Chapter 8 Selecting a Biometric

1
Chapter 8 Selecting a Biometric

• Presentation by
• Phani Dogiparthi

2
Contents

• Biometric Attributes
• Properties of Applications
• Evaluating Biometric Options
• Cost and Affordability of Biometric
• Pros and Cons of Various Biometrics
• Myths and Misrepresentations

3
Factors Associated with choosing a Biometric
4
Comparison of the Various Attributes
5
Application Requisites

• What is the protected asset
• Who are the authorized users and operators
• Who are the adversaries of the application
• What can be the costs of security violations
• What is the actual cost of the security and
• What is the added cost of inconvenience?

6
Categories of Applications

• Cooperative and Non-cooperative
• Overt and Covert
• Habituated and Non-habituated
• Attended and Non-attended
• Standard Environment
• Public and Private
• Open and Closed

7
Importance Weightings
8
Drawbacks of various Biometrics
9
Mismatch Calculation
10
Zephyr Charts
11
Characteristics Weighting
Effortless Iris gt Face gt Finger gt voice
Non-intrusive Iris gt Voice gt Face gt Finger
Accurate Iris gt Finger gt Face Voice
Inexpensive Voice gt Finger gt Face gt Iris
12
Affordability and Cost

• Total Cost of ownership
• Verification versus Identification
• Single and Multiple Authentication protocols
• Training
• -gt System Training
• -gt Enrollment
• Continued Enrollment
• Failure to Enroll
• User Education
• Supervisory Labor
• Maintenance Labor

13
Positives and Negatives of the Biometrics
14
Fingerprint - Pros

• Long Tradition of usage of fingerprint as
  immutable identification in law enforcement
• Existing of large databases of fingerprints
  comprising largely of criminals
• Forensic investigation uses fingerprint
• Sampling of the fingerprint is easy
• Conversion of fingerprints into digital images is
  getting easier, better and cheaper
• Various state departments of motor vehicles are
  working to establish a fingerprint biometric
  system for drivers license and record data

15
Fingerprint-Cons

• Finger print suffer from poor public reputation
• Social stigma such that it is related to
  illiterate
• Variation occurs due to age, cut or due to
  working conditions
• Sampling variations due to improper usage of the
  apparatus
• Rare situations like people without fingers.

16
Face - Pros

• Photos are widely used and usage is public
  accepted
• Face recognition systems are least intrusive.
• In theory this biometric works with image sources
• In theory can be applied to identify criminals in
  crowd
• It is good biometric for small scale verification
  purpose

17
Face - Cons

• Face authentication requires images taken in a
  good controlled lighted environment
• Presently it is a poor biometric for usage in
  pure identification protocol
• Disguise - a major obstacle
• Usage may hamper due to social stigma

18
Voice - Pros

• Like face voice is natural biometric
• Voice sampling can also be obtained unobtrusively
• Great public acceptance of this biometric
• Voice uses inexpensive hardware and can be
  established on current communications systems
• Voice allows incremental protocols
• Voice may achieve high accuracy and flexibility
  when combined with knowledge verification
• Voice allows continuous identity checking

19
Voice - Cons

• Imitation by possible in voice
• Possibility of creating nonexistent identities
  with text-to-speech technology
• Depends on high quality capturing of audio signal
• Possibility in rare cases where voice is lost

20
Iris - pros

• Widely believed to be the most accurate biometric
• Sample acquisition is unobtrusive no physical
  contact
• Received only little negative press
• Present claim is that iris does not involve high
  training costs

21
Iris - Cons

• There are no or few legacy iris databases
• Sampling the iris patterns requires much user
  cooperation or complex, expensive input devices
• Iris authentication is hampered by vision aid
• This biometric is not left as an evidence in the
  scene of crime, so not used for forensic
  applications
• Cases like missing eye

22
Hand - Pros

• Certain amount of public acceptance
• Termed as do-it-yourself operation
• Existing of at least one scenario evaluation of
  hand geometry
• Can be used in place in stronger biometric where
  stronger might induce public fear

23
Hand - Cons

• Physical contact induces public hygiene
• Not very distinctive
• Existence of only one scenario application rises
  doubts about Hand as Biometric
• Mutation

24
Signature - Pros

• Man-made biometric- extensive study in forgery
• Used to identify impersonation even at enrollment
  stage
• Training is fast
• Has fast response and low storage requirements
• Independent of the native language of speaker
• Combination of information and biometric
• Unaffected by high compression rates

25
Signature - Cons

• Excessive usage of signature rises perception
  that it is not secure enough for protection
• Scaling properties and identifier standards are
  hard to set
• A five-dimensional pen is needed to get desired
  accuracy which makes hardware costly
• Medical conditions prevent people from writing
  consistently

26
Myths and Misrepresentations

• Biometric X is best for all applications
• Biometric X is unique for each individual
• A single number quantifies system accuracy
• System is plug and play
• Real accuracy performance can be predicted
• The vendors reporting best FAR and FRR has the
  most accurate system
• Multiple biometrics outperform single biometrics
• Our biometric system does not use a decision
  threshold

27

• Our feature extractor can be used with any match
  engine
• Large templates mean better accuracy
• Face recognition prevents terrorism
• Biometrics means 100 percent security
• Biometric systems invade our privacy
• Biometric sensors are unhygienic or otherwise
  harmful

28
Thank you