Biometrics
Topics Biometric identifier classification
Biometric identifier characteristics comparison
Multimodal Biometrics
Biometric Standards
Challenges in Biometrics
Identifiable biometric characteristics
Biological traces DNA, blood, saliva, etc.
Biological (physiological) characteristics fingerprints, eye irises and retinas, hand
palms and geometry, and facial geometry Behavioral characteristics
signature, gait, keystroke dynamics, lip motion, voice
Classification of identifiers Physiological biometric identifiers: fingerprints,
hand geometry, eye patterns (iris and retina), facial features and other physical characteristics.
Behavioral identifiers: voice, signature typing patterns other.
Analyzers based on behavioral identifiers are often less conclusive due to limitations/complex patterns.
Example of banking application
Biometric identifiers
Courtesy of G. Bromba
Biometric Market Share
Comparison of biometric techniques
Palm
Hand vein
Facial Thermogram
Ear print
Retina
Human eye has its own totally unique pattern of blood vessels.
Because of its internal location, the retina is protected from variations caused by exposure to the external environment (unlike fingerprints).
Which Biometric is the Best? Universality (everyone should have this trait) Uniqueness (everyone has a different value) Permanence (should be invariant with time) Collectability (can be measured quantitatively) Performance (achievable recognition accuracy, re
sources required, operating environment) Acceptability (are people willing to accept it?) Circumvention (how easily can it be spoofed?)
Selecting a Biometric
Selecting the ‘right’ biometric is a complicated problem that involves more factors than just accuracy. It depends on cost, error rates, computational speed, acquitability, privacy and easy of use.
Ideal Biometric CharacteristicsThe ideal biometric characteristics have five qualities:
Robust: Unchanging on an individual over time.
Distinctive: Showing great variation over the population.
Available: The entire population should ideally have this measure in multiples.
Accessible: Easy to image using electronic sensors.
Acceptable: People do not object to having this measurement taken on them.
Quantitative measuresQuantitative measures of these five qualities have been
developed.
"Robustness" is measured by the "false non-match rate" (Type I error), the probability that a submitted sample will not match the enrollment image.
"Distinctiveness" is measured by the "false match rate" (Type II error), the probability that a submitted sample will match the enrollment image of another user.
"Availability" is measured by the "failure to enroll" rate, the probability that a user will not be able to supply a readable measure to the system upon enrollment.
"Accessibility" can be quantified by the "throughput rate" of the system, the number of individuals that can be processed in a unit time, such as a minute or an hour.
"Acceptability" is measured by polling the device users.
Biometric System Goals A biometric system can be designed to test one of only two
possible hypotheses:
The submitted samples are from an individual known to the system
The submitted samples are from an individual not known to the system
Applications to test the first hypothesis are called "positive identification" systems while applications testing the latter are called "negative identification" systems.
Types of Biometrics Overt Versus Covert: The first partition is "overt/covert". If the user is
aware that a biometric identifier is being measured, the user is overt. If unaware, the use is covert. Almost all conceivable access control and non-forensic applications are overt. Forensic applications can be covert.
Habituated Versus Non-Habituated: This applies to the intended users of the application. Users presenting a biometric trait on a daily basis can be considered habituated after a short period of time. Users who have not presented the trait recently can be considered "non-habituated".
Attended Versus Non-Attended: This partition refers to whether the use of the biometric device during operation will be observed and guided by system management.
Open Versus Closed: If a system is to be open, data collection, compression and format standards are required. A closed system can operate perfectly well on completely proprietary formats.
Generic Biometric System
A generic biometric system.
Multimodal Biometrics
Multimodal Biometric system is a system that uses more than one independent or weakly correlated biometric identifier taken from an individual (e.g., fingerprint and face of the same person, or fingerprints from two different fingers of a person)
Multi-modal Systems: Fusion Early integration or sensor fusion
Integration is performed on the feature level Classification is done on the combined
feature vector
Multi-modal Systems: Fusion
Late integration or decision fusion Each modality is first pre-classified
independently The final classification is based on the
fusion of the outputs of the different modalities
Multimodal biometrics systems Multimodal biometrics systems improve
performance A combination in a verification system
improves system accuracy A combination in an identification system
improves system speed as well as accuracy A combination of uncorrelated modalities (e.g.
fingerprint and face, two fingers of a person, etc.) is expected to result in a better improvement in performance than a combination of correlated modalities (e.g. different fingerprint matchers)
Other work: classification FBI Fingerprint card (includes information o
n gender, ethnicity, height, weight, eye color and hair color)
Wayman (1997) proposed filtering large biometric databases based on gender and age
Givens et al. (2003) and Newham (1995) showed that age, gender and ethnicity can affect the performance of a biometric system
International Standards Bodies
Application Programming Interface (API) Biometrics is the automated use of
physiological or behavioral characteristics to determine or verify an identity
Standards for interfaces and methods for performance evaluation are needed
Biometric Authentication Systems Layers of interaction with biometric authentication
systems
Scope Standardization of generic biometric technologies
to support interoperability and data interchange between applications and systems
Included: common file formats, application programming interfaces (APIs), biometric templates, template protection techniques, related application/implementation profiles, methodologies for conformity
Basic Standards BioAPI – The most popular API in the biome
trics area CBEFF – Common Biometric Exchange File
Format ANSI X9.84-2003 – Biometric Information M
anagement and Security for the Financial Services Industry
ISO/IEC 19794 – Biometric Data Interchange Formats
Challenges in Biometrics Large number of classes (~ 6 billion faces) Large intra-class variability Small inter-class variability Segmentation Noisy and distorted images Population coverage & scalability System performance (error rate, speed, cost) Attacks on the biometric systemEvery biometric characteristic has some
limitations
Threats to Biometrics
The Modern Burglar
Matsumoto’s Technique
Only a few dollars’ worth of materials
Making the Actual Clone
You can place the “gummy finger” over your real finger. Observers aren’t likely to detect it when you use it on a fingerprint reader.
Don’t try this at home! (Matsumoto)
Summary There is wide variety of biometric identifiers
that posses different characteristics Each biometric system should take into
account the end goal of application Multi-biometrics improve performance of
individual matchers and is active topic of current biometric research
Biometric standards are being developed, while biometric reliability is still a concern
Reference and Links Signal Processing Institute, Swiss Federal Ins
titute of Technology http://scgwww.epfl.ch/ Biometric Systems Lab, University of Bologn
ahttp://bias.csr.unibo.it/research/biolab/
www.sciencedierect.com Textbooks 1 and 2 CPSC 601.20