Preface......Page 5
Contents......Page 9
Acronyms......Page 13
1 Brief Introduction......Page 17
2 Basic Types......Page 19
2.1 Stacked Autoencoders (SAEs)......Page 20
2.2 Deep Belief Networks (DBNs)......Page 21
2.3 Convolutional Neural Networks (CNNs)......Page 22
2.4 Recurrent Neural Networks (RNNs)......Page 24
2.5 Generative Adversarial Nets (GANs)......Page 25
3.1 Audio Data......Page 27
3.2 Image Data......Page 28
3.3 Text Data......Page 29
4.1 Theory Challenges......Page 30
5 Conclusions......Page 31
References......Page 32
1 Introduction......Page 35
2.1 Two-Stage Methods for Deep Object Detection......Page 37
2.2 One-Stage Methods for Deep Object Detection......Page 46
3.1 Handcrafted Feature-Based Methods for Pedestrian Detection......Page 50
3.2 CNN-Based Methods for Pedestrian Detection......Page 56
4.1 Scale Variation Problem......Page 62
4.2 Occlusion Problem......Page 68
4.3 Deformation Problem......Page 69
References......Page 70
1 Introduction......Page 74
2.1.1 Engineering Designed Features......Page 76
2.1.2 Learning-Based Features......Page 79
2.2.1 2D-Based Synthesis Methods......Page 87
3.1 Image Processing-Based Methods......Page 89
3.2 Invariant Feature-Based Methods......Page 91
3.3 Illumination Model-Based Method......Page 93
4.1 Root Convolutional Layer......Page 96
4.2 Multi-hierarchical Local Feature......Page 98
5.1 Dataset......Page 99
5.2 Recognition Across Poses and Illumination......Page 100
6 Conclusion......Page 101
References......Page 102
1 Introduction......Page 106
2 Related Work......Page 108
3 Proposed Method......Page 110
3.3 Local Binary Pattern (LBP)......Page 111
3.4 Concatenating the LBP......Page 113
4.1.1 Replay-Attack......Page 114
4.1.2 CASIA-FA......Page 115
4.2.2 Data Processing......Page 116
5.1.1 Intra Test......Page 117
5.2.1 Intra Test......Page 120
5.3 Comparison Against State-of-the-Art Methods......Page 121
6 Conclusion......Page 123
References......Page 125
1 Introduction......Page 127
2 Related Work......Page 128
2.1 Methods Based on Features......Page 129
2.3 Other Methods......Page 130
3.1 Methodology......Page 131
3.2 Experimental Analysis......Page 135
4 Video-Based Kinship Verification......Page 136
4.1 Methodology......Page 137
4.2 Experimental Analysis......Page 140
References......Page 144
Deep Learning Architectures for Face Recognition in VideoSurveillance......Page 147
1 Introduction......Page 148
2 Background of Video-Based FR Through Deep Learning......Page 149
3.1.1 Cross-Correlation Matching CNN......Page 151
3.1.2 Trunk-Branch Ensemble CNN......Page 155
3.1.3 HaarNet......Page 156
3.2 Deep CNNs Using Autoencoder......Page 161
4 Performance Evaluation......Page 164
5 Conclusion and Future Directions......Page 165
References......Page 166
Deep Learning for 3D Data Processing......Page 169
1 Introduction......Page 170
2.2 Deep Learning for 3D Shapes With Raw Features......Page 172
2.3 Restricted Boltzmann Machines (RBM) and Deep Belief Network (DBN)......Page 173
2.5 Details of CRBM and CDBN......Page 174
3.1 Circle Convolution......Page 176
3.3 Example of Circle Convolution and PDD Computation......Page 180
3.4 Elimination of the Initial Location Ambiguity......Page 181
3.5 The Structure of CCRBM......Page 183
3.6 Circle Convolutional DBN (CCDBN)......Page 186
4.1 Global Shape Retrieval......Page 187
4.3 Shape Correspondence......Page 188
4.4 The Setup of Parameters for CCRBM......Page 189
5.1 Global Shape Retrieval......Page 193
5.2 Partial Shape Retrieval......Page 194
5.3 Shape Correspondence......Page 195
5.4 Significance and Analysis......Page 196
6 Conclusion......Page 197
References......Page 198
Deep Learning-Based Descriptors for Object Instance Search......Page 202
1 Introduction......Page 203
2 Related Work......Page 204
3 Compact Invariant Deep Descriptors......Page 207
3.1.1 Method......Page 208
3.1.2 Evaluation Framework......Page 212
3.1.3 Experimental Results......Page 213
3.2 Dual-Margin Siamese Fine-Tuning......Page 215
3.2.1 Method......Page 217
3.2.3 Experimental Results......Page 219
3.3.1 I-Theory in a Nutshell......Page 221
3.3.3 Multigroup-Invariant CNN Descriptors......Page 222
3.3.4 Evaluation Framework......Page 225
3.3.5 Experimental Results......Page 226
3.4 Hashing with Invariant Descriptors......Page 229
4 Conclusions and Future Works......Page 232
References......Page 233