the retina. Proc. R Soc. Lond. B Biol. Sci. 216:1982;427-459.
17. Atick J.J., Redlich A.N. Towards a theory of early visual processing. Neural. Comput. 2:1990;308-320.
18. van Hateren J.H. A theory of maximizing sensory information. Biol. Cybern. 68:1992;23-29.
19. Field D.J. What is the goal of sensory coding? Neural. Comput. 6:1994;559-601.
20. Rieke F., Bodnar D.A., Bialek W. Naturalistic stimuli increase the rate and efficiency of information transmission by primary auditory afferents. Proc. R Soc. Lond. Ser. B. 262:1995;259-265.
21. Rieke F, Warland D, de-Ruyter-van-Steveninck RR, Bialek W: Spikes: Exploring the Neural Code. Cambridge, MA: MIT Press; 1997.
22. Vinje W.E., Gallant J.L. Sparse coding and decorrelation in primary visual cortex during natural vision. Science. 287:2000;1273-1276.
23. Nirenberg S., Carcieri S.M., Jacobs A.L., Latham P.E. Retinal ganglion cells act largely as independent encoders. Nature. 411:2001;698-701 This is an important paper, as it is one of the first to attempt to directly quantify the redundancy between pairs of neurons under naturalistic stimulation conditions. The redundancy measure used is unusual and is motivated with respect to decoding (as opposed to encoding). This work has generated a lively controversy (see the discussion on Peter Latham's web page URL: http://culture.neurobio.ucla.edu/∼pel/critics/ind.rgc_nature01.html ), which should be resolved through careful discussion and comparison of alternative measures.
24. Vinje W.E., Gallant J.L. Natural stimulation of the nonclassical receptive field increases information transmission efficiency in V1. J. Neurosci. 22:2002;2904-2915 This study shows that the information rate, information per spike, and transmission efficiency increase as the spatial extent of a naturalistic stimulus is increased to extend beyond the classical receptive field.
25. Chechik G, Globerson A, Tishby N, Anderson MJ, Young ED, Nelken I: Group redundancy measures reveal redundancy reduction in the auditory pathway. In Advanced Neural Information Processing Systems (NIPS*01), vol 14. Edited by Dietterich TG, Becker S, Ghahramani Z. Cambridge, MA: MIT Press; 2002.
26. Laughlin S.B. A simple coding procedure enhances a neuron's information capacity. Z. Naturforsch. 36c:1981;910-912.
27. Dan Y., Atick J.J., Reid R.C. Efficient coding of natural scenes in the lateral geniculate nucleus: experimental test of a computational theory. J. Neurosci. 16:1996;3351-3362.
28. Baddeley R., Abbott L.F., Booth M.C., Sengpiel F., Freeman T., Wakeman E.A., Rolls E.T. Responses of neurons in primary and inferior temporal visual cortices to natural scenes. Proc. R Soc. Lond. B Biol. Sci. B264:1998;1775-1783.
29. Linsker R. Self-organization in a perceptual network. Computer. 21:1988;105-117.
30. Barlow H.B., Kaushal T.P., Mitchison G.J. Finding minimum entropy codes. Neural. Comput. 1:1989;412-423.
31. Sanger T.D. Optimal unsupervised learning in a single-layer network. Neural. Netw. 2:1989;459-473.
32. Foldiak P. Forming sparse representations by local anti-hebbian learning. Biol. Cybern. 64:1990;165-170.
33. Atick J.J. Could information theory provide an ecological theory of sensory processing? Netw. Computat. Neural. Syst. 3:1992;213-251.
34. Olshausen B.A., Field D.J. Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature. 381:1996;607-609.
35. Bell A.J., Sejnowski T.J. The 'independent components' of natural scenes are edge filters. Vision Res. 37:1997;3327-3338.
36. van Hateren J.H., Ruderman D.L. Independent component analysis of natural image sequences yields spatio-temporal filters similar to simple cells in primary visual cortex. Proc. R Soc. Lond. Ser. B. 265:1998;2315-2320.
37. Simoncelli EP, Schwartz O: Image statistics and cortical normalization models. In Advanced Neural Information Processing Systems, vol 11. Edited by Kearns MS, Solla SA, Cohn DA. Cambridge, MA: MIT Press; 1999:153-159.
38. Reinagel P. How do visual neurons respond in the real world? Curr. Opin. Neurobiol. 11:2001;437-442 The author presents a good overview of recent physiological experiments that utilize naturalistic stimuli.
39. Treves A., Panzeri S., Rolls E.T., Booth M., Wakeman E.A. Firing rate distributions and efficiency of information transmission of inferior temporal cortex neurons to natural visual stimulation. Neural. Comput. 11:1999;601-631.
40. de Polavieja G.G. Errors drive the evolution of biological signaling to costly codes. J. Theor. Biol. 214:2002;657-664.
41. Meister M., Lagnado L., Baylor D.A. Concerted signaling by retinal ganglion cells. Science. 270:1995;1207-1210.
42. Reich D.S., Mechler F., Victor J.D. Independent and redundant information of nearby cortical neurons. Science. 294:2001;2566-2568 This study shows that responses of clusters of V1 neurons to white-noise stimuli indicate a fairly low level of redundancy, for clusters of up to six neurons. Although this does not directly bear on the Efficient Coding Hypothesis (because of the choice of stimuli), it seems surprising, given the results of Rieke et al. [20].
43. Wiener M.C., Oram M.W., Liu Z., Richmond B.J. Consistency of encoding in monkey visual cortex. J. Neurosci. 21:2001;8210-8221 The authors demonstrate that neurons in areas V1 transmit roughly equal amounts of information about a variety of stimuli (including bars, gratings, Walsh patterns, and photographic images). Their information measure is based on a spike count code taken over 400 msec intervals, and an assumption that spike count variability followed a Gaussian distribution.
44. Lewen G, Bialek W, Steveninck RR: Neural coding of natural motion stimuli. Network 2001, 12:317-329.
45. Kern R., Petereit C., Egelhaaf M. Neural processing of naturalistic optic flow. J. Neurosci. 21:2001;RC139.
46. Ringach D., Hawken M., Shapley R. Receptive field structure of neurons in monkey primary visual cortex revealed by stimulation with natural image sequences. J. Vis. 2:2002;12-24.
47. Sharpee T, Rust NC, Bialek W: Maximizing informative dimensions: analyzing neural responses to natural signals. In Advanced Neural Information Processing Systems (NIPS*02), vol 15. Edited by Becker S, Thrun S, Obermayer K. Cambridge, MA: MIT Press; 2003.
48. Paninski L: Convergence properties of spike-triggered analysis techniques. In Advanced Neural Information Processing Systems (NIPS*02), vol 15. Edited by Becker S, Thrun S, Obermayer K. Cambridge, MA: MIT Press; 2003.
49. Schwartz O., Simoncelli E.P. Natural signal statistics and sensory gain control. Nat. Neurosci. 4:2001;819-825 We demonstrate that natural images and sounds exhibit higher-order statistical dependencies that may be reduced using divisive normalization, a form of local contrast gain control. We also show that when the parameters of such a model are optimized to reduce redundancy, the resulting model provides a good account of several non-linear properties of early visual and auditory neurons.
50. van Hateren J.H., Snippe H.P. Information theoretical evaluation of parametric models of gain control in blowfly photoreceptor cells. Vision Res. 41:2001;1851-1865 This study derives a nearly optimal gain control model for blowfly photoreceptors that is based on information-theoretic measures in the context of naturalistic stimuli. It remains to be seen whether the resulting model is quantitatively consistent with physiological behaviors.
51. Zetzsche C., Rohrbein F. Nonlinear and extra-classical receptive field properties and the statistics of natural scenes. Network. 12:2001;331-350 This study examines a variety of higher-order statistical properties of images, and their relationship to cortical models. In particular, the authors argue that gain-control models, and cascades of linear operators and non-linear scalar operations are well-suited for natural image structures, as well as being consistent with non-linear properties of cortical neurons.
52. Hoyer P., Hyvarinen A. A multi-layer sparse coding network learns contour coding from natural images. Vision Res. 42:2002;1593-1605 The authors use a sparse-coding model to derive an optimal basis set for processing the responses of model complex cells to natural images. The resulting optimal basis set encodes extended contours.
53. Levy W.B., Baxter R.A. Energy efficient neural codes. Neural. Comput. 8:1996;531-543.
54. Laughlin S.B., de-Ruyter-van-Steveninck R., Anderson J.C. The metabolic cost of information. Nat. Neurosci. 1:1998;36-41.
55. Balasubramanian V., Berry M.J. A test of metabolically efficient coding in the retina. Network. 13:2002;531-552.
56. Gottschalk A. Derivation of the visual contrast response function by maximizing information rate. Neural. Comput. 14:2002;527-542.
57. Barlow HB, Foldiak P: Adaptation and decorrelation in the cortex. In The Computing Neuron. Edited by Durbin R, Miall C, Mitchinson G. New York: Addison-Wellesley; 1989:54-72.
58. Barlow HB: A theory about the functional role and synaptic mechanism of visual aftereffects. In Vision: Coding and Efficiency. Edited by Blakemore C. Cambridge, UK: Cambridge University Press, 1990.
59. Wainwright M.J. Visual adaptation as optimal information transmission. Vision Res. 39:1999;3960-3974.
60. Brenner N., Bialek W., de-Ruyter-van-Steveninck R.R. Adaptive rescaling maximizes information transmission. Neuron. 26:2000;695-702.
61. Fairhall A., Lewen G., Bialek W., de-Ruyter-van-Steveninck R. Efficiency and ambiguity in an adaptive neural code. Nature. 412:2001;787-792 This study elaborates on previous work examining changes in the neural response caused by changes by contrast (variance) of white-noise stimuli. The study shows that the dynamic properties of adaptation appear to be optimized for representation of both the statistics of the stimulus ensemble and the specific attributes of the most recently presented stimulus.
62. Hyvarinen A, Hoyer P: Emergence of topography and complex cell properties from natural images using extensions of ICA. In Advanced Neural Information Processing Systems, vol 12. Edited by Solla SA, Leen TK, Muller KR. Cambridge, MA: MIT Press; 2000:827-833.
63. Kayser C, Einhauser W, Dummer O, Konig P, Kording KP: Extracting slow subspaces from natural videos leads to complex cells. In Proceedings International Conference on Artificial Neural Networks (ICANN-01). Edited by Dorfiner G, Bischof H, Hornik K. Vienna: Springer-Verlag 2001:1075-1080.
64. Berkes P, Wiskott L: Applying slow feature analysis to image sequences yields a rich repertoire of complex cell properties. In Proceedings International Conference on Artificial Neural Networks (ICANN- 02). Edited by Dorronsoro JR. Madrid: Springer 2002:81-86.
65. Balboa R.M., Grzywacz N.M. The role of early lateral inhibition: more than maximizing luminance information. Vision Res. 17:2000;77-89.
66. von-der-Twer T., MacLeod D.I.A. Optimal nonlinear codes for the perception of natural colours. Network. 12:2001;395-407.