Face Recognition: An Introduction
Face Recognition: An Introduction
An Introduction
Sharad Shivmath
Parmeet Singh
1/24
Can guess whos hand is this?
2/24
FACE
Face is the most common biometric used by humans
3/24
WHAT IS FACE RECOGNITION?
Identifying and verifying a person from a
digital image or a video frame from a video
source.
Digital Image Streaming Video
4/24
MOTIVATION
Security
Airports
Office
Risk: privacy rights
Human Machine Interface
ATMs
Computers
Cell phones
5/24
CHALLENGES
Inter-Class
Similar faces
Twins Father & Son
6/24
CHALLENGES contd..
Intra-Class
Variations between images
of the same face can be
greater than those from
different faces
Illumination
Pose (angle)
Accessories
Expressions
Aging Effects
7/24
8/24
IDENTIFICATION Vs VERIFICATION
Face Identification
Given a face image that belongs to a person in a
database, tell whose image it is.
9/24
IDENTIFICATION Vs VERIFICATION
Face Verification
Given a face image that might not belong to the
database, verify whether it is from the person it is
claimed to be in the database.
10/24
ALGORITHMS
Eigenface
PCA (Principal Component Analysis)
Fisherface
Linear Discriminate Analysis
Elastic Bunch Graph Matching
Hidden Markov model
For Sequence
Neuronal Network
11/24
EXAMPLE OF PCA
Let X(x,y) be a data
set:
x y
2.5 2.4
0.5 0.7
2.2 2.9
1.9 2.2
3.1 3
2.3 2.7
2 1.6
1 1.1
1.5 1.6
1.1 0.9
Original Data
0
0.5
1
1.5
2
2.5
3
3.5
0 1 2 3 4
x axis
y
a
x
i
s
12/24
EXAMPLE OF PCA contd..
Mean Adjusted Data (x,y)
x - /x y - /y
0.69 0.49
-1.31 -1.21
0.39 0.99
0.09 0.29
1.29 1.09
0.49 0.79
0.19 -0.31
-0.81 -0.81
-0.31 -0.31
-0.71 -1.01
Adjusted Data
-1.5
-1
-0.5
0
0.5
1
1.5
-2 -1 0 1 2
x axis
y
a
x
i
s
13/24
EXAMPLE OF PCA contd..
Original Data Vs Adjusted
Adjusted Data
-1.5
-1
-0.5
0
0.5
1
1.5
-2 -1 0 1 2
x axis
y
a
x
i
s
Original Data
0
0.5
1
1.5
2
2.5
3
3.5
0 1 2 3 4
x axis
y
a
x
i
s
14/24
EXAMPLE OF PCA contd..
Eigen Vectors and Eigen
Values
Adjusted Data
Eigen Vector (-0.735, 0678)
with Eigen Value 0.049
Eigen Vector (-0.678,-0.735)
with Eigen Value 1.284
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
-2 -1 0 1 2
15/24
EIGENFACE
Eigenfaces are a set of eigenvectors used in
the computer vision problem of human face
recognition.
It is generated by performing a mathematical
process called Principal Component Analysis
(PCA) on a large set of images depicting
different human faces.
16/24
RECOGNITION PROCESS
Face Space: Training face images are projected onto a feature space, defined by the Eigen vectors.
Face Class: The vectors formed by the linear operation between face space and input image.
17/24
18/24
RECOGNITION PROCESS contd..
Process labeled training images:
Find mean and covariance matrix
Find k principal components (eigenvectors of ) u
1
,u
k
Project each training image x
i
onto subspace spanned by
principal components:
(w
i1
,,w
ik
) = (u
1
T
(x
i
), , u
k
T
(x
i
))
Given novel image x:
Project onto subspace:
(w
1
,,w
k
) = (u
1
T
(x
), , u
k
T
(x
))
Optional: check reconstruction error x x to determine whether
image is really a face
Classify as closest training face in k-dimensional subspace
19/24
RECOGNITION EXAMPLE
Training
images
x
1
,,x
N
20/24
RECOGNITION EXAMPLE contd..
Mean : Top Eigen Vectors :
u1,uk
21/24
RECOGNITION EXAMPLE contd..
Face x in face space coordinates:
=
k
w w ,....,
1
22/24
RECOGNITION EXAMPLE contd..
Face x in face class coordinates:
Threshold Distance Where:
2
2
) (
k k
O O = c
| |
'
,....., ,
2 1
M
T
w w w = O
) ( X X u w
T
k k
=
'
,...., 2 , 1 M k =
} ) ( { max
2
1
, k j k j c
O O = u
23/24
RECOGNITION EXAMPLE contd..
c k k k c
u c c u c < = < } { min &
*
c
u c <
c
u c > Not a face
Unknown face
Face of kth individual
24/24
RECOGNITION EXAMPLE contd..
Reconstruction
x = + w1u1 + w2u2 + w3u3 +
= +
THANK YOU