Guide to Biometric Reference Systems and Performance Evaluation (eBook)

Foreword 5
Preface 7
Acknowledgments 9
Contents 10
Contributors 18
Acronyms 24
Symbols 28
Introduction—About the Need of an Evaluation Framework in Biometrics 30
1.1 Reference Software 30
1.2 Biometrics 32
1.3 Databases 33
1.4 Risks 34
1.5 Biometric “Menagerie” 35
1.6 Spoofing 35
1.7 Evaluation 35
1.8 Evaluation Campaigns 36
1.9 Outline of This Book 37
Acknowledgments 38
References 38
The BioSecure Benchmarking Methodology for Biometric Performance Evaluation 40
2.1 Introduction 40
2.2 Terminology 41
2.3 Description of the Proposed Evaluation Framework 48
2.4 Use of the Benchmarking Packages 51
2.5 Conclusions 52
References 52
Iris Recognition 54
3.1 Introduction 54
3.2 State of the Art in Iris Recognition 55
3.3 Current Issues and Challenges 58
3.4 Existing Evaluation Databases, Campaigns and Open- source Software 59
3.5 The BioSecure Evaluation Framework for Iris 63
3.6 Research Systems Evaluated within the Benchmarking Framework 69
3.7 Fusion Experiments 75
3.8 Conclusions 77
References 77
Fingerprint Recognition 79
4.1 Introduction 79
4.2 State of the Art in Fingerprint Recognition 81
4.3 Fingerprint Databases 90
4.4 Fingerprint Evaluation Campaigns 93
4.5 The BioSecure Benchmarking Framework 97
4.6 Research Algorithms Evaluated within the Benchmarking Framework 103
4.7 Experimental Results within the Benchmarking Framework 108
4.8 Conclusions 112
Acknowledgments 112
References 113
Hand Recognition 117
5.1 Introduction 117
5.2 State of the Art in Hand Recognition 118
5.3 The BioSecure Evaluation Framework for Hand Recognition 122
5.4 More Experimental Results with the Reference System 129
5.5 Appearance-Based Hand Recognition System [BU] 137
5.6 Conclusions 149
References 150
Online Handwritten Signature Verification 153
6.1 Introduction 153
6.2 State of the Art in Signature Verification 156
6.3 Databases 161
6.4 Evaluation Campaigns 167
6.5 The BioSecure Benchmarking Framework for Signature Verification 167
6.6 Research Algorithms Evaluated within the Benchmarking Framework 176
6.7 Conclusions 189
Acknowledgments 191
References 192
Text-independent Speaker Verification 194
7.1 Introduction 194
7.2 Review of Text-independent Speaker Verification 196
7.3 Speaker Verification Evaluation Campaigns and Databases 213
7.4 The BioSecure Speaker Verification Benchmarking Framework 215
7.5 How to Reach State-of-the-art Speaker Verification Performance Using Open Source Software 220
7.6 Conclusions and Perspectives 230
Acknowledgments 233
References 233
2D Face Recognition 239
8.1 State of the Art in Face Recognition: Selected Topics 239
8.2 2D Face Databases and Evaluation Campaigns 258
8.3 The BioSecure Benchmarking Framework for 2D Face 259
8.4 Method 1: Combining Gabor Magnitude and Phase Information [ TELECOM SudParis] 263
8.5 Method 2: Subject-Specific Face Verification via Shape- Driven Gabor Jets ( SDGJ) [ University of Vigo] 273
8.6 Method 3: SIFT-based Face Recognition with Graph Matching [ UNISS] 277
8.7 Comparison of the Presented Approaches 279
8.8 Conclusions 281
Acknowledgments 281
References 281
3D Face Recognition 289
9.1 Introduction 289
9.2 State of the Art in 3D Face Recognition 290
9.3 3D Face Databases and Evaluation Campaigns 305
9.4 Benchmarking Framework for 3D Face Recognition 308
9.5 More Experimental Results with the 3D Reference System 315
9.6 Conclusions 316
References 317
Talking-face Verification 322
10.1 Introduction 322
10.2 State of the Art in Talking-face Verification 323
10.3 Evaluation Framework 333
10.4 Research Systems 340
10.5 Conclusion 346
Acknowledgments 347
References 347
BioSecure Multimodal Evaluation Campaign 2007 ( BMEC’ 2007) 352
11.1 Introduction 352
11.2 Scientific Objectives 353
11.3 Existing Multimodal Databases 354
11.4 BMEC Database 356
11.5 Performance Evaluation 362
11.6 Experimental Results 366
11.7 Conclusion 390
Acknowledgements 391
Appendix 391
11.8 Equal Error Rate 391
11.9 Parametric Confidence Intervals 392
11.10 Participants 393
References 394
Index 397