Wechsler Neural Networks for Perception

Human and Machine Perception
1. Auflage 2014
ISBN: 978-1-4832-6025-9
Verlag: Elsevier Science & Techn.
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 542 Seiten, Web PDF

Neural Networks for Perception, Volume 1: Human and Machine Perception focuses on models for understanding human perception in terms of distributed computation and examples of PDP models for machine perception. This book addresses both theoretical and practical issues related to the feasibility of both explaining human perception and implementing machine perception in terms of neural network models. The book is organized into two parts. The first part focuses on human perception. Topics on network model of object recognition in human vision, the self-organization of functional architecture in the cerebral cortex, and the structure and interpretation of neuronal codes in the visual system are detailed under this part. Part two covers the relevance of neural networks for machine perception. Subjects considered under this section include the multi-dimensional linear lattice for Fourier and Gabor transforms, multiple- scale Gaussian filtering, and edge detection; aspects of invariant pattern and object recognition; and neural network for motion processing. Neuroscientists, computer scientists, engineers, and researchers in artificial intelligence will find the book useful.

Autoren/Hrsg.

Wechsler, Harry

Inhaltsverzeichnis

1;Front Cover;1
2;Human and Machine Perception;4
3;Copyright Page;5
4;Table of Contents;8
5;Contents of Volume 2;12
6;Contributors;14
7;Foreword;18
8;PART I: Human Perception;24
8.1;I. Introduction;26
8.1.1;Chapter 1.1. Visual Cortex: Window on the Biological Basis of Learning and Memory;31
8.1.1.1;I. Introduction;31
8.1.1.2;II. Discussion;44
8.1.1.3;References;46
8.1.2;Chapter 1.2. A Network Model of Object Recognition in Human Vision;48
8.1.2.1;Abstract;48
8.1.2.2;1 Introduction;49
8.1.2.3;2 Shape-Based Recognition Performance in Human Vision;50
8.1.2.4;3 The CLF Model;54
8.1.2.5;4 Discussion;61
8.1.2.6;References;61
8.1.3;Chapter 1.3. A Cortically Based Model for Integration in Visual Perception;64
8.1.3.1;1. Introduction;64
8.1.3.2;II. The RCI Model;68
8.1.3.3;II.. Implications for Perception;82
8.1.3.4;Acknowledgements;83
8.1.3.5;References;84
8.1.4;CHAPTER 1.4. THE SYMMETRIC ORGANIZATION OF PARALLEL CORTICAL SYSTEMS FOR FORM AND MOTION PERCEPTION;87
8.1.4.1;1. Introduction: Why Do Parallel Cortical Systems Exist for the Perception of Static Form and Moving Form?;87
8.1.4.2;II. The Motion BCS and the Static BCS;88
8.1.4.3;III. Joining Sensitivity to Direction-of-Motion with Insensitivity to Direction-of-Contrast;90
8.1.4.4;IV. Why is not a Motion Form Perception System Sufficient?;94
8.1.4.5;V. A Symmetry Principle for Cortical Development: Sustained-Transient Gating, Opponent Processing, and Insensitivity to Direction-of-Contrast;94
8.1.4.6;VI. Different Geometries and Afterimages for Perception of Static Form and Motion Form;95
8.1.4.7;VII. Resonance versus Reset: Cooperative Feature Linking without Destructive Smearing;95
8.1.4.8;VIII. Combining Rapid Reset and Spatial Impenetrability Predictions;96
8.1.4.9;IX. Perception of Moving-Form-in-Depth: The V1.V2.MT Pathway;97
8.1.4.10;X. Apparent Motion of Illusory Figures;98
8.1.4.11;XI. Design of a MOC Filter;99
8.1.4.12;XII. Continuous Motion Paths from Spatially Stationary Flashes;105
8.1.4.13;..II. Feature Integration and Spatial Attention Shifts;110
8.1.4.14;XIV. Augmenting the Static BCS;111
8.1.4.15;XV. Design of Simple On-Cells and Off-Cells in the SOC Filter;113
8.1.4.16;.VI. FM Symmetry;114
8.1.4.17;XVII. 90° Orientations: From V1 to V2;115
8.1.4.18;XVIII. 180° Opponent Directions from V1 to MT;117
8.1.4.19;XIX. Opponent Rebounds: Rapid Reset of Resonating Segmentations;117
8.1.4.20;XX. MacKay Afterimages, the Waterfall Effect, and Long Range Motion Aftereffects;118
8.1.4.21;XXI. Concluding Remarks: FACADE Principles of Uncertainty, Complementarity, Summetry, and Resonance as a Foundation for Biological Vision Theories;120
8.1.4.22;References;122
8.1.5;Chapter 1.5. The Structure and Interpretation of Neuronal Codes in the Visual System;127
8.1.5.1;1. Defining the stimulus-response relation;128
8.1.5.2;2. Stimulus parameter interactions;132
8.1.5.3;3. The multiplex-filter hypothesis;134
8.1.5.4;4. Decoding multiplex messages;136
8.1.5.5;5. Feature selection;138
8.1.5.6;6. Feature linking;139
8.1.5.7;7. Interpreting Neuronal Codes;140
8.1.5.8;8. Conclusion;141
8.1.5.9;REFERENCES;141
8.1.6;Chapter 1.6. Self-Organization of Functional Architecture in The Cerebral Cortex;143
8.1.6.1;I. Introduction;143
8.1.6.2;II. Self-Organization of Synaptic Connections;144
8.1.6.3;III. Receptive Fields;150
8.1.6.4;IV. Self-Organization of Receptive Fields;152
8.1.6.5;V. Simple Application;156
8.1.6.6;VI. Summary;161
8.1.6.7;Appendix 1;162
8.1.6.8;Appendix 2;163
8.1.6.9;Appendix 3;165
8.1.6.10;References;166
8.1.7;Chapter 1.7. Filters Versus Textons in Human and Machine Texture Discrimination;168
8.1.7.1;I. Abstract;168
8.1.7.2;II. Introduction;169
8.1.7.3;III. Perceptual Asymmetries;174
8.1.7.4;IV. Conclusions;193
8.1.7.5;References;197
8.1.8;Chapter 1.8. Two-Dimensional Maps and Biological Vision: Representing Three-Dimensional Space;199
8.1.8.1;1. Introduction;199
8.1.8.2;2. Biological Vision;200
8.1.8.3;3. Computational Vision;203
8.1.8.4;4. Conclusion;212
8.1.8.5;References;213
9;PART II: Machine Perception;216
9.1;II. Introduction;218
9.1.1;II.1. WISARD and other Weightless Neurons;225
9.1.2;I. Introduction;225
9.1.3;II. The RAM Node;226
9.1.4;III. Generalizing RAM Nodes;228
9.1.5;IV. Large Neurons;229
9.1.6;V. Associative Networks: A Memorandum;235
9.1.7;References;235
9.2;Chapter II.2. Multi-Dimensional Linear Lattice for Fourier and Gabor Transforms, Multiple-Scale Gaussian Filtering, and Edge Detection;237
9.2.1;I. Introduction;237
9.2.2;II. Successive Convolution in terms of the Central Limit Theorem;241
9.2.3;III. One dimensional Gaussian Convolution;244
9.2.4;IV. Two-Dimensional Gaussian Averaging;246
9.2.5;V. Three and Higher Dimensional Averaging;248
9.2.6;VI. Higher order Webs;248
9.2.7;VII. Multi-Scale Edge masks;249
9.2.8;VIII. Deriving the Gaussian-Smoothed-Laplacian;250
9.2.9;IX. Generating the Discrete Fourier Transform;251
9.2.10;X. Gabor Transforms;252
9.2.11;XI. Simulation Results;253
9.2.12;XII. Discussion;255
9.2.13;References;255
9.3;CHAPTER II.3. ASPECTS OF INVARIANT PATTERN AND OBJECT RECOGNITION;257
9.3.1;1. Introduction;257
9.3.2;II. Pattern Recognition;259
9.3.3;III. Brief Notes in Object Recognition;268
9.3.4;IV. Conclusion;269
9.3.5;REFERENCES;270
9.4;CHAPTER II.4. A NEURAL NETWORK ARCHITECTURE FOR FAST ON-LINE SUPERVISED LEARNING AND PATTERN RECOGNITION;271
9.4.1;I. Introduction;271
9.4.2;II. The ARTMAP System;273
9.4.3;III. ARTMAP Simulations: Distinguishing Edible and Poisonous Mushrooms;274
9.4.4;IV. Conclusion: Predictive ART;282
9.4.5;Acknowledgements;286
9.4.6;References;286
9.5;Chapter II.5. Neural Network Approaches to Color Vision;288
9.5.1;I. Introduction;288
9.5.2;II. Learning Lightness Algorithms;289
9.5.3;III. Finding Material Boundaries: The Use of Color in Image Segmentation;298
9.5.4;References;306
9.6;Chapter II.6. Adaptive Sensory-Motor Coordination Through Self-Consistency;308
9.6.1;1. Introduction;308
9.6.2;II. Locating Stationary Targets;311
9.6.3;III. Reaching stationary targets with multiple joints;323
9.6.4;IV. Dynamic movement control for unforeseen payloads;328
9.6.5;V. Summary;335
9.6.6;Appendix;336
9.6.7;References;336
9.7;Chapter II.7. Finding Boundaries in Images;338
9.7.1;I. Introduction;338
9.7.2;II. Local analysis of image patches by filtering;343
9.7.3;III. Texture Discrimination;348
9.7.4;IV· Brightness boundaries;355
9.7.5;Acknowledgements;360
9.7.6;References;363
9.8;Chapter II.8. Compression of Remotely Sensed Images using Self Organizing Feature Maps;368
9.8.1;I. Introduction;368
9.8.2;II. Self Organizing Feature Map (SOFM);372
9.8.3;III. MasPar Implementation of Kohonen's SOFM;375
9.8.4;IV. Results and Discussion;380
9.8.5;Acknowledgments;382
9.8.6;References;382
9.9;Chapter II.9. Self - Organizing Maps and Computer Vision;391
9.9.1;1. Introduction;391
9.9.2;2.The Self-Organizing Feature Map and Feature Detection;395
9.9.3;3. Texture segmentation by the SOFM;399
9.9.4;4. Extracting global curve features by the SOFM;403
9.9.5;5. Conclusions;406
9.9.6;References;407
9.10;Chapter II.10. Region Growing Using Neural Networks;409
9.10.1;I. Introduction;409
9.10.2;II. Region Growing Using Neural Networks;411
9.10.3;III. Experimental Results;412
9.10.4;IV. Discussion;417
9.10.5;V. Summary;419
9.10.6;References;419
9.11;Chapter II.11. Vision and Space-Variant Sensing;421
9.11.1;Abstract;421
9.11.2;1. Introduction;422
9.11.3;2. Fixation and space-variant sensing;424
9.11.4;3. Conclusion;441
9.11.5;4. References;445
9.12;Chapter II.12. Learning and Recognizing 3D Objects from Multiple Views in a Neural System;449
9.12.1;I Introduction;449
9.12.2;II Background;450
9.12.3;Ill Neural System Architecture;451
9.12.4;IV System Neurodynamics;455
9.12.5;V Demonstration;462
9.12.6;VI Conclusion;465
9.12.7;References;466
9.13;Chapter II.13. Hybrid Symbolic-Neural Methods for Improved Recognition Using High-Level Visual Features;468
9.13.1;I. Introduction;468
9.13.2;II. The KBANN Algorithm;470
9.13.3;III. Empirical Tests of KBANN-DAID;475
9.13.4;IV. Discussion;481
9.13.5;V. Related Work;482
9.13.6;VI. Conclusions;483
9.13.7;Acknowledgement;483
9.13.8;References;483
9.14;Chapter II.14. Multiscale and Distributed Visual Representations and Mappings for Invariant Low-Level Perception;485
9.14.1;I. Introduction;485
9.14.2;II. Multiscale and Distributed Image Representations;488
9.14.3;III. Invariance;490
9.14.4;IV. Mappings;493
9.14.5;V. Conclusions;495
9.14.6;References;496
9.15;Chapter II.15. Symmetry: a context free Cue for Foveated Vision;500
9.15.1;1 Introduction;500
9.15.2;2 Defining tlie Operator;502
9.15.3;3 Operation on Natural scenes;505
9.15.4;4 Psychophysical Correlates;508
9.15.5;5 Implementation by Neural Networks;510
9.15.6;6 Conclusion;512
9.15.7;References;513
9.16;Chapter II.16. A Neural Network for Motion Processing;515
9.16.1;Abstract;515
9.16.2;1. Introduction;516
9.16.3;2. A Neural Network;517
9.16.4;3. Computing Optical Flow;521
9.16.5;4. Recovering Depth;531
9.16.6;5. Discussion;536
9.16.7;References;538
10;Index;540