Dorothy Monekosso. Paolo Remagnino Yoshinori Kuno Editors Intelligent Environments Methods, Algorithms and Applications ~ Springer
Contents Preface............................................................ List of Contributors VB XV 1 Intelligent Environments: Methods, Algorithms and Applications... Dorothy N. Monekosso, Paolo Remagnino, and Yoshinori Kuno 1.1 Intelligent Environments I 1.1.1 What Is An Intelligent Environment?................. 2 1.1.2 How Is An Intelligent Environment Built?............. 2 1.2 Technology for Intelligent Environments...................... 2 1.3 Research Projects......................................... 4 1.3.1 Private Spaces 4 1.3.2 Public Spaces..................................... 5 1.3.3 Middleware...................................... 7 1.4 Chapter Themes in This Collection 8 1.5 Conclusion............................................... 9 References..................................................... 10 2 A Pervasive Sensor System for Evidence-Based Nursing Care Support............................................... 13 Toshio Hori, Yoshifumi Nishida, and Shin'ichi Murakami 2.1 Introduction.............................................. 13 2.2 Evidence-Based Nursing Care Support 14 2.2.1 Background of the Project.......................... 14 2.2.2 Concept of Evidence-Based Nursing Care Support...... 15 2.2.3 Initial Goal of the Project: Falls Prevention............ 16 2.2.4 Second Goal of the Projecl: Obtaining ADL of Inhabitants..................................... 17 2.3 Related Work............................................. 18 ix
x Contents 2.4 Overview and Implementations of the System 19 2.4.1 Overview of the Evidence-Based Nursing Care Support System................................... 19 2.4.2 System Implementations 20 2.5 Experiments and Analyses.................................. 24 2.5.1 Tracking a Wheelchair for Falls Prevention............ 24 2.5.2 Activity Transition Diagram: Transition of Activities in One Day....................................... 25 2.5.3 Quantitative Evaluation of Daily Activities 26 2.5.4 Probability of "Toi]et" Activity...................... 28 2.5.5 Discussion of the Experimental Results............... 29 2.6 Prospect of the Evidence-Based Nursing Care Support System 30 2.7 Conclusions 31 Referenccs..................................................... 32 3 Anomalous Behavior Detection: Supporting Independent Living.... 35 Dorothy N. Monekosso and Paolo Remagnino 3.1 Introduction.............................................. 35 3.2 Related work............................................. 36 3.3 Methodology............................................. 37 3.3.1 Unsupervised Classification Techniques............... 37 3.3.2 Using HMM to Model Behavior..................... 38 3.4 Experimental Setup and Data Collection...................... 39 3.4.1 Noisy Data: Sources of Error........................ 40 3.4.2 Learning activities................................. 40 3.5 Experimental Results...................................... 41 3.5.1 Instance Class Annotation 41 3.5.2 Data Preprocessing 41 3.5.3 Models: Unsuperviscd Classification: C1ustering and Allocation of Activities to Clusters 43 3.5.4 Behaviors: Discovering Patterns in Activities 45 3.5.5 Behaviors: Discovering Anomalous Patterns of Activity................................ 46 3.6 Discussion............................................... 48 3.7 Conclusions.............................................. 49 References..................................................... 49 4 Sequential Pattern Mining for Cooking-Support Robot, 51 Yasushi Nakauchi 4.1 Introduction.............................................. 5] 4.2 System Design 53 4.2.1 Inference from Series of Human Actions, 53 4.2.2 Time Sequence Data Mining........................ 54 4.2.3 Human Behavior Inference Algorithm................ 54 4.2.4 Activity Support of Human 57
Contents XI 4.3 Implementation........................................... 59 4.3.1 IC Tag System.................................... 59 4.3.2 Inference of Human's Next Action 60 4.3.3 Cooking Support Interface.......................... 61 4.4 Experimental Results...................................... 63 4.5 Conc1usions.............................................. 65 References..................................................... 66 5 Robotic, Sensory and Problem-Solving Ingredients for the Future Horne......................................... 69 Amedeo Cesta, Luca locchi, G. Riccardo Leone, Daniele Nardi, Federico Pecora, and Riccardo Rasconi 5.1 Introduction.............................................. 69 5.1.1 Components of the Multiagent System 70 5.2 The Robotic Platform Mobility Subsystem.................... 71 5.3 The Interaction Manager................................... 73 5.4 Environmental Sensors for People Tracking and Posture Recognition 74 5.5 Monitoring Activities of Daily Living 76 5.5.1 Schedule Representation and Execution Monitoring..... 77 5.5.2 Constraint Management in the ROBOCARE Context..... 78 5.5.3 From Constraint Violations to VerbalInteraction 81 5.6 Multiagent Coordination Infrastructure....................... 82 5.6.1 Casting the MAC Problem to DCOP.................. 83 5.6.2 Cooperatively Solving the MAC Problem 86 5.7 Conc1usions.............................................. 87 References..................................................... 88 6 Ubiquitous Stereo Vision for Human Sensing.................... 91 Ikushi Yoda and Katsuhiko Sakae 6.1 Introduction.............................................. 91 6.2 Ubiquitous Stereo Vision................................... 93 6.2.1 Concept of Ubiquitous Stereo Vision " 93 6.2.2 Server-Client Model for USV 93 6.2.3 Real Utilization Cases.............................. 94 6.3 Hierarchical Utilization of 3D Data and Personal Recognition.... 95 6.3.1 Acquisition of 3D Range Information................. 95 6.3.2 Projection to Floor Plane 96 6.4 Recognition of Multiple Persons and Interface................. 98 6.4.1 Pose Recognition for Multiple People................. 99 6.4.2 Personalldentification 100 6.4.3 Interface for Space Control 101 6.5 Human Monitoring in Open Space (Safety Management Application) 101 6.5.1 Monitoring Railroad Crossing 101 6.5.2 Station Platform Edge Safety Management 103
XII Contents 6.5.3 Monitoring Huge Spaee 104 6.6 Conclusion and Future Work 105 Referenecs 106 7 Augmenting Professional Training, an Ambient Intelligence Approach, 109 B. Zhan, D.N. Monekosso, S. Rush, P. Remagnino, and S.A. Velastin 7.1 Introduetion 109 7.2 Color Tracking of People 112 7.3 Counting People by Spatial Relationship Analysis 113 7.3.1 Simple People Counting Algorithm 113 7.3.2 Graphs of Blobs 114 7.3.3 Estimation of Distanee Between Blobs 116 7.3.4 Temporal Pyramid für Distance Estimation 117 7.3.5 Probabilistic Estimation of Groupings 119 7.3.6 Grouping Blobs 120 7.4 Experimental Results 121 7.5 Conclusions 124 References..................................................... 124 8 Stereo Omnidirectional System (SOS) and Its Applications 127 Yutaka Satoh and Katsuhiko Sakaue 8.1 Introduction 127 8.2 System Configuration 128 8.3 Image integration 131 8.4 Generation of Stable Images at Arbitrary Rotation 133 8.5 An example Applieation: Intelligent Electric Wheelchair 136 8.5.1 Overview 136 8.5.2 System Configuration 136 8.5.3 Obstacle Detection 138 8.5.4 Gesture / Posture Detection 140 8.6 Conclusions 140 References 140 9 Video Analysis for Ambient Intelligence in Urban Environments..., 143 Andrea Prati and Rita Cueehiara 9.1 Introduction 143 9.2 Visual Data for Urban AmI 144 9.2.1 Video Surveillance in Urban Environment 145 9.2.2 The LAICA Project 148 9.3 Automatie Video Proeessing for People Traeking 149 9.3.1 People Detection and Traeking from Single Static Camera 150 9.3.2 People Detection and Tracking from Distributed Cameras 152
Contents Xlll 9.3.3 People Detection and Tracking from Moving Cameras 154 9A Privacy and Ethical Issues 155 References 157 10 From Monomodal to Multimodal: Affect Recognition Using Visual Modalities " 161 Hatice Gunes and Massimo Piccardi 10.1 Introduction 161 10.2 Organization of the Chapter 163 10.3 From Monomodal to Multimodal: Changes and Challenges 164 10.3.1 Background Research 164 10.3.2 Data Collection 168 10.3.3 Data Annotation 169 1O.3A Synchrony/Asynchrony Between Modalities 171 10.3.5 Data Integration/Fusion 172 10.3.6 Information Complementarity/Redundancy 174 10.3.7 Information Content of Modalities 176 IOA Monomodal Systems Recognizing Affective Face or Body Movement 177 10.5 Multimodal Systems Recognizing Affect from Face and Body Movement 179 10.5.1 Project 1: Multimodal Affect Analysis for Future Cars " 179 10.5.2 Project 2: Emotion Analysis in Man-Machine Interaction Systems 182 10.5.3 Project 3: Multimodal Affect Recognition in Learning Environments 183 10.5A Project 4: FABO-Fusing Face and Body Gestures for Bimodal Emotion Recognition 184 10.6 Multimodal Affect Systems: The Future 185 References 187 11 Importance of Vision in Human-Robot Communication: Understanding Speech Using Robot Vision and Demonstrating Proper Actions to Human Vision " 191 Yoshinori Kuno, Michie Kawashima, Keiichi Yamazaki, and Akiko Yamazaki 11.1 Introduction 191 11.2 Understanding Simplified Utterances Using Robot Vision 193 11.2.1 Inexplicit Utterances 193 11.2.2 Information Obtained by Vision 194 11.2.3 Language Processing 195 11.2A Vision Processing 195 11.2.5 Synchronization Between Speech and Vision 197 11.2.6 Experiments 199
xiv Contents 11.3 Communicative Head Gestures for Museum Guide Robots 200 11.3.1 Observations from Guide-Visitor Interaction 201 11.3.2 Prototype Museum Guide Robot 203 11.3.3 Experiments at a Museum 206 IIA Conclusion 208 References 209 Index 211