EE Dept. IIT Delhi 1

Face Recognition Using Fuzzy Fisherface Classifier

Presenters:Nilesh PadwalVivek K.Rajat Rastogi

EE Dept. IIT Delhi 2

Contents

Fuzzy Fisherface AppproachAlgorithmFlowchartYale DatabaseORL DatabaseComparison of Recognition RatesConclusionReferences

EE Dept. IIT Delhi 3

Fuzzy Fisherface Approach

More sophisticated usage of class assignment of patterns (faces)

Classification results affect the within-class and between-class scatter matrices

EE Dept. IIT Delhi 4

The Computations Of Membership Degrees

Compute the membership grade to class for pattern ,

0.51 0.49( / ) if same as the label of the pattern

0.49( / ) if same as the label of the patternij

ij

n k i jth

n k i jth

+ =⎧⎪⎨ ≠⎪⎩

i jth

where is number of the neighbors of theijnjth data that belong to ith class

ijµ

EE Dept. IIT Delhi 5

AlgorithmResults of FKNN are used in computations of mean value and scatter covariance matrices,Mean vector of each class

The between class and within class fuzzy scatter matrices are respectively,

~1

1

N

i j jj

i N

i jj

Xm

µ

µ

=

=

=∑

∑

~ ~

1

~ ~

1 1

( )( )

( )( )i

k i

cT

i iFB ii

c cT

i iFW k k FWi x C i

S N m m m m

S x m x m S

=

= ∈ =

= − −

= − − =

∑

∑ ∑ ∑

EE Dept. IIT Delhi 6

AlgorithmThe optimal fuzzy projection WF-FLD and feature vector transformed by fuzzy fisherfacemethod are given by

~

arg max

( )

TFB

F FLD TWFB

T T Ti F FLD i F FLD i

W S WW

W S W

v W X W E z z

−

− −

=

= = −

EE Dept. IIT Delhi 7

Flowchart

courtesy:Source [1]

EE Dept. IIT Delhi 8

Yale DatabaseTotal Images- 165, total Classes- 15 (11 Images For Each Class) One Image for each configuration: Center-light, glasses/no glasses, happy, normal, left/right light, sad, sleepy, surprised, wink.

EE Dept. IIT Delhi 9

Yale Database

Mugshots were acquired using digicam,Each image was digitized and presented by a 243 X 320 pixel array

EE Dept. IIT Delhi 10

ORL DatabaseTotal Images: 400, total classes: 40 (10 Images for each class)Mugshots were acquired using DigiCam, varying facial detailsEach image was digitized and presented by a 112 X 92 pixel array

EE Dept. IIT Delhi 11

Comparison of Mean Recognition Rates (Yale)

93.8791.9470.85Case 3

(5-training,6-testing)

96.2496.0471.66Case 2

(7-training,4-testing)

94.293.472.44Case 1

(6-training,5-testing)

Fuzzy Fisherface(Fuzzy+PCA+LDA) (%)

Fisherface(PCA+LDA)(%)

Eigenface (PCA)(%)

EE Dept. IIT Delhi 12

Comparison of Mean For Recognition Rates (ORL)

93.5093.3886.94Case 3

(4-training,6-testing)

95.594.7591.13Case 2

(5-training,5-testing)

97.1295.5990.94Case 1

(6-training,4-testing)

Fuzzy Fisherface(Fuzzy+PCA+LDA)

(%)

Fisherface(PCA+LDA)(%)

Eigenface (PCA)(%)

EE Dept. IIT Delhi 13

Comparison of Recognition Rate For Yale Database

0

20

40

60

80

100

Case 1 Case 2 Case 3

EigenfaceFisherfaceFuzzy-Fisherface

EE Dept. IIT Delhi 14

Comparison of Recognition Rate For ORL Database

80828486889092949698

Case 1 Case 2 Case 3

EigenfaceFisherfaceFuzzy-Fisherface

EE Dept. IIT Delhi 15

Confusion Matrix (Yale)Case 2(7/4)

Fisherface Fuzzy Fisherface

EE Dept. IIT Delhi 16

Comparison

Fisherface

Fuzzy Fisherface

Eigenface

Input Image Matched Image

EE Dept. IIT Delhi 17

Comparison

Fisherface

Fuzzy Fisherface

Eigenface

Input Image Matched Image

EE Dept. IIT Delhi 18

Comparison

Fisherface

Fuzzy Fisherface

Eigenface

Input Image Matched Image

EE Dept. IIT Delhi 19

ConclusionFuzzy fisherface approach outperforms the other two methods for the datasets considered.

Sensitive to variations in illumination and facial expression reduced substantially.

EE Dept. IIT Delhi 20

ReferencesKeun-Chang Kwak, Witold Pedrycz : Face Recognition Using Fuzzy Fisherface Classifier, Journal of Pattern Recognition 38(2005),1717-1732Turk, M., Pentland, A.: Eignefaces for Recognition. Journal of Cognitive Neuroscience, Vol.3, (1991) 72-86Turk, M., Pentland, A.: Face Recognition Using Eignefaces. In Proc. IEEE Conf. On Computer Vision and Pattern Recognition. (1991) 586-591Belhumeur, P., Hespanha, J., Kriegman, D.: Eigenfacesvs. Fisherfaces: Face Recognition using class specific linear projection. In Proc. ECCV, (1996) 45-58Yale Face Database, http://cvc.yale.edu/projects/yalefaces/yalefaces.htmlORL Face Database, http://www.uk.research.att.com/facedatabase.html

EE Dept. IIT Delhi 21

Thank You