Integrating Information

1
Integrating Information

• Dr. Pushkin Kachroo

2
Integration
Match
Matcher 1
B1
Integration
Decision
B2
Matcher 2
No Match
3
Expanding a Biometric
Multiple Biometrics
Multiple Fingers
Multiple Tokens
One Finger
Multiple Matchers
Multiple Samples
Multiple Sensors
4
Coupling
Match Decision
Match Decision
Integration
Integration
Process 2
Process 1
Process 2
Process 1
Sensor 1
Sensor 2
Sensor 1
Sensor 2
Loosely Coupled
Tightly Coupled
5
Boolean Combinations
Biometric a
Accept/Reject
AND
Biometric b
Biometric a
Accept/Reject
OR
Biometric b
6
Boolean Convenience/Security
Biometric a
Accept/Reject
AND
Biometric b
Biometric a
Accept/Reject
OR
Biometric b



Improve Convenience Lower FRR (OR) Improve
Security Lower FAR (AND)
7
Filtering-Binning

• Filtering using non-biometric, e.g. using last
  name. (P,B)
• Binning using biometric, e.g. some whorl pattern
  (B,B)

• Penetration Rate Ppr The fraction of database
  being matched on average
• Binning Error Rate Pbe

Tradeoff
8
Filtering Error-Negative Identification

• Adding Pn for subject dn to negative
  identification prescribes narrowing down on a
  smaller set of biometric template gt
• Since we are comparing over a smaller set, the
  chance of false positives goes down. However,
  false negatives goes up because you might say the
  person is not in the database (looking at the
  smaller set) when the person might be in the full
  database.

9
Filtering Error-Positive Identification

• The probability that a person is who she/he says
  she/he is equals the probability of a match
  between stored biometric template and a newly
  acquired biometric sample. This match
  probability does not change if additional
  knowledge or possession is supplied.

10
Dynamic Authentication

• Example Conversational biometric.allows for
  natural filtering by asking knowledge information
  during conversation could include possession
  while speaker recognition is taking place.

11
Boolean Score Level Integration
12
Normal Distribution
Accept
Reject
13
Normal Distribution Problems

• Covariance Matrix is assumed to be diagonal okay
  for disparate biometrics but not for similar ones
  e.g. two fingers.
• Gaussian gives non-zero probability to negative
  scores.

14
Distance based
B and Bm are templates from the same biometric
per means person
15
Degenerate Cases
16
ROC based Methods
Mismatch
Match
Compare to
FM
FNM
T