You are on page 1of 20

Score Normalization in

Multimodal Biometric Systems

Karthik Nandakumar and Anil K. Jain


Michigan State University, East Lansing, MI

Arun A. Ross
West Virginia University, Morgantown, WV

http://biometrics.cse.mse.edu
Multimodal Biometric Systems
• Multiple sources of biometric information are integrated to
enhance matching performance

• Increases population coverage by reducing failure to enroll rate

• Anti-spoofing; difficult to spoof multiple traits simultaneously

Fingerprint Face Hand geometry Iris


Fusion in Multimodal Biometrics
Biometric Fusion

Prior to matching After matching

Sensor Feature Dynamic Classifier


Level Level Classifier Fusion
Selection
Raw Data Feature
Sets

Confidence Rank Abstract


Level Level Level
Matching Class Class
Scores Ranks Labels
Fusion at the Matching Score Level
• Fusion at the matching score level offers the best tradeoff
in terms of information content and ease in fusion

Classification Approach Combination Approach

• Experiments indicate that the combination approach


performs better than the classification approach*
*A. Ross, A.K. Jain, “Information Fusion in Biometrics”, Pattern Recognition Letters, Sep. 2003
Fusion Rules*
• Problem: Classify input pattern Z into one of m possible
classes (c1, …,cm) based on evidence provided by R classifiers

• Let xi be the feature vector for the ith classifier derived from Z;
xi´s are independent

• Assign Z → cj, if g(cj) ≥ g(ck), 1 ≤ k ≤ m, k ≠ j


R
• Product rule: g(cr) = ∏ P (c
i =1
r | xi )
R

• Sum rule: g(cr) = ∑ P (c


i =1
r | xi )

• Max rule: g(cr) = max


i
P (cr | xi )

• Min rule: g(cr) = min


i
P(cr | xi )

*Kittler et al., “On Combining Classifiers”, IEEE PAMI, March 1998


Computing the Posteriori Probability
• To make use of the fusion rules, we need to compute P(cj|xi)

• At the matching score level, we have only the score sij (not xi)

• Verlinde et al. proposed that

sij = f(P(cj|xi)) + η(xi)

– f is a monotonic function

– η is the error introduced by the biometric system due to


problems in acquisition and feature extraction processes

• To estimate η, the biometric system should output a matching


score along with a confidence measure on that score
indicating the quality of the input feature vector xi

• If η is known, we can estimate P(cj|xi) from sij using non-


parametric density estimation methods (e.g., Parzen window)
Score Normalization
• If η is unknown, the errors in the estimation of P(cj|xi) will be
very large; hence, it is better to combine the scores directly

• Combination of scores has the following problems:

• Non-homogeneous scores: distance or similarity


ƒ Ranges may be different; e.g., [0,100] or [0,1000]
ƒ Distributions may be different
• Modify the location and scale parameters of score
distributions of individual matchers to transform the scores
into a common domain

• Robustness: Should not be affected by the outliers

• Efficiency: Estimated parameters of the score distribution


should be close to the true values
Score Normalization Techniques
• Min-max normalization: Given matching scores {sk}, k=1,2,..,n
the normalized scores are given by:

s − min{s k }
s'=
max{s k } − min{s k }

• Decimal scaling: Used when scores of different matchers differ by


a logarithmic factor; e.g., one matcher has scores in the range
[0,1] and the other matcher has scores in the range [0, 1000]

s
s' = n
,
10
n = log 10 max{s k }
Score Normalization Techniques
• Z-score:
s−µ
s' =
σ
• Median and Median Absolute Deviation (MAD):
( s − median )
s' =
MAD
MAD = median ( { s k } − median )

• Double Sigmoid function:


1
s' =
  s −t 
1 + exp  −2  
  r 
r = r1, if s < t
r = r2, otherwise
Score Normalization Techniques
• Tanh estimators:

  ( s − µ GH )  
s ' = 0.5  tanh  0.01  + 1 ,
  σ GH  

where µGH and σGH are the mean


and standard deviation estimates
of the genuine score distribution
as given by Hampel estimators*

• Min-max, Z-score, and Tanh normalization schemes are efficient


• Median, Double Sigmoid, and Tanh methods are robust

*Hampel et al., Robust Statistics: The Approach Based on Influence Functions, 1986
Experimental Setup

Face

Hand-geometry

Fingerprint

Database of 100 users with three modalities (5 samples/user)


Experimental Setup
Feature df Sf´
Extraction Normalization
and Matching

Eigenfaces
Fusion Decision
Rule
Feature
Extraction Normalization
and Matching sp sp´

Minutiae

Feature dh sh´
Extraction Normalization
and Matching

14-D hand feature vector Fusion Rules


Sum rule: mean(sf´, sp´, sh´)
Min rule: min(sf´, sp´, sh´)
Max rule: max(sf´, sp´, sh´)
Distribution of Matching Scores

Face Fingerprint

Hand-geometry
Performance of Individual Modalities
Normalization followed by Sum Rule
Sensitivity of Min-max to Outliers
Sensitivity of Z-score to Outliers
Sensitivity of Tanh to Outliers
Summary
• The effects of different score normalization techniques in a
multimodal biometric system have been studied

• Min-max, Z-score, and Tanh normalization techniques followed


by sum rule resulted in the best recognition performance

• Min-max and Z-score methods are efficient but sensitive to


outliers in the training data; Tanh is both efficient and robust

• If the parameters of the matching score distribution are known,


simple methods like Min-max and Z-score would suffice

• If the parameters are to be estimated from noisy training data,


one should choose robust methods like Tanh normalization

• If the system can provide both score and some confidence


measure on that score, non-parametric density estimation
methods can be used to estimate the posteriori probability
Efficiency of a Normalization Scheme

Before Normalization After Sigmoid Normalization

After Min-max Normalization After Tanh Normalization

You might also like