This book presents a systematic study of visual pattern discovery, from unsupervised to semi-supervised manner approaches, and from dealing with a single feature to multiple types of features. Furthermore, it discusses the potential applications of discovering visual patterns for visual data analytics, including visual search, object and scene recognition.
It is intended as a reference book for advanced undergraduates or postgraduate students who are interested in visual data analytics, enabling them to quickly access the research world and acquire a systematic methodology rather than a few isolated techniques to analyze visual data with large variations. It is also inspiring for researchers working in computer vision and pattern recognition fields. Basic knowledge of linear algebra, computer vision and pattern recognition would be helpful to readers.
|
|
|
1 | (14) |
|
|
|
1 | (2) |
|
1.2 Discovering Spatial Co-occurrence Patterns |
|
|
3 | (2) |
|
1.3 Discovering Feature Co-occurrence Patterns |
|
|
5 | (2) |
|
|
|
7 | (8) |
|
|
|
8 | (7) |
|
2 Context-Aware Discovery of Visual Co-occurrence Patterns |
|
|
15 | (14) |
|
|
|
15 | (1) |
|
2.2 Multi-context-aware Clustering |
|
|
16 | (5) |
|
2.2.1 Regularized k-means Formulation with Multiple Contexts |
|
|
16 | (3) |
|
2.2.2 Self-learning Optimization |
|
|
19 | (2) |
|
|
|
21 | (6) |
|
2.3.1 Spatial Visual Pattern Discovery |
|
|
21 | (2) |
|
2.3.2 Image Region Clustering Using Multiple Contexts |
|
|
23 | (4) |
|
2.4 Summary of this
Chapter |
|
|
27 | (2) |
|
|
|
28 | (1) |
|
3 Hierarchical Sparse Coding for Visual Co-occurrence Discovery |
|
|
29 | (16) |
|
|
|
29 | (1) |
|
3.2 Spatial Context-Aware Multi-feature Sparse Coding |
|
|
30 | (6) |
|
3.2.1 Learning Spatial Context-Aware Visual Phrases |
|
|
30 | (5) |
|
3.2.2 Learning Multi-feature Fused Visual Phrases |
|
|
35 | (1) |
|
|
|
36 | (7) |
|
3.3.1 Spatial Visual Pattern Discovery |
|
|
36 | (2) |
|
|
|
38 | (2) |
|
3.3.3 Scene Categorization |
|
|
40 | (3) |
|
3.4 Summary of this
Chapter |
|
|
43 | (2) |
|
|
|
43 | (2) |
|
4 Feature Co-occurrence for Visual Labeling |
|
|
45 | (22) |
|
|
|
45 | (2) |
|
4.2 Multi-feature Collaboration for Transductive Learning |
|
|
47 | (7) |
|
4.2.1 Spectral Embedding of Multi-feature Data |
|
|
48 | (1) |
|
4.2.2 Embedding Co-occurrence for Data Representation |
|
|
49 | (1) |
|
4.2.3 Transductive Learning with Feature Co-occurrence Patterns |
|
|
50 | (2) |
|
4.2.4 Collaboration Between Pattern Discovery and Label Propagation |
|
|
52 | (2) |
|
|
|
54 | (10) |
|
4.3.1 Experimental Setting |
|
|
54 | (1) |
|
4.3.2 Label Propagation on Synthetic Data |
|
|
54 | (2) |
|
|
|
56 | (2) |
|
|
|
58 | (1) |
|
4.3.5 Body Motion Recognition |
|
|
59 | (3) |
|
|
|
62 | (2) |
|
4.4 Summary of this
Chapter |
|
|
64 | (3) |
|
|
|
64 | (3) |
|
5 Visual Clustering with Minimax Feature Fusion |
|
|
67 | (18) |
|
|
|
67 | (2) |
|
5.2 Minimax Optimization for Multi-feature Spectral Clustering |
|
|
69 | (5) |
|
5.2.1 Spectral Embedding for Regularized Data-Cluster Similarity Matrix |
|
|
69 | (1) |
|
|
|
69 | (2) |
|
5.2.3 Minimax Optimization |
|
|
71 | (3) |
|
|
|
74 | (7) |
|
5.3.1 Datasets and Experimental Setting |
|
|
74 | (1) |
|
5.3.2 Baseline Algorithms |
|
|
74 | (2) |
|
|
|
76 | (1) |
|
5.3.4 Experimental Results |
|
|
76 | (3) |
|
5.3.5 Convergence Analysis |
|
|
79 | (1) |
|
5.3.6 Sensitivity of Parameters |
|
|
79 | (2) |
|
5.4 Summary of this
Chapter |
|
|
81 | (4) |
|
|
|
82 | (3) |
|
|
|
85 | |
|
|
|
86 | |
Hongxing Wang received his B.S. and M.S. degrees from Chongqing University, China, and his Ph.D. degree from Nanyang Technological University, Singapore. He is currently a faculty member at the School of Software Engineering, Chongqing University. Before joining Chongqing University, he worked as a research fellow/associate at the School of Electrical and Electronic Engineering (EEE) at Nanyang Technological University, and as a visiting student at The Institute of Scientific and Industrial Research (ISIR), Osaka University, Japan. His research interests include computer vision, pattern recognition, and machine learning.
Chaoqun Weng received his B.E. degree in Computer Science and Technology from Nankai University, China, in 2010. He is currently pursuing his Ph.D. degree at the School of Electrical & Electronics Engineering, Nanyang Technological University, Singapore. His research interests include computer vision and machine learning.
JunsongYuan received his Ph.D. from Northwestern University and M.Eng. from the National University of Singapore. Before that, he graduated from the Special Class for the Gifted Young of Huazhong University of Science and Technology in China. He is currently an associate professor and program director of video analytics at the School of Electrical and Electronics Engineering, Nanyang Technological University (NTU). He serves as guest editor of International Journal of Computer Vision (IJCV), and is currently an associate editor of IEEE Trans. on Image Processing (T-IP), IEEE Trans. on Circuits and Systems for Video Technology (T-CSVT) and The Visual Computer journal (TVC). He also serves as area chair of various conferences including CVPR/ACCV/WACV/ICPR/ICME. He received Best Paper Award from IEEE Trans. on Multimedia, the Doctoral Spotlight Award from IEEE Conf. on Computer Vision and Pattern Recognition (CVPR'09), a Nanyang Assistant Professorship from NTU, and Outstanding EECS Ph.D. Thesis award from Northwestern University.
Lisainfo e-raamatute kohta