search for books and compare prices
Tables of Contents for Guide to Biometrics
Chapter/Section Title
Page #
Page Count
List of Figures
xiii
List of Tables
xxi
Foreword
xxv
Preface
xxvii
I Basics of Biometrics
1
60
1 Introduction
3
14
1.1 Person authentication, in general
4
2
1.2 A quick overview of biometrics
6
5
1.2.1 Biometric identifiers
6
1
1.2.2 Biometric subsystems
7
1
1.2.3 System performance and design issues
8
2
1.2.4 Competing system design issues
10
1
1.3 Biometric identification
11
1
1.4 Biometric verification
12
1
1.5 Biometric enrollment
13
1
1.6 Biometric system security
14
3
2 Authentication and Biometrics
17
14
2.1 Secure authentication protocols
17
1
2.2 Access control security services
18
2
2.3 Authentication methods
20
2
2.4 Authentication protocols
22
2
2.5 Matching biometric samples
24
3
2.5.1 Identification
25
1
2.5.2 Screening
26
1
2.5.3 Verification
26
1
2.5.4 Continuity of identity
26
1
2.6 Verification by humans
27
1
2.7 Passwords versus biometrics
28
1
2.8 Hybrid methods
29
2
3 The Common Biometrics
31
20
3.1 Fingerprint recognition
31
5
3.1.1 Acquisition devices
32
2
3.1.2 Matching approaches
34
1
3.1.3 Minutiae
35
1
3.2 Face recognition
36
4
3.2.1 Imaging
36
1
3.2.2 Local and global models
37
2
3.2.3 Challenges
39
1
3.3 Speaker recognition
40
3
3.3.1 Application categories
41
1
3.3.2 Acoustic features
42
1
3.4 Iris recognition
43
2
3.5 Hand geometry
45
2
3.6 Signature verification
47
4
4 Additional Biometrics
51
10
4.1 DNA
52
1
4.2 Retina recognition
53
1
4.3 Thermograms
54
1
4.4 Gait
55
1
4.5 Keystroke
56
1
4.6 Ear recognition
56
1
4.7 Skin reflectance
57
1
4.8 Lip motion
57
1
4.9 Body odor
58
3
II Performance and Selection
61
94
5 Basic System Errors
63
24
5.1 Matching
63
7
5.1.1 Two kinds of errors
65
1
5.1.2 Score distributions
66
2
5.1.3 Estimating errors from data
68
1
5.1.4 Error rates of match engines
69
1
5.1.5 Definitions of FAR and FRR, positive authentication
70
1
5.2 The Receiver Operating Characteristic (ROC)
70
8
5.2.1 Variations on ROCS
72
1
5.2.2 Using the ROC curve
73
1
5.2.3 Expressing the "quality" of a ROC curve
74
4
5.3 Error conditions "specific" to biometrics
78
1
5.4 Negative authentication
79
2
5.5 Trade-offs
81
6
5.5.1 Convenience versus security
81
1
5.5.2 Cost versus security of positive authentication
82
1
5.5.3 Cost of negative authentication
83
4
6 Identification System Errors
87
18
6.1 Problem overview
88
6
6.1.1 Winnowing
89
1
6.1.2 Approaches and implementations
89
1
6.1.3 Open and closed worlds
90
1
6.1.4 Ranking versus scoring
91
3
6.1.5 Identities versus templates
94
1
6.2 Generic evaluation criteria
94
3
6.2.1 Reliability and selectivity
94
2
6.2.2 The recall precision curve
96
1
6.3 Threshold-based identification
97
4
6.3.1 Simple FAR(m) and FRR(m)
98
1
6.3.2 Unambiguous answers
99
2
6.4 Rank-based identification
101
4
6.4.1 Rank-order statistics
101
1
6.4.2 Rank Probability Mass (RPM) function
102
1
6.4.3 Choosing the list length
103
2
7 Performance Testing
105
24
7.1 Measuring performance
105
6
7.1.1 Technology evaluations
106
2
7.1.2 Scenario evaluations
108
1
7.1.3 Comparison of the methods
108
2
7.1.4 Limits to evaluation
110
1
7.2 Implications of error rates
111
2
7.3 Face, finger, and voice
113
1
7.4 Iris and hand
114
3
7.5 Signature
117
3
7.6 Summary of verification accuracies
120
1
7.7 Identification system testing
121
6
7.7.1 Biometric data and ROC, CMC
121
1
7.7.2 Biometric search engines
122
1
7.7.3 1: m search engine testing
123
2
7.7.4 Face Recognition and Verification Test 2000
125
1
7.7.5 FRVT 2002
126
1
7.8 Caveats
127
2
8 Selecting a Biometric
129
26
8.1 Biometric attributes
130
3
8.1.1 Sensor properties
131
1
8.1.2 Template sizes
131
1
8.1.3 Scalability
132
1
8.2 Application properties
133
5
8.2.1 Wayman's application taxonomy
133
2
8.2.2 Weighting the factors
135
3
8.3 Evaluating options
138
6
8.3.1 Mismatch calculation
138
3
8.3.2 Zephyr™ charts
141
1
8.3.3 Open competitions for comparing biometrics
142
2
8.4 Affordability and cost
144
1
8.5 Positives and negatives of the biometrics
145
5
8.5.1 Fingerprint
146
1
8.5.2 Face
146
1
8.5.3 Voice
147
1
8.5.4 Iris
148
1
8.5.5 Hand
148
1
8.5.6 Signature
149
1
8.6 Biometric myths and misrepresentations
150
5
III System Issues
155
88
9 Creating and Maintaining Databases
157
20
9.1 Enrollment policies
158
4
9.1.1 Enrollment for positive authentication
159
1
9.1.2 Enrollment for screening
160
1
9.1.3 Social issues
161
1
9.2 The zoo
162
2
9.3 Biometric sample quality control
164
2
9.4 Training
166
3
9.4.1 Training and modeling
167
1
9.4.2 There are two aspects to training
167
2
9.5 Enrollment is system training
169
8
9.5.1 Database integrity
170
2
9.5.2 Probabilistic enrollment
172
1
9.5.3 Modeling the world
172
1
9.5.4 Modeling the rest of the world-cohorts
173
2
9.5.5 Updating the probabilities
175
1
9.5.6 Use of the probabilities
176
1
10 Large-Scale Applications
177
16
10.1 Example systems
177
2
10.2 Required accuracies
179
3
10.2.1 Enrollment False Positives
179
2
10.2.2 Enrollment False Negatives
181
1
10.2.3 Authentication accuracy
181
1
10.3 Matcher sizing
182
2
10.3.1 Match throughput
182
1
10.3.2 Total number of matches
183
1
10.4 Exception handling
184
2
10.5 Voter registration
186
2
10.6 National ID systems
188
3
10.7 How feasible is large scale?
191
2
11 Integrating Information
193
18
11.1 Integration methods
193
3
11.2 Decision level integration
196
7
11.2.1 Boolean combination
196
2
11.2.2 Binning and filtering
198
3
11.2.3 Dynamic authentication protocols
201
2
11.3 Score level integration
203
7
11.3.1 Normal distributions
205
2
11.3.2 Degenerate cases
207
1
11.3.3 From thresholds to boundaries
208
2
11.4 Alternatives
210
1
12 Thwarting Attacks
211
18
12.1 Pattern recognition model
211
2
12.2 Attacking biometric identifiers
213
2
12.3 Front-end attacks
215
1
12.4 Circumvention
216
1
12.5 Back-end attacks
217
2
12.6 Other attacks
219
1
12.7 Combining smartcards and biometrics
220
1
12.8 Challenge and response
221
1
12.9 Cancellable biometrics
222
7
12.9.1 Privacy
222
2
12.9.2 Intentional, repeatable transforms
224
1
12.9.3 Signal domain distortions
224
2
12.9.4 Feature domain distortions
226
1
12.9.5 Relation to compression and encryption
226
3
13 APIs, Standards, and Databases
229
14
13.1 Interface standards
229
4
13.1.1 Application programming interfaces
230
1
13.1.2 Data structure and security standards
231
2
13.2 Databases
233
6
13.2.1 Fingerprint databases
233
2
13.2.2 Face databases
235
2
13.2.3 Speaker recognition databases
237
1
13.2.4 Other databases
238
1
13.3 Certifications
239
2
13.4 Legislation
241
2
IV Mathematical Analyses
243
92
14 A Biometric's Individuality
245
24
14.1 Approaches to individuality
245
2
14.2 Empirical individuality studies
247
1
14.3 A partial iris model
248
5
14.3.1 FAR modeling
249
2
14.3.2 FRR calculation
251
1
14.3.3 Numeric evaluation
251
2
14.4 Fingerprint individuality
253
10
14.4.1 A simple model
254
2
14.4.2 Probabilistic scoring
256
1
14.4.3 A more complex model
257
3
14.4.4 Model comparison
260
1
14.4.5 Imposing structure
261
1
14.4.6 Minutiae distributions
262
1
14.5 Standards for legal evidence
263
1
14.6 Expert fingerprint testimony
264
3
14.7 Remarks
267
2
15 System Errors Revisited
269
24
15.1 Estimating the match score mean
270
7
15.1.1 Confidence intervals
271
1
15.1.2 Parametric method
271
1
15.1.3 Nonparametric methods
272
2
15.1.4 Estimating the quantiles
274
1
15.1.5 The Bootstrap
275
2
15.2 The data are not independent
277
3
15.2.1 How are the data obtained?
277
1
15.2.2 The subsets bootstrap
278
1
15.2.3 Estimating the match score mean (continued)
279
1
15.3 Confidence intervals on the FRR and FAR
280
6
15.3.1 Distributions F and a G specify the engine
280
2
15.3.2 Error in the FRR estimate
282
1
15.3.3 Error in the FAR estimate
283
3
15.4 How good are the confidence intervals?
286
7
15.4.1 The set of samples X
287
1
15.4.2 More independent samples X1, X2
288
1
15.4.3 The match score set X
289
1
15.4.4 Confidence interval validation
290
1
15.4.5 Sample means versus the true mean
291
2
16 Advanced Topics
293
32
16.1 What is needed to compare matchers?
294
1
16.2 The data needed to compare matchers
295
1
16.3 Cost functions in detail
296
6
16.3.1 Statistical tests using costs
299
1
16.3.2 Estimating operating thresholds
300
2
16.4 Statistical tests at fixed FRR or FAR
302
3
16.4.1 The FAR is known
303
1
16.4.2 Estimating the FAR first
304
1
16.5 Biometric searching and sorting-the CMC
305
10
16.5.1 Estimating the CMC
306
2
16.5.2 Closed universe
308
1
16.5.3 The CMC, FAR/FRR relation
309
5
16.5.4 Opening up the closed world
314
1
16.6 Biometric searching and ranking
315
7
16.6.1 Matching versus searching
315
1
16.6.2 Biometric search engine
316
1
16.6.3 Rank engine
317
2
16.6.4 Search errors
319
1
16.6.5 Probability of correct identification
320
2
16.7 ROC versus CMC
322
3
17 What's next?
325
10
17.1 Recommendations
326
2
17.2 Current issues
328
2
17.3 The future
330
3
17.4 A final word
333
2
References
335
18
Index
353