Thomas S. Huang - Böcker

Deep Learning through Sparse Representation and Low-Rank Modeling bridges classical sparse and low rank models-those that emphasize problem-specific Interpretability-with recent deep network models that have enabled a larger learning capacity and better utilization of Big Data. It shows how the toolkit of deep learning is closely tied with the sparse/low rank methods and algorithms, providing a rich variety of theoretical and analytic tools to guide the design and interpretation of deep learning models. The development of the theory and models is supported by a wide variety of applications in computer vision, machine learning, signal processing, and data mining.

This book will be highly useful for researchers, graduate students and practitioners working in the fields of computer vision, machine learning, signal processing, optimization and statistics.

Combines classical sparse and low-rank models and algorithms with the latest advances in deep learning networks Shows how the structure and algorithms of sparse and low-rank methods improves the performance and interpretability of Deep Learning models Provides tactics on how to build and apply customized deep learning models for various applications

200Ts Vision of Vision One of my formative childhood experiences was in 1968 stepping into the Uptown Theater on Connecticut Avenue in Washington, DC, one of the few movie theaters nationwide that projected in large-screen cinerama. I was there at the urging of a friend, who said I simply must see the remarkable film whose run had started the previous week. "You won't understand it," he said, "but that doesn't matter. " All I knew was that the film was about science fiction and had great special eflPects. So I sat in the front row of the balcony, munched my popcorn, sat back, and experienced what was widely touted as "the ultimate trip:" 2001: A Space Odyssey. My friend was right: I didn't understand it. . . but in some senses that didn't matter. (Even today, after seeing the film 40 times, I continue to discover its many subtle secrets. ) I just had the sense that I had experienced a creation of the highest aesthetic order: unique, fresh, awe inspiring. Here was a film so distinctive that the first half hour had no words whatsoever; the last half hour had no words either; and nearly all the words in between were banal and irrelevant to the plot - quips about security through Voiceprint identification, how to make a phonecall from a space station, government pension plans, and so on.

"Exploration of Visual Data" presents research efforts in the area of content-based exploration of image and video data. The main objective is to bridge the semantic gap between high-level concepts in the human mind and low-level features extractable by the machines. The two key issues emphasized are "content-awareness" and "user-in-the-loop". The authors provide a comprehensive review on algorithms for visual feature extraction based on colour, texture, shape and structure, and techniques for incorporating such information to aid browsing, exploration, search and streaming of image and video data. They also discuss issues related to the mixed use of textual and low-level visual features to facilitate more effective access of multimedia data. To bridge the semantic gap, significant recent research efforts have also been put on learning during user interactions, which is also known as "relevance feedback". The difficulty and challenge also come from the personalized information need of each user and a small amount of feedbacks the machine could obtain through real-time user interaction.The authors present and discuss several recently proposed classification and learning techniques that are specifically designed for this problem, with kernel- and boosting-based approaches for nonlinear extensions. "Exploration of Visual Data" provides state-of-the-art materials on the topics of content-based description of visual data, content-based low-bitrate video streaming, and latest asymmetric and nonlinear relevance feedback algorithms, which to date are unpublished.

This monograph is an essential reference for both academic and professional researchers working in the fields of computer vision, surveillance, HCI, image processing, and modeling. The human face provides important visual cues for human-human communication and human-computer interaction (HCI). With the advances in new information techniques and media, 3D face processing techniques will help an intelligent HCI system interpret and deliver more natural multimedia information related to human faces. 3D Face Processing: Modeling, Analysis and Synthesis introduces the frontiers of 3D face processing techniques. It reviews existing 3D face processing techniques, including techniques for 3D face geometry modeling; 3D face motion modeling; and 3D face motion tracking and animation. Then it discusses a unified framework for face modeling, analysis and synthesis. In this framework, the authors present new methods for modeling complex natural facial motion, as well as face appearance variations due to illumination and subtle motion. Then the authors apply the framework to face tracking, expression recognition and face avatar for HCI interface.They conclude this book with comments on future work in the 3D face processing framework. 3D Face Processing: Modeling, Analysis and Synthesis will interest those working in face processing for intelligent human computer interaction and video surveillance. It contains a comprehensive survey on existing face processing techniques, which can serve as a reference for students and researchers. It also covers in-depth discussion on face motion analysis and synthesis algorithms, which will benefit more advanced graduate students and researchers.

The need for natural and effective Human-Computer Interaction (HCI) is increasingly important due to the prevalence of computers in human activities. Computer vision and pattern recognition continue to play a dominant role in the HCI realm. However, computer vision methods often fail to become pervasive in the field due to the lack of real-time, robust algorithms, and novel and convincing applications. This state-of-the-art contributed volume is comprised of articles by prominent experts in computer vision, pattern recognition and HCI. It is the first published text to capture the latest research in this rapidly advancing field with exclusive focus on real-time algorithms and practical applications in diverse and numerous industries, and it outlines further challenges in these areas. Real-Time Vision for Human-Computer Interaction is an invaluable reference for HCI researchers in both academia and industry, and a useful supplement for advanced-level courses in HCI and Computer Vision.

3D Face Processing: Modeling, Analysis and Synthesis introduces the frontiers of 3D face processing techniques. It reviews existing 3D face processing techniques, including techniques for 3D face geometry modeling; 3D face motion modeling; and 3D face motion tracking and animation. Then it discusses a unified framework for face modeling, analysis and synthesis. In this framework, the authors present new methods for modeling complex natural facial motion, as well as face appearance variations due to illumination and subtle motion. Then the authors apply the framework to face tracking, expression recognition and face avatar for HCI interface. They conclude this book with comments on future work in the 3D face processing framework. 3D Face Processing: Modeling, Analysis and Synthesis will interest those working in face processing for intelligent human computer interaction and video surveillance. It contains a comprehensive survey on existing face processing techniques, which can serve as a reference for students and researchers. It also covers in-depth discussion on face motion analysis and synthesis algorithms, which will benefit more advanced graduate students and researchers.

Exploration of Visual Data presents latest research efforts in the area of content-based exploration of image and video data. The main objective is to bridge the semantic gap between high-level concepts in the human mind and low-level features extractable by the machines.The two key issues emphasized are "content-awareness" and "user-in-the-loop". The authors provide a comprehensive review on algorithms for visual feature extraction based on color, texture, shape, and structure, and techniques for incorporating such information to aid browsing, exploration, search, and streaming of image and video data. They also discuss issues related to the mixed use of textual and low-level visual features to facilitate more effective access of multimedia data.Exploration of Visual Data provides state-of-the-art materials on the topics of content-based description of visual data, content-based low-bitrate video streaming, and latest asymmetric and nonlinear relevance feedback algorithms, which to date are unpublished.

This book constitutes the refereed proceedings of the International Workshop on Human-Computer Interaction, HCI 2004, held at ECCV 2004 in Prague, Czech Republic in May 2004.The 19 revised full papers presented together with an introductory overview and an invited paper were carefully reviewed and selected from 45 submissions. The papers are organized in topical sections on human-robot interaction, gesture recognition and body tracking, systems, and face and head.

Human-Computer Interaction (HCI) lies at the crossroads of many scienti?c areas including arti?cial intelligence, computer vision, face recognition, motion tracking, etc. In order for HCI systems to interact seamlessly with people, they need to understand their environment through vision and auditory input. Mo- over, HCI systems should learn how to adaptively respond depending on the situation. The goal of this workshop was to bring together researchers from the ?eld of computer vision whose work is related to human-computer interaction. The selected articles for this workshop address a wide range of theoretical and - plication issues in human-computer interaction ranging from human-robot - teraction, gesture recognition, and body tracking, to facial features analysis and human-computer interaction systems. This year 74 papers from 18 countries were submitted and 22 were accepted for presentation at the workshop after being reviewed by at least 3 members of the Program Committee. We had therefore a very competitive acceptance rate of less than 30% and as a consequence we had a very-high-quality workshop. Wewouldliketo thankallmembersofthe ProgramCommitteefor their help in ensuring the quality of the papers accepted for publication. We are grateful to Dr. Jian Wang for giving the keynote address. In addition, we wish to thank the organizers of the 10th IEEE International Conference on Computer Vision and our sponsors, University of Amsterdam, Leiden Institute of Advanced Computer Science, and the University of Illinois at Urbana-Champaign, for support in setting up our workshop.

The interests and goals of HCI (human–computer interaction) include understanding, designing, building, and evaluating complex interactive systems - involving many people and technologies. Developments in software and hardware technologies are continuously driving applications in supporting our collaborative and communicative needs as social beings, both at work and at play. At the same time, similar developments are pushing the human–computer interface beyond the desktop and into our pockets, streets, and buildings. Developments in mobile, wearable, and pervasive communications and computing technologies provide exciting challenges and opportunities for HCI. This volume presents the proceedings of the HCI 2006 Workshop, held in conjunction with ECCV 2006 (European Conference on Computer Vision) in Graz, Austria. The goal of this workshop was to bring together researchers from the field of computer vision whose work is related to human–computer interaction.The 11 papers presented in this volume were carefully reviewed and selected from 27 submissions. They address a wide range of theoretical and application issues in human–computer interaction. There were four thematic sessions named: face analysis; gesture and emotion recognition; event detection; and applications.

The refereed proceedings of the Second International Conference on Image and Video Retrieval, CIVR 2003, held in Urbana-Champaign, IL, USA in July 2003. The 46 revised full papers presented together with an introduction and 2 invited papers were carefully reviewed and selected from 110 submissions. The papers are organized in topical sections on image retrieval, indexing strategies and structures, feature based retrieval, semantic and learning, video retrieval, user studies, applications, video summarization and analysis, and performance.

This text brings together concepts and approaches from the fields of photogrammetry and computer vision. In particular, it examines techniques relating to quantitative image analysis, such as orientation, camera modelling, system calibration, self-calibration and error handling. The chapters have been contributed by experts in the relevant fields, and there are examples from automated inspection systems and other real-world cases. The book provides study material for students, researchers, developers and practitioners.

This volume in the Lecture Notes of Artificial Intelligence represents the first book on human computing. We introduced the notion of human computing in 2006 and organized two events that were meant to explain this notion and the research conducted worldwide in the context of this notion. The first of these events was a Special Session on Human Computing that took place during the Eighth International ACM Conference on Multimodal Interfaces (ICMI 2006), held in Banff, Canada, on November 3, 2006. The theme of the conference was multimodal collaboration and our Special Session on Human Computing was a natural extension of the discussion on this theme. We are grateful to the organizers of ICMI 2006 for supporting our efforts to organize this Special Session during the conference. The second event in question was a Workshop on AI for Human Computing organized in conjunction with the 20th International Joint Conference on Artificial Intelligence (IJCAI 2007), held in Hyderabad (India), on January 6, 2007. The main theme of IJCAI 2007 was AI and its benefits to society. Our workshop presented a vision of the future of computing technology in which AI, in particular machine learning and agent technology, plays an essential role. We want to thank the organizers of IJCAI 2007 for their support in the organization of the Workshop on AI for Human Computing.

Welcome to the International Workshop on Multimedia Content Analysis and Mining, MCAM 2007. Our workshop gives a snapshot of the current wor- wide research in multimedia analysis. Through recent advances in computing, networking, and data storage, multimedia will create new interesting technical possibilities in a wide range of ?elds, such as entertainment, commerce, science, medicine, and public safety. To bene?t from this potential, developers need - liable techniques for the analysis, search, and management of multimedia data, as well as distributed system architectures in which these techniques can be embedded to e?ectively help the users. Recently, there have been many workshops and meetings dedicated to M- timedia, but most of them followed the standard pattern consisting mainly of oral and poster presentations. We all feel that there is a need for a meeting that really is a workshop, i.e., one that provides a lot of opportunities for discussions and dissemination. Consequently, the format of the workshop consists of 4 p- els on hot topics in multimedia followedby discussions and poster presentations. The topics of the panels are: Multimedia Analysis and Applications, Multi- dia Search and Mining, P2P Streaming, and Security. This volume contains a number of invited contributions from experts in the area as well as the selected regular contributions. This year 139 submissions from 13 countries were submitted and 46 were accepted for presentation at the conference after being reviewed by the Program Committee members.

This book constitutes the refereed proceedings of the International Workshop on Human Computer Interaction, HCI 2007. Coverage in the 16 revised full papers presented includes affective detection and recognition, human motion tracking, multimedia data modeling and visualization, HCI issues in image/video retrieval, learning in HCI, input and interaction techniques, perceptual user interfaces, wearable and pervasive technologies in HCI and intelligent virtual environments.

This book brings together concepts and approaches from the fields of photogrammetry and computer vision. In particular, it examines techniques relating to quantitative image analysis, such as orientation, camera modelling, system calibration, self-calibration and error handling. The chapters have been contributed by experts in the relevant fields, and there are examples from automated inspection systems and other real-world cases. The book provides study material for students, researchers, developers and practitioners.

The NATO Advanced Study Institute (ASI) on Face Recognition: From Theory to Applications took place in Stirling, Scotland, UK, from June 23 through July 4, 1997. The meeting brought together 95 participants (including 18 invited lecturers) from 22 countries. The lecturers are leading researchers from academia, govemment, and industry from allover the world. The lecturers presented an encompassing view of face recognition, and identified trends for future developments and the means for implementing robust face recognition systems. The scientific programme consisted of invited lectures, three panels, and (oral and poster) presentations from students attending the AS!. As a result of lively interactions between the participants, the following topics emerged as major themes of the meeting: (i) human processing of face recognition and its relevance to forensic systems, (ii) face coding, (iii) connectionist methods and support vector machines (SVM), (iv) hybrid methods for face recognition, and (v) predictive learning and performance evaluation. The goals of the panels were to provide links among the lectures and to emphasis the themes of the meeting. The topics of the panels were: (i) How the human visual system processes faces, (ii) Issues in applying face recognition: data bases, evaluation and systems, and (iii) Classification issues involved in face recognition. The presentations made by students gave them an opportunity to receive feedback from the invited lecturers and suggestions for future work.

Motion and Structure from Image Sequences is invaluablereading for researchers, graduate students, and practicingengineers dealing with computer vision. It presents abalanced treatment of the theoretical and practical issues,including very recent results - some of which are publishedhere for the first time.The topics covered in detail are:- image matching and optical flow computation- structure from stereo- structure from motion- motion estimation- integration of multiple views- motion modeling and predictionAspects such as uniqueness of the solution, degeneracyconditions, error analysis, stability, optimality, androbustness are also investigated. These details togetherwith the fact that the algorithms are accessible withoutnecessarily studying the rest of the material, make thisbook particularly attractive to practitioners.

This book provides a broader introduction to the theories and applications of sparse coding techniques in computer vision research. It introduces sparse coding in the context of representation learning, illustrates the fundamental concepts, and summarizes the most active research directions. A variety of applications of sparse coding are discussed, ranging from low-level image processing tasks such as super-resolution and de-blurring to high-level semantic understanding tasks such as image recognition, clustering and fusion.The book is suitable to be used as an introductory overview to this field, with its theoretical part being both easy and precious enough for quick understanding. It is also of great value to experienced researchers as it offers new perspective to the underlying mechanism of sparse coding, and points out potential future directions for different applications.