Computational Maps in the Visual Cortex
     Table of Contents
MiikkulainenBednarChoeSirosh
Home    
About the Authors
Back Cover    
Table of Contents 
Sample Chapter 
Figures    
References    
Errata    
Demos     
Talks/Courses 
Software    
Credits    
Purchase online at:

springeronline.com
Foreword--vii
Preface--xi
List of Figures--xxv
List of Tables--xxxi

Part I     FOUNDATIONS

     1     Introduction -- 3
          1.1     Input-Driven Self-Organization -- 4
          1.2     Constructing Visual Function -- 7
          1.3     Perceptual Grouping -- 8
          1.4     Approach -- 10
          1.5     Guide for the Reader -- 13
     2     Biological Background -- 15
          2.1     Visual System Organization -- 15
          2.2     Lateral Connections -- 23
          2.3     Genetic Versus Environmental Factors in Development -- 28
          2.4     Temporal Coding -- 33
          2.5     Conclusion -- 36
     3     Computational Foundations -- 39
          3.1     Computational Units -- 39
          3.2     Temporal Coding -- 46
          3.3     Adaptation -- 48
          3.4     Self-Organizing Maps -- 50
          3.5     Knowledge Representation in Maps -- 57
          3.6     Conclusion -- 63

Part II     INPUT-DRIVEN SELF-ORGANIZATION

     4     LISSOM: A Computational Map Model of V1 -- 67
          4.1     Motivation: Cortical Maps -- 67
          4.2     The LISSOM Architecture -- 68
          4.3     Response Generation -- 73
          4.4     Learning -- 76
          4.5     Self-Organizing Process -- 78
          4.6     Conclusion -- 82
     5     Development of Maps and Connections -- 85
          5.1     Biological Background -- 85
          5.2     Computational Models -- 91
          5.3     Orientation Maps -- 95
          5.4     Ocular Dominance Maps -- 106
          5.5     Direction Selectivity Maps -- 113
          5.6     Combined Maps of Multiple Features -- 119
          5.7     Discussion -- 130
          5.8     Conclusion -- 132
     6     Understanding Plasticity -- 133
          6.1     Biological and Computational Background -- 133
          6.2     The Reduced LISSOM Model -- 138
          6.3     Retinal Lesions -- 142
          6.4     Cortical Lesions -- 146
          6.5     Discussion -- 152
          6.6     Conclusion -- 153
     7     Understanding Visual Performance: The Tilt Aftereffect -- 155
          7.1     Psychophysical and Computational Background -- 155
          7.2     Method -- 160
          7.3     Results -- 164
          7.4     Analysis -- 166
          7.5     Discussion -- 170
          7.6     Conclusion -- 171

Part III     CONSTRUCTING VISUAL FUNCTION

     8     HLISSOM: A Hierarchical Model -- 175
          8.1     Motivation: Synergy of Nature and Nurture -- 175
          8.2     The Hierarchical Architecture -- 178
          8.3     Inputs, Activation and Learning -- 181
          8.4     Effect of Input Sequence and Initial Organization -- 184
          8.5     Conclusion -- 185
     9     Understanding Low-Level Development: Orientation Maps -- 189
          9.1     Biological Motivation -- 189
          9.2     Prenatal Development -- 190
          9.3     Postnatal Development -- 194
          9.4     Prenatal and Postnatal Contributions -- 199
          9.5     Discussion -- 201
          9.6     Conclusion -- 202
     10     Understanding High-Level Development: Face Detection -- 203
          10.1     Psychophysical and Computational Background -- 203
          10.2     Prenatal Development -- 213
          10.3     Postnatal Development -- 227
          10.4     Discussion -- 233
          10.5     Conclusion -- 238

Part IV     PERCEPTUAL GROUPING

     11     PGLISSOM: A Perceptual Grouping Model -- 241
          11.1     Motivation: Temporal Coding -- 241
          11.2     The Self-Organization and Grouping Architecture -- 243
          11.3     Spiking Unit Model -- 244
          11.4     Learning -- 247
          11.5     Self-Organizing Process -- 249
          11.6     Conclusion -- 255
     12     Temporal Coding -- 257
          12.1     Method -- 257
          12.2     Binding Through Synchronization -- 258
          12.3     Segmentation Through Desynchronization -- 263
          12.4     Robustness Against Variation and Noise -- 265
          12.5     Discussion -- 270
          12.6     Conclusion -- 271
     13     Understanding Perceptual Grouping: Contour Integration -- 273
          13.1     Psychophysical and Computational Background -- 273
          13.2     Contour Integration and Segmentation -- 280
          13.3     Contour Completion and Illusory Contours -- 288
          13.4     Influence of Input Distribution on Anatomy and Performance -- 295
          13.5     Discussion -- 299
          13.6     Conclusion -- 304

Part V     EVALUATION AND FUTURE DIRECTIONS

     14     Computations in Visual Maps -- 307
          14.1     Visual Coding in the Cortex -- 307
          14.2     Visual Coding in LISSOM -- 309
          14.3     Visual Coding for High-Level Tasks -- 314
          14.4     Discussion -- 319
          14.5     Conclusion -- 324
     15     Scaling LISSOM simulations -- 325
          15.1     Parameter Scaling Approach -- 325
          15.2     Scaling Equations -- 326
          15.3     Forming Large Maps: The GLISSOM Approach -- 332
          15.4     GLISSOM Scaling -- 332
          15.5     Scaling to Cortical Dimensions -- 339
          15.6     Discussion -- 342
          15.7     Conclusion -- 343
     16     Discussion: Biological Assumptions and Predictions -- 345
          16.1     Self-Organization -- 345
          16.2     Genetically Driven Development -- 354
          16.3     Temporal Coding -- 358
          16.4     Predictions -- 362
          16.5     Conclusion -- 373
     17     Future Work: Computational Directions -- 375
          17.1     Extensions to the LISSOM Mechanisms -- 375
          17.2     Modeling New Phenomena with LISSOM -- 379
          17.3     New Research Directions -- 396
          17.4     The Topographica Cortical Map Simulator -- 403
          17.5     Conclusion -- 406
     18     Conclusion -- 409
          18.1     Contributions -- 409
          18.2     Conclusion -- 412

Appendices    

     A     LISSOM Simulation Specifications -- 415
          A.1     Generalized Activation Equation -- 415
          A.2     Default Parameters -- 416
          A.3     Choosing Parameters for New Simulations -- 419
          A.4     Retinotopic Maps -- 422
          A.5     Orientation Maps -- 422
          A.6     Ocular Dominance Maps -- 423
          A.7     Direction Maps -- 424
          A.8     Combined Orientation / Ocular Dominance Maps -- 424
          A.9     Combined Orientation / Ocular Dominance / Direction Maps -- 424
     B     Reduced LISSOM Simulation Specifications -- 427
          B.1     Plasticity -- 427
          B.2     Tilt Aftereffect -- 428
          B.3     Scaling -- 428
     C     HLISSOM Simulation Specifications -- 429
          C.1     V1 Only -- 429
          C.2     Face-Selective Area Only -- 430
          C.3     Combined V1 and Face-Selective Area -- 432
     D     PGLISSOM Simulation Specifications -- 435
          D.1     Self-Organization -- 435
          D.2     Grouping -- 437
          D.3     Synchronization -- 438
     E     SOM Simulation Specifications -- 439
     F     Visual Coding Simulation Specifications -- 441
          F.1     Sparse Coding and Reconstruction -- 441
          F.2     Handwritten Digit Recognition -- 442
     G     Calculating Feature Maps -- 445
          G.1     Preference Map Algorithms -- 445
          G.2     Retinotopic Maps -- 449
          G.3     Orientation Maps -- 449
          G.4     Ocular Dominance Maps -- 449
          G.5     Direction Maps -- 450
          G.6     Orientation Gradients -- 450
References--451
Author Index--503
Subject Index--523