In this book, the authors present the latest research results in the multimedia and semantic web communities, bridging the "Semantic Gap"

This book explains, collects and reports on the latest research results that aim at narrowing the so-called multimedia "Semantic Gap": the large disparity between descriptions of multimedia content that can be computed automatically, and the richness and subjectivity of semantics in user queries and human interpretations of audiovisual media. Addressing the grand challenge posed by the "Semantic Gap" requires a multi-disciplinary approach (computer science, computer vision and signal processing, cognitive science, web science, etc.) and this is reflected in recent research in this area. In addition, the book targets an interdisciplinary community, and in particular the Multimedia and the Semantic Web communities. Finally, the authors provide both the fundamental knowledge and the latest state-of-the-art results from both communities with the goal of making the knowledge of one community available to the other.

Key Features:

• Presents state-of-the art research results in multimedia semantics: multimedia analysis, metadata standards and multimedia knowledge representation, semantic interaction with multimedia
• Contains real industrial problems exemplified by user case scenarios
• Offers an insight into various standardisation bodies including W3C, IPTC and ISO MPEG
• Contains contributions from academic and industrial communities from Europe, USA and Asia
• Includes an accompanying website containing user cases, datasets, and software mentioned in the book, as well as links to the K-Space NoE and the SMaRT society web sites (http://www.multimediasemantics.com/)

This book will be a valuable reference for academic and industry researchers /practitioners in multimedia, computational intelligence and computer science fields. Graduate students, project leaders, and consultants will also find this book of interest.

List of Figures xiii

List of Tables xvii

List of Contributors xix

1 Introduction 1
Raphaël Troncy, Benoit Huet and Simon Schenk

2 Use Case Scenarios 7
Werner Bailer, Susanne Boll, Oscar Celma, Michael Hausenblas and Yves Raimond

2.1 Photo Use Case 8

2.1.1 Motivating Examples 8

2.1.2 Semantic Description of Photos Today 9

2.1.3 Services We Need for Photo Collections 10

2.2 Music Use Case 10

2.2.1 Semantic Description of Music Assets 11

2.2.2 Music Recommendation and Discovery 12

2.2.3 Management of Personal Music Collections 13

2.3 Annotation in Professional Media Production and Archiving 14

2.3.1 Motivating Examples 15

2.3.2 Requirements for Content Annotation 17

2.4 Discussion 18

Acknowledgements 19

3 Canonical Processes of Semantically Annotated Media Production 21
Lynda Hardman, Z¡êljko Obrenovic´ and Frank Nack

3.1 Canonical Processes 22

3.1.1 Premeditate 23

3.1.2 Create Media Asset 23

3.1.3 Annotate 23

3.1.4 Package 24

3.1.5 Query 24

3.1.6 Construct Message 25

3.1.7 Organize 25

3.1.8 Publish 26

3.1.9 Distribute 26

3.2 Example Systems 27

3.2.1 CeWe Color Photo Book 27

3.2.2 SenseCam 29

3.3 Conclusion and Future Work 33

4 Feature Extraction for Multimedia Analysis 35
Rachid Benmokhtar, Benoit Huet, Gaël Richard and Slim Essid

4.1 Low-Level Feature Extraction 36

4.1.1 What Are Relevant Low-Level Features? 36

4.1.2 Visual Descriptors 36

4.1.3 Audio Descriptors 45

4.2 Feature Fusion and Multi-modality 54

4.2.1 Feature Normalization 54

4.2.2 Homogeneous Fusion 55

4.2.3 Cross-modal Fusion 56

4.3 Conclusion 58

5 Machine Learning Techniques for Multimedia Analysis 59
Slim Essid, Marine Campedel, Gaël Richard, Tomas Piatrik, Rachid Benmokhtar and Benoit Huet

5.1 Feature Selection 61

5.1.1 Selection Criteria 61

5.1.2 Subset Search 62

5.1.3 Feature Ranking 63

5.1.4 A Supervised Algorithm Example 63

5.2 Classification 65

5.2.1 Historical Classification Algorithms 65

5.2.2 Kernel Methods 67

5.2.3 Classifying Sequences 71

5.2.4 Biologically Inspired Machine Learning Techniques 73

5.3 Classifier Fusion 75

5.3.1 Introduction 75

5.3.2 Non-trainable Combiners 75

5.3.3 Trainable Combiners 76

5.3.4 Combination of Weak Classifiers 77

5.3.5 Evidence Theory 78

5.3.6 Consensual Clustering 78

5.3.7 Classifier Fusion Properties 80

5.4 Conclusion 80

6 Semantic Web Basics 81
Eyal Oren and Simon Schenk

6.1 The Semantic Web 82

6.2 RDF 83

6.2.1 RDF Graphs 86

6.2.2 Named Graphs 87

6.2.3 RDF Semantics 88

6.3 RDF Schema 90

6.4 Data Models 93

6.5 Linked Data Principles 94

6.5.1 Dereferencing Using Basic Web Look-up 95

6.5.2 Dereferencing Using HTTP 303 Redirects 95

6.6 Development Practicalities 96

6.6.1 Data Stores 97

6.6.2 Toolkits 97

7 Semantic Web Languages 99
Antoine Isaac, Simon Schenk and Ansgar Scherp

7.1 The Need for Ontologies on the Semantic Web 100

7.2 Representing Ontological Knowledge Using OWL 100

7.2.1 OWL Constructs and OWL Syntax 100

7.2.2 The Formal Semantics of OWL and its Different Layers 102

7.2.3 Reasoning Tasks 106

7.2.4 OWL Flavors 107

7.2.5 Beyond OWL 107

7.3 A Language to Represent Simple Conceptual Vocabularies: SKOS 108

7.3.1 Ontologies versus Knowledge Organization Systems 108

7.3.2 Representing Concept Schemes Using SKOS 109

7.3.3 Characterizing Concepts beyond SKOS 111

7.3.4 Using SKOS Concept Schemes on the Semantic Web 112

7.4 Querying on the Semantic Web 113

7.4.1 Syntax 113

7.4.2 Semantics 118

7.4.3 Default Negation in SPARQL 123

7.4.4 Well-Formed Queries 124

7.4.5 Querying for Multimedia Metadata 124

7.4.6 Partitioning Datasets 126

7.4.7 Related Work 127

8 Multimedia Metadata Standards 129
Peter Schallauer, Werner Bailer, Raphaël Troncy and Florian Kaiser

8.1 Selected Standards 130

8.1.1 MPEG-7 130

8.1.2 EBU P_Meta 132

8.1.3 SMPTE Metadata Standards 133

8.1.4 Dublin Core 133

8.1.5 TV-Anytime 134

8.1.6 METS and VRA 134

8.1.7 MPEG-21 135

8.1.8 XMP, IPTC in XMP 135

8.1.9 EXIF 136

8.1.10 DIG35 137

8.1.11 ID3/MP3 137

8.1.12 NewsML G2 and rNews 138

8.1.13 W3C Ontology for Media Resources 138

8.1.14 EBUCore 139

8.2 Comparison 140

8.3 Conclusion 143

9 The Core Ontology for Multimedia 145
Thomas Franz, Raphaël Troncy and Miroslav Vacura

9.1 Introduction 145

9.2 A Multimedia Presentation for Granddad 146

9.3 Related Work 149

9.4 Requirements for Designing a Multimedia Ontology 150

9.5 A Formal Representation for MPEG-7 150

9.5.1 DOLCE as Modeling Basis 151

9.5.2 Multimedia Patterns 151

9.5.3 Basic Patterns 155

9.5.4 Comparison with Requirements 157

9.6 Granddad’s Presentation Explained by COMM 157

9.7 Lessons Learned 159

9.8 Conclusion 160

10 Knowledge-Driven Segmentation and Classification 163
Thanos Athanasiadis, Phivos Mylonas, Georgios Th. Papadopoulos, Vasileios Mezaris, Yannis Avrithis, Ioannis Kompatsiaris and Michael G. Strintzis

10.1 Related Work 164

10.2 Semantic Image Segmentation 165

10.2.1 Graph Representation of an Image 165

10.2.2 Image Graph Initialization 165

10.2.3 Semantic Region Growing 167

10.3 Using Contextual Knowledge to Aid Visual Analysis 170

10.3.1 Contextual Knowledge Formulation 170

10.3.2 Contextual Relevance 173

10.4 Spatial Context and Optimization 177

10.4.1 Introduction 177

10.4.2 Low-Level Visual Information Processing 177

10.4.3 Initial Region-Concept Association 178

10.4.4 Final Region-Concept Association 179

10.5 Conclusions 181

11 Reasoning for Multimedia Analysis 183
Nikolaos Simou, Giorgos Stoilos, Carsten Saathoff, Jan Nemrava, Vojt¡ech Sv´atek, Petr Berka and Vassilis Tzouvaras

11.1 Fuzzy DL Reasoning 184

11.1.1 The Fuzzy DL f-SHIN 184

11.1.2 The Tableaux Algorithm 185

11.1.3 The FiRE Fuzzy Reasoning Engine 187

11.2 Spatial Features for Image Region Labeling 192

11.2.1 Fuzzy Constraint Satisfaction Problems 192

11.2.2 Exploiting Spatial Features Using Fuzzy

Constraint Reasoning 193

11.3 Fuzzy Rule Based Reasoning Engine 196

11.4 Reasoning over Resources Complementary to Audiovisual Streams 201

12 Multi-Modal Analysis for Content Structuring and Event Detection 205
Noel E. O’Connor, David A. Sadlier, Bart Lehane, Andrew Salway, Jan Nemrava and Paul Buitelaar

12.1 Moving Beyond Shots for Extracting Semantics 206

12.2 A Multi-Modal Approach 207

12.3 Case Studies 207

12.4 Case Study 1: Field Sports 208

12.4.1 Content Structuring 208

12.4.2 Concept Detection Leveraging Complementary Text Sources 213

12.5 Case Study 2: Fictional Content 214

12.5.1 Content Structuring 215

12.5.2 Concept Detection Leveraging Audio Description 219

12.6 Conclusions and Future Work 221

13 Multimedia Annotation Tools 223
Carsten Saathoff, Krishna Chandramouli, Werner Bailer, Peter Schallauer and Raphaël Troncy

13.1 State of the Art 224

13.2 SVAT: Professional Video Annotation 225

13.2.1 User Interface 225

13.2.2 Semantic Annotation 228

13.3 KAT: Semi-automatic, Semantic Annotation of Multimedia Content 229

13.3.1 History 231

13.3.2 Architecture 232

13.3.3 Default Plugins 234

13.3.4 Using COMM as an Underlying Model: Issues and Solutions 234

13.3.5 Semi-automatic Annotation: An Example 237

13.4 Conclusions 239

14 Information Organization Issues in Multimedia Retrieval Using Low-Level Features 241
Frank Hopfgartner, Reede Ren, Thierry Urruty and Joemon M. Jose

14.1 Efficient Multimedia Indexing Structures 242

14.1.1 An Efficient Access Structure for Multimedia Data 243

14.1.2 Experimental Results 245

14.1.3 Conclusion 249

14.2 Feature Term Based Index 249

14.2.1 Feature Terms 250

14.2.2 Feature Term Distribution 251

14.2.3 Feature Term Extraction 252

14.2.4 Feature Dimension Selection 253

14.2.5 Collection Representation and Retrieval System 254

14.2.6 Experiment 256

14.2.7 Conclusion 258

14.3 Conclusion and Future Trends 259

Acknowledgement 259

15 The Role of Explicit Semantics in Search and Browsing 261
Michiel Hildebrand, Jacco van Ossenbruggen and Lynda Hardman

15.1 Basic Search Terminology 261

15.2 Analysis of Semantic Search 262

15.2.1 Query Construction 263

15.2.2 Search Algorithm 265

15.2.3 Presentation of Results 267

15.2.4 Survey Summary 269

15.3 Use Case A: Keyword Search in ClioPatria 270

15.3.1 Query Construction 270

15.3.2 Search Algorithm 270

15.3.3 Result Visualization and Organization 273

15.4 Use Case B: Faceted Browsing in ClioPatria 274

15.4.1 Query Construction 274

15.4.2 Search Algorithm 276

15.4.3 Result Visualization and Organization 276

15.5 Conclusions 277

16 Conclusion 279
Raphaël Troncy, Benoit Huet and Simon Schenk

References 281

Author Index 301

Subject Index 303