From the foreword by Thomas Huang: "During the past decade, researchers in computer vision have found that probabilistic machine learning methods are extremely powerful. This book describes some of these methods. In addition to the Maximum Likelihood framework, Bayesian Networks, and Hidden Markov models are also used. Three aspects are stressed: features, similarity metric, and models. Many interesting and important new results, based on research by the authors and their collaborators, are presented. Although this book contains many new results, it is written in a style that suits both experts and novices in computer vision."
Author(s): N. Sebe, M.S. Lew
Series: Computational Imaging and Vision
Edition: 1
Publisher: Springer
Year: 2003
Language: English
Pages: 234
Foreword......Page 12
Preface......Page 14
1. INTRODUCTION......Page 18
1 Visual Similarity......Page 19
1.1 Color......Page 21
1.2 Texture......Page 24
1.3 Shape......Page 26
1.4 Stereo......Page 28
1.6 Facial expression......Page 30
1.7 Summary......Page 32
2 Evaluation of Computer Vision Algorithms......Page 33
3 Overview of the Book......Page 36
1 Introduction......Page 42
2 Statistical Distributions......Page 43
2.1 Gaussian Distribution......Page 44
2.2 Exponential Distribution......Page 55
2.3 Cauchy Distribution......Page 58
3 Robust Statistics......Page 60
3.1 Outliers......Page 61
4 Maximum Likelihood Estimators......Page 62
5 Maximum Likelihood in Relation to Other Approaches......Page 64
6 Our Maximum Likelihood Approach......Page 67
6.1 Scale Parameter Estimation in a Cauchy Distribution......Page 71
7 Experimental Setup......Page 74
8 Concluding Remarks......Page 76
1 Introduction......Page 78
3 Color Models......Page 81
3.1 RGB Color System......Page 82
3.2 HSV Color System......Page 83
3.3 l1l2l3 Color System......Page 84
4 Color Based Retrieval......Page 85
4.1 Color Indexing......Page 86
5.1 Early Experiments......Page 90
5.2 Usability Issues......Page 91
5.3 Printer-Scanner Noise Experiments......Page 92
5.5 Quantization......Page 93
5.6 Distribution Analysis......Page 94
6 Experiments with the Objects Database......Page 96
7 Concluding Remarks......Page 98
1 Introduction......Page 100
2 Human Perception of Texture......Page 103
3 Texture Features......Page 104
3.1 Texture Distribution Models......Page 105
3.1.3 Center-symmetric covariance measures......Page 106
3.1.5 Complementary feature pairs......Page 108
3.2 Gabor and Wavelet Models......Page 109
4 Texture Classification Experiments......Page 112
4.2 Distribution Analysis......Page 114
4.3 Misdetection Rates......Page 116
5 Texture Retrieval Experiments......Page 121
5.1 Texture Features......Page 122
5.3 Similarity Noise for QMF-Wavelet Transform......Page 123
5.4 Similarity Noise for Gabor Wavelet Transform......Page 125
6 Concluding Remarks......Page 126
1 Introduction......Page 128
2 Human Perception of Visual Form......Page 130
3 Active Contours......Page 135
3.1 Behavior of Traditional Active Contours......Page 137
3.2 Generalized Force Balance Equations......Page 141
3.3 Gradient Vector Flow......Page 142
4 Invariant Moments......Page 147
5 Experiments......Page 148
6 Conclusions......Page 151
1 Introduction......Page 152
1.1 Stereoscopic Vision......Page 154
2 Stereo Matching......Page 155
2.1 Related Work......Page 159
3.1 Template Based Algorithm......Page 161
3.2 Multiple Windows Algorithm......Page 163
3.3 Cox’ Maximum Likelihood Algorithm......Page 164
4 Stereo Matching Experiments......Page 167
4.2 Stereo Matching Results......Page 168
5 Motion Tracking Experiments......Page 174
6 Concluding Remarks......Page 177
7. FACIAL EXPRESSION RECOGNITION......Page 180
1 Introduction......Page 181
2 Emotion Recognition......Page 183
2.2 Review of Facial Expression Recognition......Page 184
3 Face Tracking and Feature Extraction......Page 188
4 The Static Approach: Bayesian Network Classifiers......Page 190
4.1 Continuous Naive-Bayes: Gaussian and Cauchy Naive Bayes Classifiers......Page 192
4.2 Beyond the Naive-Bayes Assumption: Finding Dependencies among Features Using a Gaussian TAN Classifier......Page 193
5 The Dynamic Approach: Expression Recognition Using Multi-level HMMs......Page 196
5.1 Hidden Markov Models......Page 199
5.2 Expression Recognition Using Emotion-Specific HMMs......Page 200
5.3 Automatic Segmentation and Recognition of Emotions Using Multi-level HMM.......Page 201
6 Experiments......Page 204
6.1.1 Person-Dependent Tests......Page 208
6.1.2 Person-Independent Tests......Page 210
6.2 Results Using the Cohn-Kanade Database......Page 211
7 Summary and Discussion......Page 212
References......Page 216
Index......Page 228
