Online Visual Tracking

This book presents the state of the art in online visual tracking, including the motivations, practical algorithms, and experimental evaluations. Visual tracking remains a highly active area of research in Computer Vision and the performance under complex scenarios has substantially improved, driven by the high demand in connection with real-world applications and the recent advances in machine learning. A large variety of new algorithms have been proposed in the literature over the last two decades, with mixed success. Chapters 1 to 6 introduce readers to tracking methods based on online learning algorithms, including sparse representation, dictionary learning, hashing codes, local model, and model fusion. In Chapter 7, visual tracking is formulated as a foreground/background segmentation problem, and tracking methods based on superpixels and end-to-end deep networks are presented. In turn, Chapters 8 and 9 introduce the cutting-edge tracking methods based on correlation filter and deep learning. Chapter 10 summarizes the book and points out potential future research directions for visual tracking. The book is self-contained and suited for all researchers, professionals and postgraduate students working in the fields of computer vision, pattern recognition, and machine learning. It will help these readers grasp the insights provided by cutting-edge research, and benefit from the practical techniques available for designing effective visual tracking algorithms. Further, the source codes or results of most algorithms in the book are provided at an accompanying website. Preface 5 Acknowledgements 6 Contents 7 1 Introduction to Visual Tracking 9 1.1 Basic Components of Visual Tracking Algorithms 9 1.2 Challenges in Visual Tracking 12 1.3 Tracking Datasets 12 1.4 Evaluation Metrics for Visual Tracking 15 1.5 Outline of This Book 16 References 16 2 Visual Tracking Based on Sparse Representation 19 2.1 Introduction 19 2.2 Sparse Representation in Occlusion Handling 20 2.3 Sparse Collaborative Appearance Model 26 2.4 Experimental Results and Discussions 30 2.5 Summary 32 References 32 3 Visual Tracking Based on Local Model 34 3.1 Introduction 34 3.2 Visual Tracking Based on Local Similarity 35 3.3 Visual Tracking Based on Local Sparse Representation 41 3.4 Experimental Results and Discussions 46 3.5 Summary 48 References 48 4 Visual Tracking Based on Model Fusion 50 4.1 Introduction 50 4.2 Visual Tracking Based on Adaptive Multi-feature Combination 51 4.3 Visual Tracking Based on Multi-attention Module 55 4.4 Experimental Results and Discussions 64 4.5 Summary 66 References 66 5 Tracking by Segmentation 68 5.1 Introduction 68 5.2 Color-Based Object Tracking Based on Boosted Color Distribution 70 5.3 Visual Tracking Based on Superpixel Segmentation 77 5.4 Tracking via End-to-End Fully Convolutional Networks (FCNs) 82 5.5 Experimental Results and Discussions 90 5.6 Summary 91 References 91 6 Correlation Tracking 93 6.1 Introduction 93 6.2 Correlation Tracking via Joint Discrimination and Reliability Learning 95 6.3 Experimental Results and Discussions 102 6.4 Summary 105 References 105 7 Visual Tracking Based on Deep Learning 107 7.1 Introduction 107 7.2 Visual Tracking Based on Convolutional Networks 109 7.3 Visual Tracking Based on Structured Siamese Network 118 7.4 Visual Tracking Based on Deep Reinforcement Learning 122 7.5 Experimental Results and Discussions 129 7.6 Summary 130 References 131 8 Conclusions and Future Work 133 8.1 Summary 133 8.2 Future Work 133 Front Matter ....Pages i-x Introduction to Visual Tracking (Huchuan Lu, Dong Wang)....Pages 1-10 Visual Tracking Based on Sparse Representation (Huchuan Lu, Dong Wang)....Pages 11-25 Visual Tracking Based on Local Model (Huchuan Lu, Dong Wang)....Pages 27-42 Visual Tracking Based on Model Fusion (Huchuan Lu, Dong Wang)....Pages 43-60 Tracking by Segmentation (Huchuan Lu, Dong Wang)....Pages 61-85 Correlation Tracking (Huchuan Lu, Dong Wang)....Pages 87-100 Visual Tracking Based on Deep Learning (Huchuan Lu, Dong Wang)....Pages 101-126 Conclusions and Future Work (Huchuan Lu, Dong Wang)....Pages 127-128