1. CUBS, University at Buffalo
Biometrics
CUBS, University at Buffalo

2. Conventional Security Measures
Possession or Token Based
Passport, IDs, Keys
License ,Smart cards,Swipe cards, Credit Cards
Knowledge Based
Username/password
PIN
Combination(P,K)
ATM
Disadvantages of Conventional Measures
Do not authenticate the user
Tokens can be lost or misused
Passwords can be forgotten
Multiple tokens and passwords difficult to manage
Repudiation
Conventional security systems rely on PINs, passwords and other token or key based methods for authentication and identification of users. Though these systems are easy to use, they are insecure as the tokens can be lost, stolen or used by more than one person. With each service requiring different form and means of identification, the multiplicity of authentication schemes becomes difficult to manage.

3. Biometrics Definition Examples Physical Biometrics
Biometrics is the science of verifying and establishing the identity of an individual through physiological features or behavioral traits
Examples
Physical Biometrics
Fingerprint, Hand Geometry,Iris,Face
Measurement Biometric
Dependent on environment/interaction
Behavioral Biometrics
Handwriting, Signature, Speech, Gait
Performance/Temporal biometric
Dependent on state of mind
Chemical Biometrics
DNA, blood-glucose
Biometrics offers a promising solution for reliable and uniform identification and verification of an individual. Biometrics is the science of verifying and establishing the identity of an individual through physiological features or behavioral traits. Physical biometrics rely on physiological features such as fingerprints, hand geometry, iris pattern, facial features etc. for identity verification. Behavioral biometrics depends upon behavioral features such as speech patterns, handwriting, signature, walking gait etc. for authentication. These traits are unique to an individual and hence cannot be misused, lost or stolen. Biometrics are based on established scientific principles as a basis for authentication.

4. Requirements of Biometrics
Universality
Each person should have the biometric
Uniqueness
Any two persons should have distinctive characteristics
Permanence
Characteristic should be invariant over time
Collectability
Characteristic should be easy to acquire
Acceptability
Is non-intrusive
Non repudiation
User cannot deny having accessed the system

5. General Biometric System
Sensor
Feature Extraction
Database
Enrollment
ID : 8809
Biometric
Sensor
Feature Extraction
Matching
Authentication
Result

6. Types of Authentication
Verification
Answers the question “Am I whom I claim to be?”
Identity of the user is known
1:1 matching
Identification
Answers the question “Who am I?”
Identity of the user is not known
1:N matching
Positive Recognition
Determines if an individual is in the database
Prevents multiple users from assuming same identity
Negative Recognition
Determines if an individual is NOT in the given database
Prevents single user from assuming multiple indentities

7. Aspects of a Biometric Systems
Sensor and devices
Types of sensors
Electrical and mechanical design
Feature representation and matching
Enhancement, preprocessing
Developing invariant representations
Developing matching algorithms
Evaluation
Testing
System Issues
Large Scale databases
Securing Biometric Systems
Ethical, Legal and Privacy Issues

8. Applications And Scope of Biometrics
Technologies
Horizontal Applications
Key Vertical Markets
Fingerprint
Civil ID
Government Sector
Facial Recognition
Surveillance and Screening
Travel and Transportation
Iris Scan
PC / Network Access
Financial Sector
Middleware
Retail / ATM / Point of Sale
Health Care
AFIS
Criminal ID
Law Enforcement
Voice Scan
eCommerce / Telephony
Hand Geometry
Physical Access / Time and Attendance
Signature Verification
Keystroke Dynamics

9. Biometric Modalities Common modalities Iris Fingerprint Face
Voice Verification
Hand Geometry
Signature
Other modalities
Retinal Scan
Odor
Gait
Keystroke dynamics
Ear recognition
Lip movement
The corrugated surface of the fingerprint is made up of ridges and valleys cover that the entire palmer surface of the hand. The flow pattern of these ridges and valleys are unique to each individual. These patterns that are used for identification and authentication. The image below shows the image of a fingerprint along with the distinguishing features on the print.
The flow of the ridges and patterns has been classified into 5 broad classes. This classification is used to catalog the fingerprints and also in authenticating two prints. Henry systems follow an elaborate classification scheme of cataloging and filing forensic prints. Fig1 shows the different classes of ridge flows. This methods cannot be used to distinguish between two fingerprints.
Fig 2 shows the distinguishing features on the fingerprint. These features are discontinuities or anomalies in the normal flow of ridges on the surface of the finger. These features are termed as minutiae (small details). There are eighteen different types of minutiae. Fig1 shows the most commonly encountered ones and their names. Fig 2 shows a thumbprint captured on paper. These are typically the kind of images that forensic AFIS (Automatic Fingerprint Identification Systems) have to deal with. The quality of the print not only deteriorates during capture but also during storage and hence AFIS systems are more sophisticated than their biometric counterparts.

10. Fingerprint Verification
Fingerprints can be classified based on the ridge flow pattern
The corrugated surface of the fingerprint is made up of ridges and valleys cover that the entire palmer surface of the hand. The flow pattern of these ridges and valleys are unique to each individual. These patterns that are used for identification and authentication. The image below shows the image of a fingerprint along with the distinguishing features on the print.
The flow of the ridges and patterns has been classified into 5 broad classes. This classification is used to catalog the fingerprints and also in authenticating two prints. Henry systems follow an elaborate classification scheme of cataloging and filing forensic prints. Fig1 shows the different classes of ridge flows. This methods cannot be used to distinguish between two fingerprints.
Fig 2 shows the distinguishing features on the fingerprint. These features are discontinuities or anomalies in the normal flow of ridges on the surface of the finger. These features are termed as minutiae (small details). There are eighteen different types of minutiae. Fig1 shows the most commonly encountered ones and their names. Fig 2 shows a thumbprint captured on paper. These are typically the kind of images that forensic AFIS (Automatic Fingerprint Identification Systems) have to deal with. The quality of the print not only deteriorates during capture but also during storage and hence AFIS systems are more sophisticated than their biometric counterparts.
Fingerprints can be distinguished based on the ridge characteristics

11. Feature Extraction X Y θ T
106 26
320 R
153 50
335
255 81
215 B

12. Matching T(ΔX, ΔY , Δθ)? X Y θ T X Y θ T Rotation Scaling Translation
Elastic distortion X Y θ T
106 26
320 R
153 50
335
255 81
215 B X Y θ T
215 08
120 R
213 20
145
372 46
109 B
T(ΔX, ΔY , Δθ)?

13. Face Recognition:Eigen faces approach
Face detection and localization
Eigen faces
Normalization

14. Face Feature Representations
Facial Parameters
Eigen faces
Semantic model

15. Speaker Identification
Speaker Recognition
Speaker Recognition
Speaker Identification
Speaker Verification
Speaker Detection
Text
Dependent
Independent
Speech Codecs
IVR
Computer Access
Transactions over phone
Forensics
Caller identification

16. Cepstral feature approach
Silence Removal
Cepstrum Coefficients
Cepstral Normalization
Long time average
Polynomial Function
Expansion
Dynamic Time Warping
Distance Computation
Reference Template
Preprocessing
Feature Extraction
Speaker model
Matching

17. Smoothened Signal Spectrum
Vocal Tract modeling
Speech signal
Signal Spectrum
Smoothened Signal Spectrum

18. Speaker Model F1 = [a1…a10,b1…b10] F2 = [a1…a10,b1…b10] …………….
FN = [a1…a10,b1…b10]
…………….

19. Signature Verification
Online Signature verification
Off line Signature Verification

20. Matching – Similarity Measure
Simple Regression Model
Y = (y1 , y2 , …, yn)
X = (x1 , x2 , …, xn)
Similarity by R2 : 91%
R2=
Similarity by R2 : 31%

21. Dynamic Alignment Dynamic alignment
( y2 is matched x2, x3, so we extend it to be two points in Y sequence.)
Dynamic alignment
Similarity = R2
Where (x1i, y1i, v1i) are points in the sequence
And a, b, c are the weights, e.g., 0.5, 0.5, 0.25
DTW warping path in a n-by-m matrix is the path which has min cumulative cost.
The unmarked area is the constrain that path is allowed to go.

22. Iris Recognition Sharbat Gula:The Afghan Girl
Iriscode used to verify the match

23. Iris Recognition
Iris Image
Choosing the bits
Gabor Kernel

24. Collage

25. Hand Geometry

26. Evaluation of Biometric Systems
Technology Evaluation
Compare competing algorithms
All algorithms evaluated on a single database
Repeatable
FVC2002, FRVT2002, SVC2004 etc.
Scenario Evaluation
Overall performance
Each system has its own device but same subjects
Models real world environment
Operational Evaluation
Not easily repeatable
Each system is tested against its own population

27. System Errors FAR/FMR(False Acceptance Ratio)
FRR/FNMR(False Reject Ratio)
FTE(Failure to Enroll)
FTA(Failure to Authenticate)
Genuine (w1)
Impostor
(w2)
Genuine
No error
False Reject
False Accept
Confusion matrix

28. Performance Curves: Score Distribution

29. Performance curves: FAR/FRR

30. Performance curves: ROC

31. State of the art Biometrics State of the art Research Problems
Fingerprint
0.15% FRR at 1% FAR
(FVC 2002)
Fingerprint Enhancement
Partial fingerprint matching
Face Recognition
10% FRR at 1% FAR
(FRVT 2002)
Improving accuracy
Face alignment variation
Handling lighting variations
Hand Geometry
4% FRR at 0% FAR
(Transport Security Administration Tests)
Developing reliable models
Identification problem
Signature Verification
1.5%(IBM Israel)
Developing offline verification systems
Handling skillful forgeries
Voice Verification
<1% FRR
(Current Research)
Handling channel normalization
User habituation
Text and language independence
Chemical Biometrics
No open testing done yet
Development of sensors
Materials research

32. Thank You