DOCSLIB.ORG

A Neural Model of Early Vision: Contrast, Contours, Corners and Surfaces

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
A Neural Model of Early Vision: Contrast, Contours, Corners and Surfaces A Neural Model of Early Vision: Contrast, Contours, Corners and Surfaces Contributions toward an Integrative Architecture of Form and Brightness Perception Thorsten Hansen A University of Ulm Faculty of Computer Science Dept. of Neural Information Processing A Neural Model of Early Vision: Contrast, Contours, Corners and Surfaces Contributions toward an Integrative Architecture of Form and Brightness Perception A Neural Model of Early Vision: Contrast, Contours, Corners and Surfaces Contributions toward an Integrative Architecture of Form and Brightness Perception Thorsten Hansen aus Leer Dissertation zur Erlangung des Doktorgrades Dr. rer. nat. 2003 A University of Ulm Faculty of Computer Science Dept. of Neural Information Processing Dekan: Prof. Dr. G¨unther Palm Erster Gutachter: Prof. Dr. Heiko Neumann Zweiter Gutachter: Prof. Dr. G¨unther Palm Tag der Promotion: 22. September 2002 Abstract The thesis is concerned with the functional modeling of information processing in early and mid- level vision. The mechanisms can be subdivided into two systems, a system for the processing of discontinuities (such as contrast, contours and corners), and a complementary system for the representation of homogeneous surface properties such as brightness. For the robust processing of oriented contrast signals, a mechanism of dominating opponent inhi- bition (DOI) is proposed and integrated into an existing nonlinear simple cell model. We demon- strate that the model with DOI can account for physiological data on luminance gradient reversal. For the processing of both natural and artificial images we show that the new mechanism results in a significant suppression of responses to noisy regions, largely independent of the noise level. This adaptive processing is further examined by a stochastic analysis and numerical evaluations. We also show that contrast-invariant orientation tuning can be achieved in a purely feedforward model based on inhibition. DOI results in a sharpening of the tuning width of model simple cells which are in accordance with empirical findings. The results lead to the proposal of a new functional role of the dominant inhibition as observed empirically, namely to sharpen the orientation tuning and to allow for robust contrast processing under suboptimal, noisy viewing conditions. For the processing of contours, a model of recurrent colinear long-range interaction in V1 is pro- posed. The key properties of the model are excitatory long-range interactions between cells with colinear receptive fields, inhibitory unoriented short-range interactions and modulating feedback. In the model, initial noisy feedforward responses which are part of a more global contour are enhanced, while other responses are suppressed. We show for a number of artificial and natural images that the recurrent long-range processing results in a selective enhancement of coherent activity at contour locations. The competencies of the model are further quantitatively evaluated using two measures of contour saliency and orientation significance. The model also qualita- tively reproduces empirical data on surround suppression and facilitation. We further suggest and examine a model variant using early feedback of grouped responses, showing an even stronger enhancement of contour saliency compared to the standard model. These results may suggest a functional role for the layout of different feedback projections in V1. Regions of intrinsically 2D structures such as corners and junctions are important for both bio- logical and artificial vision systems. We propose a novel scheme for the robust and reliable repre- sentation and detection of junction points, where junctions are characterized by high responses at multiple orientations within an model hypercolumn. The recurrent long-range interaction results in a robust extraction of orientation information. A measurement of circular variance is used to detect and localize junctions. We show for a number of artificial and natural images that localiza- tion accuracy and positive correctness of the junction detection is improved compared to a purely feedforward computation of contour orientations. We also use ROC analysis to compare the new scheme with two other junction detector schemes based on Gaussian curvature and the structure tensor, showing that the new approach performs superior to the standard schemes. Brightness surfaces are reconstructed by a diffusive spreading or filling-in of sparse contrast mea- surements, which is locally controlled by contour signals. A mechanism of confidence-based filling- in is proposed, where a confidence measure ensures a robust selection of sparse contrast signals. We show that the model with confidence-based filling-in can generate brightness surfaces which are invariant against size and shape transformations and can also generate smooth brightness sur- faces even from noisy contrast data, in contrast to standard filling-in. The model can also account for psychophysical data on human brightness perception. We further suggest a new approach for the reconstruction of reference levels, where sparse contrast signals are modulated to carry an additional luminance component. We show for a number of test stimuli that the newly proposed scheme can successfully predict human brightness perception. Overall, we show that basic tasks in early vision can be robustly and efficiently implemented by biologically motivated mechanisms. This leads to a deeper understanding of the functional role of the particular mechanisms and provides the basis for practical applications in technical vision systems. Zusammenfassung Die Dissertation behandelt die Modellierung der Informationsverarbeitung bei der fr¨uhenvisuellen Wahrnehmung. The verschiedenen Mechanismen lassen sich zwei Systemen zuordnen: ein System zur Verarbeitung von Diskontinuit¨aten(Kontraste, Konturen, Ecken) und ein komplement¨ares System zur Repr¨asentation homogener Bereiche (Oberfl¨achen). Im Bereich der Kontrastverarbeitung wird motiviert durch empirische Daten ein Mechanismus der dominanten opponenten Inhibition (DOI) vorschlagen und in ein bestehendes nichtlineares sim- ple cell Modell integriert. Der neue Mechanismus erlaubt die Simulation experimentell gemessener Antworten auf Hell-dunkel-Balken. Bei der Verarbeitung von nat¨urlichen und synthetischen Bildern f¨uhrtder Mechanismus zu einer signifikanten Unterdr¨uckung von Rauschen, weitgehend unabh¨an- gig von der H¨ohedes Rauschlevels. Diese adaptive Verarbeitung wird in einer stochastischen Ana- lyse und in umfangreichen numerischen Evaluationen ausf¨uhrlich untersucht. Weiter wird gezeigt das kontrast-invariantes Orientierungstuning durch eine reine feedforward Verarbeitung realisiert werden kann. DOI f¨uhrtdabei zu einer Versch¨arfungder Tuningkurven in Ubereinstimmung¨ mit empirischen Daten. Aufgrund der Ergebnisse wird als eine neue funktionelle Rolle der empirisch beobachteten dominanten Inhibition die Versch¨arfungdes Orientierungstunings und die robuste Verarbeitung von orientierten Kontrastsignalen vorgeschlagen. Die Konturverarbeitung wird realisiert durch ein rekurrentes Modell langreichweitiger, kolinearer Verbindungen im prim¨arenvisuellen Kortex. Wesentliche Mechanismen sind exzitatorische lang- reichweitige Interaktionen zwischen Zellen mit kolinearen rezeptiven Feldern, inhibitorische orien- tierungsunspezifische kurzreichweitige Interaktionen und modulatorisches Feedback. In dem Mo- dell werden initiale, verrauschte feedforward Messungen durch modulatorisches Feedback verst¨arkt, wenn sie Teil einer zusammenh¨angendenKontur sind, andernfalls abgeschw¨acht. Diese Verst¨arkung koh¨arenter Konturaktivit¨atwird f¨ureine Reihe von nat¨urlichen und synthetischen Bildern gezeigt. Zur quantitativen Evaluierung der Modelleigenschaften wird ein Maß f¨urSalienz und Orien- tierungsvarianz verwendet. Das Modell erlaubt ausserdem die Reproduktion empirischer Daten zu Kontexteffekten. Weiter wird eine Modellvariante vorgeschlagen und untersucht, bei der eine fr¨uheR¨uckkopplung der gruppierten Kontursignale verwendet wird, was zu einer noch gr¨oßeren Verst¨arkungder Salienz von Konturen f¨uhrt. Regionen mit intrinsisch zweidimensionaler Struktur wie Ecken und Kreuzungspunkte spielen eine wichtige Rolle f¨urbiologische und artifizielle Sehsysteme. Wir schlagen ein neues Schema f¨ur die robuste Repr¨asentation und Detektion von Ecken und Kreuzungspunkten vor, basierend auf starken Antworten f¨urmehrere Orientierungen an einem Ort, d. h. innerhalb einer kortikalen Hy- perkolumne. Diese Orientierungsinformation kann robust durch rekurrente langreichweitige In- teraktionen generiert werden. Ausgehend von dieser Repr¨asentation wird ein Maß f¨urdie Orien- tierungsvarianz zur Detektion und Lokalisation der Kreuzungspunkte verwendet. Wir zeigen f¨ur synthetische und nat¨urliche Bilder, dass die Lokalisation und die positive Korrektheit durch die langreichweitige Interaktion verbessert wird, verglichen mit einer reinen feedforward Verarbeitung. Eine ROC Analyse zeigt eine bessere Detektionsleistung des neuen Verfahrens im Vergleich mit zwei anderen Verfahren, basierend auf Gaußscher Kr¨ummung und dem Strukturtensor. Helligkeitsoberfl¨achen werden durch eine laterale Diffusion von sp¨arlichen Kontrastsignalen rekon- struiert, deren Ausbreitung lokal durch Kontursignale geblockt wird. Durch einen Mechanismus des konfidenzbasierten Filling-in werden Positionen, an denen keine Kontrastsignale vorliegen, nicht als Eingabe f¨urden Diffusionsprozess genutzt. Auf diese Weise kann auch bei sp¨arlichen Kontrasten
Load more

Recommended publications
  • A Detailed Model of the Primary Visual Pathway in The
    The Journal of Neuroscience, July 1991, 7 7(7): 1959-l 979 A Detailed Model of the Primary Visual Pathway in the Cat: Comparison of Afferent Excitatory and lntracortical Inhibitory Connection Schemes for Orientation Selectivity Florentin WMg6tter” and Christof Koch Computation and Neural Systems Program, California Institute of Technology, Pasadena, California 91125 In order to arrive at a quantitative understanding of the dy- two inhibitory schemes, local and circular inhibition (a weak namics of cortical neuronal networks, we simulated a de- form of cross-orientation inhibition), is in good agreement tailed model of the primary visual pathway of the adult cat. with observed receptive field properties. The specificity re- This computer model comprises a 5”~ 5” patch of the visual quired to establish these connections during development field at a retinal eccentricity of 4.5” and includes 2048 ON- is low. We propose that orientation selectivity is caused by and OFF-center retinal B-ganglion cells, 8 192 geniculate at least three different mechanisms (“eclectic” model): a X-cells, and 4096 simple cells in layer IV in area 17. The weak afferent geniculate bias, broadly tuned cross-orien- neurons are implemented as improved integrate-and-fire tation inhibition, and some iso-orientation inhibition. units. Cortical receptive fields are determined by the pattern The most surprising finding is that an isotropic connection of afferent convergence and by inhibitory intracortical con- scheme, circular inhibition, in which a cell inhibits all of its nections. Orientation columns are implemented continuously postsynaptic target cells at a distance of approximately 500 with a realistic receptive field scatter and jitter in the pre- pm, enhances orientation tuning and leads to a significant ferred orientations.
    [Show full text]
  • Michael Paul Stryker
    BK-SFN-NEUROSCIENCE_V11-200147-Stryker.indd 372 6/19/20 2:19 PM Michael Paul Stryker BORN: Savannah, Georgia June 16, 1947 EDUCATION: Deep Springs College, Deep Springs, CA (1964–1966) University of Michigan, Ann Arbor, MI, BA (1968) Massachusetts Institute of Technology, Cambridge, MA, PhD (1975) Harvard Medical School, Boston, MA, Postdoctoral (1975–1978) APPOINTMENTS: Assistant Professor of Physiology, University of California, San Francisco (1978–1983) Associate Professor of Physiology, UCSF (1983–1987) Professor of Physiology, UCSF (1987–present) Visiting Professor of Human Anatomy, University of Oxford, England (1987–1988) Co-Director, Neuroscience Graduate Program, UCSF (1988–1994) Chairman, Department of Physiology, UCSF (1994–2005) Director, Markey Program in Biological Sciences, UCSF (1994–1996) William Francis Ganong Endowed Chair of Physiology, UCSF (1995–present) HONORS AND AWARDS (SELECTED): W. Alden Spencer Award, Columbia University (1990) Cattedra Galileiana (Galileo Galilei Chair) Scuola Normale Superiore, Italy (1993) Fellow of the American Association for the Advancement of Science (1999) Fellow of the American Academy of Arts and Sciences (2002) Member of the U.S. National Academy of Sciences (2009) Pepose Vision Sciences Award, Brandeis University (2012) RPB Stein Innovator Award, Research to Prevent Blindness (2016) Krieg Cortical Kudos Discoverer Award from the Cajal Club (2018) Disney Award for Amblyopia Research, Research to Prevent Blindness (2020) Michael Stryker’s laboratory demonstrated the role of spontaneous neural activity as distinguished from visual experience in the prenatal and postnatal development of the central visual system. He and his students created influential and biologically realistic theoretical mathematical models of cortical development. He pioneered the use of the ferret for studies of the central visual system and used this species to delineate the role of neural activity in the development of orientation selectivity and cortical columns.
    [Show full text]
  • Columnar Specificity of Intrinsic Horizontal and Corticocortical Connections in Cat Visual Cortex
    The Journal of Neuroscience, July 1989, g(7): 2432-2442 Columnar Specificity of Intrinsic Horizontal and Corticocortical Connections in Cat Visual Cortex Charles D. Gilbert and Torsten N. Wiesel The Rockefeller University, New York, New York 10021 A prominent and stereotypical feature of cortical circuitry in O’Leary, 194 1; Lorente de No, 1949; Lund and Boothe, 1975). the striate cortex is a plexus of long-range horizontal con- This rather comfortable agreement between anatomy and phys- nections, running for 6-6 mm parallel to the cortical surface, iology was disrupted by the discovery that cortical connections which has a clustered distribution. This is seen for both instead run for many millimeters parallel to the cortical surface, intrinsic cortical connections within a particular cortical area in seeming violation of the principles of receptive field size and and the convergent and divergent connections running be- functional architecture (Fisken et al., 1975; Gilbert and Wiesel, tween area 17 and other cortical areas. To determine if these 1979, 1983). An intriguing pattern of these connections, re- connections are related to the columnar functional architec- vealed by intracellular HRP injections, was that axon collaterals ture of cortex, we combined labeling of the horizontal con- of individual pyramidal cells were distributed in discrete clus- nections by retrograde transport of rhodamine-filled latex ters, spaced from 0.5 to 1 mm apart (Gilbert and Wiesel, 1979, microspheres (beads) and labeling of the orientation col- 1983; Martin and Whitteridge, 1984). Extracellular HRP injec- umns by 2-deoxyglucose autoradiography. We first mapped tions showed that the horizontal connections were convergent the distribution of orientation columns in a small region of as well as divergent and that cells projecting to the injection site area 17 or 16, then made a small injection of beads into the were distributed in a lattice-like pattern (Rockland and Lund, center of an orientation column of defined specificity, and 1982, 1983).
    [Show full text]
  • An Emergent Model of Orientation Selectivity in Cat Visual Cortical Simple Cells
    The Journal of Neuroscience, August 1995, 75(8): 5448-5465 An Emergent Model of Orientation Selectivity in Cat Visual Cortical Simple Cells David C. Sometql Sacha B. Nelson,2 and Mriganka Surl ‘Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 and *Department of Biology and Center for Complex Systems, Brandeis University, Waltham, Massachusetts 02254 It is well known that visual cortical neurons respond vigor- thalamocortical inputs (feedforward excitation) or a combination ously to a limited range of stimulus orientations, while their of intracortical inhibition and thalamocortical convergence. primary afferent inputs, neurons in the lateral geniculate nu- However, “feedforward” and “inhibitory” models (see Fig. 1) cleus (LGN), respond well to all orientations. Mechanisms are inconsistent with key pieces of physiological data and ne- based on intracortical inhibition and/or converging thala- glect the effects of connectionsfrom intracortical excitatory neu- mocottical afferents have previously been suggested to un- rons, which provide the majority of synapsesonto cells in all derlie the generation of cortical orientation selectivity; how- cortical layers (LeVay and Gilbert, 1976; Peters and Payne, ever, these models conflict with experimental data. Here, a 1993; Ahmed et al., 1994). I:4 scale model of a 1700 pm by 200 pm region of layer IV Feedforward models suggestthat cortical neuronsobtain ori- of cat primary visual cortex (area 17) is presented to dem- entation selectivity
    [Show full text]
  • 1 Introduction
    Self-Organization, Plasticity, and Low-level Visual Phenomena in a Laterally Connected Map Mo del of the Primary Visual Cortex Risto Miikkulainen, James A. Bednar, Yo onsuck Cho e, and Joseph Sirosh Department of Computer Sciences The UniversityofTexas at Austin, Austin, TX 78712 fristo,jb ednar,yscho e,[email protected] Abstract Based on a Hebbian adaptation pro cess, the a erent and lateral connections in the RF-LISSOM mo del organize simultaneously and co op eratively, and form structures such as those observed in the primary visual cortex. The neurons in the mo del develop lo cal receptive elds that are organized into orientation, o cular dominance, and size selectivity columns. At the same time, patterned lateral connections form b etween neurons that follow the receptive eld organization. This structure is in a continuously-adapting dynamic equilibrium with the external and intrinsic input, and can account for reorganization of the adult cortex following retinal and cortical lesions. The same learning pro cesses may b e resp onsible for a number of low-level functional phenomena such as tilt aftere ects, and combined with the leaky integrator mo del of the spiking neuron, for segmentation and binding. The mo del can also b e used to verify quantitatively the hyp othesis that the visual cortex forms a sparse, redundancy-reduced enco ding of the input, which allows it to pro cess massive amounts of visual information eciently. 1 Intro duction The primary visual cortex, like many other regions of the neo cortex, is a top ographic map, organized so that adjacent neurons resp ond to adjacent regions of the visual eld.
    [Show full text]
  • Inhibition, Spike Threshold, and Stimulus Selectivity in Primary Visual Cortex
    Neuron Review Inhibition, Spike Threshold, and Stimulus Selectivity in Primary Visual Cortex Nicholas J. Priebe1 and David Ferster2,* 1Section of Neurobiology, The University of Texas at Austin, 1 University Station C0920, Austin, TX 78712, USA 2Department of Neurobiology and Physiology, Northwestern University, 2205 Tech Drive, Evanston, IL 60208, USA *Correspondence: [email protected] DOI 10.1016/j.neuron.2008.02.005 Ever since Hubel and Wiesel described orientation selectivity in the visual cortex, the question of how precise selectivity emerges has been marked by considerable debate. There are essentially two views of how selec- tivity arises. Feed-forward models rely entirely on the organization of thalamocortical inputs. Feedback models rely on lateral inhibition to refine selectivity relative to a weak bias provided by thalamocortical inputs. The debate is driven by two divergent lines of evidence. On the one hand, many response properties appear to require lateral inhibition, including precise orientation and direction selectivity and crossorientation sup- pression. On the other hand, intracellular recordings have failed to find consistent evidence for lateral inhibi- tion. Here we demonstrate a resolution to this paradox. Feed-forward models incorporating the intrinsic non- linear properties of cortical neurons and feed-forward circuits (i.e., spike threshold, contrast saturation, and spike-rate rectification) can account for properties that have previously appeared to require lateral inhibition. Since Hartline described inhibition between adjacent photore- et al., 2000a; Tan et al., 2004; Wehr and Zador, 2003). In addition, ceptors in the limulus retina (Hartline, 1949), the principle of lat- inactivation of the cortical circuit (including both excitatory and eral inhibition has become deeply embedded in neuroscience.
    [Show full text]
  • Neural Connections and Receptive Field Properties in the Primary
    REVIEW I Neural Connections and Receptive Field Properties in the Primary Visual Cortex JOSE-MANUEL ALONSO Department of Psychology University of Connecticut Storrs, Connecticut A cubic millimeter of primary visual cortex contains about 100,000 neurons that are heavily interconnected by intrinsic and extrinsic afferents. The effort of many neuroanatomists over the past has revealed the gen- eral outline of these connections; however, their function remains a mystery. Recently, combined physio- logical and anatomical approaches are beginning to reveal the role of these connections in the generation of cortical receptive fields. A common theme emerges from all these studies: cortical connections are remarkably specific and this specificity is determined in great extent by the type of connection and the neu- ronal response properties. Feedforward connections follow relatively rigid rules of wiring selectively target- ing neurons with receptive fields matched in position and contrast polarity (thalamus → cortical layer 4) or position and orientation selectivity (layer 4 → layers 2 + 3). In contrast, horizontal connections follow more flexible rules connecting distant cells that are not retinotopically aligned and neighboring cells with differ- ent orientation preferences. These differences in connectivity may give a hint on how visual stimuli are processed in the primary visual cortex. An attractive hypothesis is that local stimuli use the highly selective feedforward inputs to reliably drive cortical neurons while background stimuli modulate
    [Show full text]
  • 1 Vision III: Cortical Mechanisms of Vision First You Tell Them What You're
    Vision III: Cortical mechanisms of vision Please sit where you can examine a partner. Michael E. Goldberg, M.D. First you tell them what you’re gonna tell them • The cortical visual system is composed of multiple visual areas with different functions. • V1 neurons describe object features. • The principle of columnar organization. • Two visual streams – ‘what’ and ‘how’ (or ‘where’). • MT neurons describe motion and depth (dorsal stream). • IT neurons describe objects (ventral stream). See the triangle? 1 See the white bar? See the wavy line? Which small square is darker? 2 So • Your visual system does not measure and report the exact physical nature of the visual world. • It collects some data, and makes guesses. • Optical illusions take advantage of the guessing strategies. Roughly 40% of cerebral cortex is involved in vision Remember • Receptive fields in the retina and the lateral geniculate are circular, with center – surround organization. Off surround - inhibits On center - excites 3 The striate cortex – V1 – builds more sophisticated receptive fields from these basic building blocks. Cells describe specific • Contour orientations. • Binocular interaction. • Speed and direction of motion. • Color. David Hubel and Torsten Wiesel won a Nobel Prize in 1981 for describing the properties of striate cortical neurons V1 simple cell is most responsive to an oriented line Off-response On-response 4 Orientation tuning in a V1 simple cell Spikes/second Stimulus Angle (from max) V1 complex cells are sensitive to orientation of stimuli But
    [Show full text]
  • 1 the Death of the Cortical Column?
    Preprint to appear in Studies in History and Philosophy of Science, doi: 10.1016/j.shpsa.2020.09.010. Please quote published version. The Death of the Cortical Column? Patchwork structure and conceptual retirement in neuroscientific practice Philipp Haueis Bielefeld University Department of Philosophy Postfach 100131 D- 33501 Bielefeld Email: [email protected] Phone: +49 52 106 4585 ABSTRACT: In 1981, David Hubel and Torsten Wiesel received the Nobel Prize for their research on cortical columns—vertical bands of neurons with similar functional properties. This success led to the view that “cortical column” refers to the basic building block of the mammalian ne- ocortex. Since the 1990s, however, critics questioned this building block picture of “cortical column” and debated whether this concept is useless and should be replaced with successor concepts. This paper inquires which experimental results after 1981 challenged the building block picture and whether these challenges warrant the elimination “cortical column” from neu- roscientific discourse. I argue that the proliferation of experimental techniques led to a patch- work of locally adapted uses of the column concept. Each use refers to a different kind of cor- tical structure, rather than a neocortical building block. Once we acknowledge this diverse- kinds picture of “cortical column”, the elimination of column concept becomes unnecessary. Rather, I suggest that “cortical column” has reached conceptual retirement: although it cannot be used to identify a neocortical building block, column research is still useful as a guide and cautionary tale for ongoing research. At the same time, neuroscientists should search for alter- native concepts when studying the functional architecture of the neocortex.
    [Show full text]
  • 1 Receptive Fields and Maps in the Visual Cortex: Models of Ocular
    Receptive Fields and Maps in the Visual Cortex Mo dels of Ocular Dominance and Orientation Columns Kenneth D Miller Published in Mo dels of Neural Networks IIIEDomanyJLvan Hemmen and K Schulten Eds SpringerVerlag NY pp Anearlier and briefer version of this article appeared in The Handbook of Neural Networks MAArbib Ed The MIT Press under the ti tle Models of Ocular Dominance and Orientation Columns Reused by permission ABSTRACT The formation of o cular dominance and orientation columns in the mammalian visual cortex is briey reviewed Correlationbased mo d els for their development are then discussed b eginning with the mo dels of Von der Malsburg For the case of semilinear mo dels mo del b ehavior is well understo o d correlations determine receptive eld structure intracor tical interactions determine pro jective eld structure and the knitting together of the two determines the cortical map This provides a ba sis for simple but powerful mo dels of o cular dominance and orientation column formation o cular dominance columns form through a correlation based comp etition b etween lefteye and righteye inputs while orientation columns can form through a comp etition between ONcenter and OFF center inputs These mo dels accountwell for receptive eld structure but are not completely adequate to account for the details of cortical map struc ture Alternative approaches to map structure including the selforganizing feature map of Kohonen are discussed Finally theories of the computa tional function of correlationbased and selforganizing rules
    [Show full text]
  • Relation of Cortical Cell Orientation Selectivity to Alignment Of
    The Journal of Neuroscience. May 1991, 17(5): 1347-1359 Relation of Cortical Cell Orientation Selectivity to Alignment of Receptive Fields of the Geniculocortical Afferents that Arborize within a Single Orientation Column in Ferret Visual Cortex Barbara Chapman, Kathleen R. Zahs, and Michael P. Stryker Department of Physiology, University of California, San Francisco, California 94143-0444 Neurons in the primary visual cortex of higher mammals are and do not exhibit marked orientation preferences.When they arranged in columns, and the neurons in each column re- first describedcortical orientation specificity, Hubel and Wiesel spond best to light-dark borders of particular orientations. (1962) proposeda model to explain how cortical cells with sim- The basis of cortical cell orientation selectivity is not known. ple-type receptive fields, which predominate in the layers of One possible mechanism would be for cortical cells to re- carnivore cortex that receive direct geniculate input, could ac- ceive input from several lateral geniculate nucleus (LGN) quire orientation selectivity from their non-orientation-selec- neurons with receptive fields that are aligned in the visual tive inputs. According to this model, a given cell in the visual field (Hubel and Wiesel, 1962). We have investigated the cortex would receive excitatory inputs from geniculate cells, the relationship between the arrangement of the receptive fields receptive fields of which are aligned in the visual field. A stim- of geniculocortical afferents and the orientation preferences ulus falling along a line of the correct orientation to excite all of cortical cells in the orientation columns to which the af- of thesegeniculate cells simultaneouslywould produce the best ferents provide visual input.
    [Show full text]
  • DAVID H. HUBEL Harvard Medical School, Department of Neurobiology, Boston, Massachusetts, U.S.A
    EVOLUTION OF IDEAS ON THE PRIMARY VISUAL CORTEX, 1955-1978: A BIASED HISTORICAL ACCOUNT Nobel lecture, 8 December 1981 by DAVID H. HUBEL Harvard Medical School, Department of Neurobiology, Boston, Massachusetts, U.S.A. INTRODUCTION In the early spring of 1958 I drove over to Baltimore from Washington, D.C., and in a cafeteria at Johns Hopkins Hospital met Stephen Kuffler and Torsten Wiesel, for a discussion that was more momentous for Torsten’s and my future than either of us could have possibly imagined. I had been at Walter Reed Army Institute of Research for three years, in the Neuropsychiatry Section headed by David Rioch, working under the supervi- sion of M.G.F. Fuortes. I began at Walter Reed by developing a tungsten microelectrode and a technique for using it to record from chronically implant- ed cats, and I had been comparing the firing of cells in the visual pathways of sleeping and waking animals. It was time for a change in my research tactics. In sleeping cats only diffuse light could reach the retina through the closed eyelids. Whether the cat was asleep or awake with eyes open, diffuse light failed to stimulate the cells in the striate cortex. In waking animals I had succeeded in activating many cells with moving spots on a screen, and had found that some cells were very selective in that they responded when a spot moved in one direction across the screen (e.g. from left to right) but not when it moved in the opposite direction (1) (Fig. 1). There were many cells that I could not influence at all.
    [Show full text]