I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
rin
g
(
I
J
E
CE
)
Vo
l.
7
,
No
.
3
,
J
u
n
e
201
7
,
p
p
.
1594
~
1
6
0
1
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
7
i
3
.
p
p
1
5
9
4
-
1601
1594
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
jo
u
r
n
a
l.c
o
m/o
n
lin
e/in
d
ex
.
p
h
p
/I
JE
C
E
Face Recog
nition
Using
Co
m
plete
d
Lo
ca
l Terna
ry
P
a
tt
ern
(C
LTP
)
Tex
ture
Descriptor
T
a
ha
H
.
Ra
s
s
e
m
1
,
Na
s
rin M
.
M
a
k
bo
l
2
,
Sa
m
Yin
Yee
3
1,
3
F
a
c
u
l
ty
o
f
Co
m
p
u
ter S
y
ste
m
s a
n
d
S
o
f
tw
a
r
e
En
g
in
e
e
rin
g
,
Un
iv
e
rsiti
M
a
l
a
y
sia
P
a
h
a
n
g
,
Ku
a
n
tan
,
M
a
la
y
sia
1
IBM
Ce
n
tre
o
f
Ex
c
e
ll
e
n
c
e
,
Un
iv
e
rsiti
M
a
l
a
y
sia
P
a
h
a
n
g
,
K
u
a
n
tan
,
M
a
la
y
sia
2
S
c
h
o
o
l
o
f
El
e
c
tri
c
a
l
a
n
d
El
e
c
tro
n
ic E
n
g
in
e
e
ri
n
g
,
Un
iv
e
rsiti
S
a
in
s
M
a
la
y
sia
,
P
e
n
a
n
g
,
M
a
lay
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
2
1
,
2
0
1
7
R
ev
i
s
ed
A
p
r
2
1
,
2
0
1
7
A
cc
ep
ted
Ma
y
6
,
2
0
1
7
No
w
a
d
a
y
s,
f
a
c
e
re
c
o
g
n
it
io
n
b
e
c
o
m
e
s
o
n
e
o
f
th
e
i
m
p
o
rtan
t
to
p
i
c
s
in
th
e
c
o
m
p
u
ter
v
isio
n
a
n
d
im
a
g
e
p
ro
c
e
ss
in
g
a
re
a
.
T
h
is
is
d
u
e
to
it
s
i
m
p
o
rtan
c
e
w
h
e
re
c
a
n
b
e
u
se
d
in
m
a
n
y
a
p
p
li
c
a
ti
o
n
s.
T
h
e
m
a
in
k
e
y
in
th
e
fa
c
e
re
c
o
g
n
it
io
n
is
h
o
w
to
e
x
trac
t
d
isti
n
g
u
ish
a
b
le
f
e
a
tu
re
s
f
ro
m
th
e
im
a
g
e
to
p
e
rf
o
r
m
h
ig
h
re
c
o
g
n
it
io
n
a
c
c
u
ra
c
y
.
L
o
c
a
l
b
in
a
r
y
p
a
tt
e
rn
(L
BP
)
a
n
d
m
a
n
y
o
f
it
s
v
a
rian
ts
u
se
d
a
s
te
x
tu
re
f
e
a
t
u
re
s
in
m
a
n
y
o
f
f
a
c
e
r
e
c
o
g
n
it
io
n
s
y
ste
m
s.
A
lt
h
o
u
g
h
L
BP
p
e
rf
o
rm
e
d
w
e
ll
i
n
m
a
n
y
f
ield
s,
it
is
se
n
siti
v
e
to
n
o
ise
,
a
n
d
d
if
fe
re
n
t
p
a
tt
e
rn
s
o
f
L
BP
m
a
y
c
las
sify
in
to
t
h
e
sa
m
e
c
las
s
th
a
t
re
d
u
c
e
s
it
s
d
isc
rim
in
a
ti
n
g
p
ro
p
e
rty
.
Co
m
p
let
e
d
L
o
c
a
l
T
e
rn
a
r
y
P
a
tt
e
rn
(CL
T
P
)
is
o
n
e
o
f
th
e
n
e
w
p
ro
p
o
se
d
tex
tu
re
f
e
a
tu
re
s
to
o
v
e
rc
o
m
e
th
e
d
ra
w
b
a
c
k
s
o
f
th
e
L
BP
.
T
h
e
CLT
P
o
u
tp
e
rf
o
rm
e
d
L
BP
a
n
d
so
m
e
o
f
it
s
v
a
rian
ts
in
m
a
n
y
f
ie
ld
s
su
c
h
a
s
tex
tu
re
,
sc
e
n
e
,
a
n
d
e
v
e
n
t
i
m
a
g
e
c
las
si
f
ica
ti
o
n
.
In
th
is
stu
d
y
,
w
e
stu
d
y
a
n
d
in
v
e
stig
a
te
th
e
p
e
rf
o
r
m
a
n
c
e
o
f
C
LT
P
o
p
e
ra
to
r
f
o
r
f
a
c
e
re
c
o
g
n
it
io
n
tas
k
.
T
h
e
Ja
p
a
n
e
se
F
e
m
a
le
F
a
c
ial
Ex
p
re
ss
io
n
(JA
F
F
E),
a
n
d
F
EI
f
a
c
e
d
a
tab
a
se
s
a
r
e
u
se
d
in
t
h
e
e
x
p
e
ri
m
e
n
ts.
In
th
e
e
x
p
e
ri
m
e
n
tal
re
su
lt
s,
CLT
P
o
u
t
p
e
rf
o
r
m
e
d
so
m
e
p
re
v
io
u
s
tex
tu
re
d
e
sc
rip
to
r
s
a
n
d
a
c
h
iev
e
s
h
ig
h
e
r
c
las
si
f
ica
ti
o
n
ra
te
f
o
r
f
a
c
e
re
c
o
g
n
it
io
n
tas
k
w
h
ich
h
a
s
re
a
c
h
e
d
u
p
9
9
.
3
8
%
a
n
d
8
5
.
2
2
%
in
JA
F
F
E
a
n
d
F
EI
,
re
sp
e
c
ti
v
e
ly
.
K
ey
w
o
r
d
:
Face
r
ec
o
g
n
itio
n
C
o
m
p
leted
l
o
ca
l
b
in
ar
y
p
atter
n
(
C
L
B
P
)
C
o
m
p
leted
l
o
ca
l
t
er
n
ar
y
p
atter
n
(
C
L
T
P
)
Face
d
ataset
I
m
ag
e
c
la
s
s
i
f
icatio
n
Co
p
y
rig
h
t
©
2
0
1
7
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
T
ah
a
H.
R
ass
e
m
,
Facu
lt
y
o
f
C
o
m
p
u
ter
S
y
s
te
m
s
an
d
So
f
t
w
ar
e
E
n
g
i
n
ee
r
in
g
,
Un
i
v
er
s
iti Ma
la
y
s
ia
P
ah
an
g
,
Ku
a
n
tan
,
Ma
la
y
s
ia
.
E
m
ail:
ta
h
ah
u
s
s
ei
n
@
u
m
p
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
Au
to
m
a
tic
f
ac
e
r
ec
o
g
n
itio
n
h
as
b
ee
n
a
f
o
cu
s
r
esear
c
h
to
p
ic
in
p
ast
f
e
w
d
ec
ad
es.
T
h
is
i
s
d
u
e
to
t
h
e
ad
v
an
ta
g
es
o
f
f
ac
e
r
ec
o
g
n
itio
n
a
n
d
th
e
p
o
ten
tial
n
ee
d
f
o
r
h
ig
h
s
ec
u
r
it
y
i
n
co
m
m
er
cial
a
n
d
la
w
e
n
f
o
r
ce
m
e
n
t
ap
p
licatio
n
s
.
T
o
d
ay
,
t
h
e
f
ac
e
i
s
th
e
m
o
s
t
co
m
m
o
n
b
i
o
m
etr
ic
u
s
ed
b
y
h
u
m
a
n
s
.
Face
r
ec
o
g
n
itio
n
is
a
tas
k
t
h
at
h
u
m
a
n
s
p
er
f
o
r
m
r
o
u
ti
n
el
y
a
n
d
ef
f
o
r
tles
s
l
y
i
n
o
u
r
d
ail
y
liv
es.
Hu
m
an
s
ar
e
v
er
y
g
o
o
d
at
r
ec
o
g
n
i
s
in
g
f
ac
e
s
an
d
co
m
p
le
x
p
atter
n
s
.
H
u
m
an
s
o
f
ten
u
s
e
f
ac
es
to
r
ec
o
g
n
i
s
e
in
d
i
v
id
u
al
s
a
n
d
ad
v
an
ce
m
en
ts
in
co
m
p
u
tin
g
ca
p
ab
ilit
y
o
v
er
th
e
p
ast
f
e
w
d
ec
ad
es
n
o
w
en
ab
le
s
i
m
i
lar
r
ec
o
g
n
itio
n
s
a
u
to
m
atica
ll
y
.
Face
R
ec
o
g
n
itio
n
m
o
r
e
ea
s
il
y
to
ap
p
l
y
in
s
tead
o
f
u
s
i
n
g
f
in
g
er
p
r
in
t
d
etec
tio
n
,
ir
is
r
ec
o
g
n
itio
n
,
s
ig
n
at
u
r
e
r
ec
o
g
n
itio
n
etc.
b
ec
au
s
e
th
is
s
o
r
t
o
f
b
io
m
etr
ic
also
h
as
s
o
m
e
d
is
ad
v
an
tag
e
s
f
o
r
n
o
n
-
co
llab
o
r
ativ
e
in
d
i
v
id
u
als.
Ma
n
y
f
ea
t
u
r
es
h
av
e
b
ee
n
p
r
o
p
o
s
ed
an
d
u
s
e
d
to
d
esig
n
f
ac
e
r
ec
o
g
n
itio
n
s
y
s
te
m
s
s
u
ch
a
s
P
r
in
cip
al
C
o
m
p
o
n
en
t
An
al
y
s
i
s
(
P
C
A
)
[
1
]
,
L
in
ea
r
Di
s
cr
i
m
i
n
a
n
t
An
al
y
s
is
(
L
D
A
)
[
2
]
,
I
n
d
ep
en
d
en
t
C
o
m
p
o
n
en
t
A
n
a
l
y
s
is
(
I
C
A
)
[
3
]
,
L
o
ca
l
B
in
ar
y
P
atter
n
(
L
B
P
)
[
4
]
,
etc.
L
P
B
is
o
n
e
o
f
th
e
f
a
m
o
u
s
te
x
t
u
r
e
d
escr
ip
to
r
p
r
o
p
o
s
ed
in
2
0
0
2
b
y
Oj
ala
[
5
]
f
o
r
tex
tu
r
e
class
i
f
icatio
n
.
L
B
P
d
escr
ip
to
r
an
d
m
a
n
y
o
f
i
t
s
v
ar
ia
n
t
s
ar
e
u
s
ed
f
o
r
d
if
f
er
en
t c
o
m
p
u
ter
v
i
s
io
n
tas
k
s
,
s
u
c
h
as o
b
j
ec
t a
n
d
s
ce
n
e
r
ec
o
g
n
itio
n
[
6
]
,
h
u
m
a
n
d
ete
ctio
n
s
[
7
]
,
o
b
j
ec
t
tr
ac
k
in
g
[
8
]
,
an
d
f
ac
e
r
ec
o
g
n
it
io
n
[
4
]
,
[
9
]
-
[
1
0
]
.
T
h
e
L
B
P
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
F
a
ce
R
ec
o
g
n
itio
n
Usi
n
g
C
o
m
p
leted
Lo
ca
l Tern
a
r
y
P
a
tter
n
(
C
LT
P
)
Textu
r
e
De
s
crip
to
r
(
Ta
h
a
H.
R
a
s
s
em)
1595
h
is
to
g
r
a
m
is
co
m
p
u
ted
o
v
er
u
s
er
-
d
ef
i
n
ed
p
atter
n
s
(
g
r
id
o
f
c
ells
)
.
T
h
e
f
ir
s
t
s
tep
is
th
e
t
h
r
es
h
o
ld
in
g
s
tep
w
h
er
e
th
e
ce
n
tr
e
o
f
th
e
p
atter
n
is
co
m
p
ar
ed
w
it
h
it
s
p
ix
el
n
ei
g
h
b
o
u
r
h
o
o
d
to
co
n
v
er
t
t
h
eir
v
al
u
es
to
b
in
ar
y
v
al
u
e
s
(
0
o
r
1
)
.
T
h
is
s
tep
ai
m
s
to
f
in
d
th
e
b
i
n
ar
y
d
if
f
er
en
ce
s
.
T
h
e
n
ex
t
s
tep
is
t
h
e
e
n
co
d
in
g
s
tep
,
w
h
ic
h
e
n
co
d
es
t
h
e
b
in
ar
y
n
u
m
b
er
o
f
ea
c
h
p
atter
n
an
d
co
n
v
er
t
s
it
to
t
h
e
eq
u
i
v
al
en
t
d
ec
i
m
a
l
n
u
m
b
er
th
a
t
c
h
ar
a
cter
is
es
a
s
tr
u
ct
u
r
al
p
atter
n
.
T
h
e
L
B
P
is
o
n
e
d
is
tr
ib
u
tio
n
-
b
ased
d
escr
ip
to
r
b
e
ca
u
s
e
all
th
e
p
atter
n
s
’
d
ec
i
m
al
v
al
u
es
ar
e
th
e
n
r
ep
r
esen
ted
as
a
h
is
to
g
r
a
m
.
I
n
ad
d
itio
n
to
t
h
at,
th
e
L
B
P
is
co
m
p
u
ta
tio
n
all
y
s
i
m
p
le,
s
h
o
w
in
g
g
o
o
d
p
er
f
o
r
m
a
n
ce
an
d
ex
ce
l
len
t
r
e
s
u
lt
s
in
te
x
tu
r
e
clas
s
i
f
icatio
n
.
E
x
a
m
p
les
o
f
L
B
P
v
ar
ian
ts
ar
e
L
o
ca
l
T
e
r
n
ar
y
P
atter
n
(
L
T
P
)
[
1
1
]
,
C
o
m
p
lete
d
L
B
P
(
C
L
B
P
)
[
1
2
]
,
an
d
C
o
m
p
leted
L
o
ca
l
B
in
ar
y
C
o
u
n
t
(
C
L
B
C
)
[
1
3
]
.
I
n
[
4
]
,
th
e
L
B
P
is
u
s
ed
f
o
r
f
ac
e
d
esc
r
ip
tio
n
.
T
h
e
f
ac
e
i
s
d
i
v
id
in
g
i
n
to
s
e
v
er
al
b
lo
ck
s
,
L
B
P
as
a
lo
ca
l
d
escr
ip
to
r
is
ex
tr
ac
ted
f
r
o
m
ea
ch
b
lo
ck
,
an
d
th
en
all
b
lo
ck
s
d
escr
ip
to
r
s
ar
e
co
m
b
in
ed
as
a
g
lo
b
al
d
escr
ip
to
r
.
T
h
e
n
ea
r
est
n
eig
h
b
o
u
r
alg
o
r
it
h
m
i
s
u
s
ed
as
a
class
i
f
ier
.
I
n
[
1
4
]
,
th
e
au
th
o
r
s
h
av
e
p
r
o
p
o
s
ed
a
f
ac
e
r
ec
o
g
n
itio
n
s
y
s
te
m
b
ased
o
n
C
L
B
P
.
T
h
ey
u
s
ed
Mu
lti
-
C
las
s
S
u
p
p
o
r
t
Vec
to
r
Ma
ch
i
n
e
as
a
cla
s
s
i
f
ier
to
a
ch
iev
e
a
h
i
g
h
f
ac
e
r
ec
o
g
n
itio
n
ac
cu
r
ac
y
.
T
h
e
co
m
b
in
at
io
n
o
f
C
L
B
P
an
d
s
p
ar
s
e
r
ep
r
esen
tatio
n
is
u
s
ed
in
to
p
r
o
p
o
s
e
a
n
e
w
f
ac
e
r
ec
o
g
n
itio
n
s
y
s
te
m
i
n
[
1
5
]
.
A
b
r
ief
e
v
al
u
atio
n
o
f
d
if
f
er
e
n
t
f
a
ce
r
ec
o
g
n
i
tio
n
s
y
s
te
m
s
b
ased
o
n
L
B
P
an
d
d
if
f
er
e
n
t v
ar
ia
n
t
s
o
f
L
B
P
tex
tu
r
e
d
escr
ip
to
r
s
h
ad
b
ee
n
d
o
n
e
i
n
[
1
6
]
.
A
lt
h
o
u
g
h
t
h
e
L
B
P
s
h
o
w
ed
a
g
o
o
d
r
esp
o
n
s
e
an
d
p
er
f
o
r
m
a
n
ce
in
m
an
y
f
ield
s
,
it
s
u
f
f
er
s
f
r
o
m
s
o
m
e
d
r
a
w
b
ac
k
s
.
Ma
n
y
o
f
te
x
t
u
r
e
f
ea
t
u
r
es
ar
e
p
r
o
p
o
s
ed
b
as
ed
o
n
L
B
P
an
d
in
h
er
it
th
e
d
r
a
wb
ac
k
s
.
T
h
e
L
B
P
is
s
en
s
iti
v
e
to
n
o
is
e,
an
d
d
if
f
er
en
t
p
atter
n
s
o
f
L
B
P
m
a
y
b
e
class
if
ied
in
to
th
e
s
a
m
e
class
th
at
r
ed
u
ce
s
it
s
d
is
cr
i
m
i
n
ati
n
g
p
r
o
p
er
ty
[
1
2
]
.
T
o
o
v
er
co
m
e
L
B
P
d
r
a
w
b
ac
k
s
,
w
e
p
r
o
p
o
s
ed
a
n
ew
te
x
t
u
r
e
d
escr
ip
to
r
,
ca
lled
C
o
m
p
leted
L
o
ca
l
ter
n
ar
y
P
att
er
n
(
C
L
T
P
)
[
1
7
]
.
C
L
T
P
s
h
o
w
ed
g
o
o
d
ac
cu
r
ac
y
r
ate
s
i
n
m
a
n
y
f
ield
s
r
at
h
er
th
a
n
L
B
P
an
d
C
L
B
P
[
1
7
]
-
[
1
8
]
.
I
n
[
1
7
]
,
th
e
C
L
T
P
o
u
tp
er
f
o
r
m
ed
L
B
P
,
C
L
B
P
,
an
d
C
L
B
C
in
ter
m
o
f
te
x
t
u
r
e
class
i
f
icatio
n
ac
cu
r
ac
y
.
Mo
r
e
o
v
er
,
in
[
1
8
]
,
th
e
C
L
T
P
is
u
s
ed
f
o
r
i
m
ag
e,
e
v
en
t,
s
ce
n
e
an
d
m
ed
ical
i
m
ag
e
class
i
f
icatio
n
an
d
ac
h
ie
v
ed
h
i
g
h
er
clas
s
i
f
icatio
n
ac
cu
r
ac
y
c
o
m
p
ar
ed
w
it
h
L
B
P
,
C
L
B
P
,
an
d
C
L
B
C
.
I
n
th
is
p
ap
er
,
th
e
C
L
T
P
tex
tu
r
e
d
escr
ip
to
r
is
s
tu
d
ied
a
n
d
i
n
v
e
s
ti
g
ated
f
o
r
f
ac
e
r
ec
o
g
n
iti
o
n
s
y
s
te
m
.
Dif
f
er
en
t
s
ta
n
d
a
r
d
f
ac
e
d
atas
ets
ar
e
u
s
ed
in
th
i
s
s
tu
d
y
s
u
c
h
as
J
A
F
FE
an
d
P
E
I
d
atasets
.
T
h
e
ex
p
er
i
m
en
tal
r
esu
lt
s
illu
s
tr
ate
th
at
C
L
T
P
is
m
o
r
e
r
o
b
u
s
t
an
d
ac
h
ie
v
es
h
ig
h
er
f
ac
e
r
ec
o
g
n
itio
n
ac
cu
r
ac
y
r
ate
co
m
p
ar
ed
w
it
h
C
L
B
P
.
T
h
e
r
est
o
f
th
i
s
p
ap
er
is
o
r
g
a
n
is
ed
as
f
o
llo
w
s
.
Sec
tio
n
2
b
r
ief
l
y
r
ev
ie
w
s
t
h
e
L
B
P
an
d
C
L
B
P
.
Ou
r
p
r
o
p
o
s
ed
C
L
T
P
tex
tu
r
e
d
escr
ip
to
r
s
ar
e
ex
p
lain
ed
in
Sectio
n
3
.
T
h
en
in
Sectio
n
4
,
th
e
ex
p
er
i
m
en
tal
r
esu
l
ts
o
f
p
r
o
p
o
s
ed
f
ac
e
r
ec
o
g
n
itio
n
s
y
s
te
m
u
s
in
g
C
L
T
P
ar
e
r
ep
o
r
ted
an
d
d
i
s
cu
s
s
ed
.
F
in
all
y
,
Secti
o
n
4
co
n
cl
u
d
es
t
h
e
p
ap
er
.
2.
RE
L
AT
E
D
WO
RK
S
I
n
th
i
s
s
ec
tio
n
,
a
b
r
ief
r
ev
ie
w
o
f
th
e
L
B
P
an
d
C
L
B
P
ar
e
p
r
o
v
id
ed
.
2
.
1
.
L
o
ca
l B
ina
ry
P
a
t
t
er
n (
L
B
P
)
T
h
e
L
B
P
ca
lcu
latio
n
ca
n
b
e
d
escr
ib
ed
m
at
h
e
m
atica
l
l
y
as
f
o
l
lo
w
:
,
0
,
0
,
0
,
1
)
(
),
(
2
1
0
,
x
x
x
s
i
i
s
L
B
R
c
p
p
P
p
R
P
(
1
)
W
h
er
e
i
c
an
d
i
p
(
p
=
0
,
.
.
.
,
P
−
1
)
d
en
o
te
th
e
g
r
e
y
v
al
u
es
o
f
th
e
ce
n
tr
e
p
ix
e
l
an
d
th
e
n
eig
h
b
o
u
r
p
ix
el
o
n
a
cir
cle
o
f
r
ad
iu
s
R
,
r
esp
ec
tiv
el
y
,
a
n
d
P
d
en
o
tes
t
h
e
n
u
m
b
er
o
f
th
e
n
ei
g
h
b
o
u
r
s
.
T
o
esti
m
ate
t
h
e
n
ei
g
h
b
o
u
r
s
th
at
d
o
n
o
t
lie
ex
ac
tl
y
i
n
t
h
e
ce
n
tr
e
o
f
t
h
e
p
i
x
els,
t
h
e
b
ili
n
ea
r
in
ter
p
o
latio
n
est
i
m
at
io
n
m
et
h
o
d
is
u
s
ed
.
T
h
e
L
B
P
is
s
h
o
w
n
i
n
Fi
g
u
r
e
1
.
I
n
ad
d
itio
n
to
L
B
P
,
Oj
ala
et
al.
[
5
]
also
im
p
r
o
v
ed
t
h
e
o
r
ig
i
n
al
L
B
P
to
r
o
tatio
n
in
v
ar
ia
n
t
L
B
P
(
ri
R
P
L
B
P
,
)
an
d
u
n
if
o
r
m
r
o
tatio
n
in
v
ar
ian
t
L
B
P
(
2
,
r
i
u
R
P
L
B
P
)
.
Af
ter
d
o
in
g
t
h
e
en
co
d
i
n
g
s
tep
i
n
an
y
o
f
th
e
s
e
L
B
P
t
y
p
es;
i.e
.
,
L
B
P
,
ri
R
P
L
B
P
,
an
d
2
,
r
i
u
R
P
L
B
P
,
th
e
d
escr
ip
to
r
h
i
s
to
g
r
a
m
is
co
n
s
tr
u
cted
ac
co
r
d
in
g
to
E
q
u
atio
n
(
2
)
as f
o
llo
w
s
:
,
,
0
,
,
1
)
,
(
],
,
0
[
),
),
,
(
(
)
(
0
0
,
o
t
h
e
r
w
i
s
e
y
x
y
x
f
K
k
k
j
i
L
B
P
f
k
H
I
i
J
j
R
P
(
2
)
W
h
er
e
K
is
th
e
m
ax
i
m
a
l L
B
P
p
atter
n
v
al
u
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
3
,
J
u
n
e
2
0
1
7
:
1
5
9
4
–
1
6
0
1
1596
Fig
u
r
e
1
.
L
B
P
d
escr
ip
to
r
A
lt
h
o
u
g
h
,
m
a
n
y
r
esear
ch
er
s
t
ar
g
eted
th
e
L
B
P
an
d
d
id
m
a
n
y
i
m
p
r
o
v
e
m
e
n
ts
o
n
it.
T
h
e
d
r
a
w
b
ac
k
s
o
f
th
e
L
B
P
ar
e
in
h
er
ited
to
all
tex
t
u
r
e
d
escr
ip
to
r
s
i
n
s
p
ir
ed
f
r
o
m
th
e
L
B
P
d
escr
ip
to
r
.
T
h
e
f
ir
s
t
d
r
a
w
b
ac
k
i
s
s
en
s
iti
v
it
y
to
n
o
is
e
as
s
h
o
w
n
in
t
h
e
ex
a
m
p
le
i
n
Fi
g
u
r
e
2
wh
ile
th
e
s
ec
o
n
d
d
r
a
w
b
ac
k
i
s
s
h
o
w
n
in
Fi
g
u
r
e
3
w
h
er
e
d
if
f
er
en
t
p
atter
n
s
o
f
L
B
P
m
a
y
b
e
w
r
o
n
g
l
y
clas
s
i
f
ied
i
n
to
th
e
s
a
m
e
cla
s
s
t
h
at
r
ed
u
ce
s
its
d
is
cr
i
m
i
n
ati
n
g
p
r
o
p
er
ty
.
Fig
u
r
e
2
.
T
h
e
ex
a
m
p
le
f
o
r
L
B
P
o
p
e
r
ato
r
’
s
n
o
is
e
s
e
n
s
it
iv
i
t
y
Fig
u
r
e
3
.
Si
m
ilar
L
B
P
co
d
es f
o
r
t
w
o
d
if
f
er
e
n
t te
x
tu
r
e
p
atter
n
s
2
.
2
.
C
o
m
plet
ed
L
o
ca
l B
ina
ry
P
a
t
t
er
n (
CL
B
P
)
I
n
2
0
1
0
,
Gu
o
et
al.
[
1
2
]
p
r
o
p
o
s
ed
th
e
co
m
p
leted
L
B
P
(
C
L
B
P
)
d
escr
ip
to
r
.
I
n
C
L
B
P
,
th
e
i
m
ag
e
lo
ca
l
d
if
f
er
e
n
ce
i
s
d
ec
o
m
p
o
s
ed
i
n
t
o
t
w
o
co
m
p
le
m
en
tar
y
co
m
p
o
n
en
t
s
;
th
e
s
i
g
n
co
m
p
o
n
en
t
s
p
an
d
t
h
e
m
a
g
n
it
u
d
e
co
m
p
o
n
e
n
t
m
p
.
|
)
(
|
,
)
(
c
p
p
c
p
p
i
i
m
i
i
s
s
(
3
)
5
2
t
o
5
6
46
51
46
45
52
52
52
53
50
46
51
46
45
52
56
52
53
50
No
is
e
L
B
P
1
0
0
0
0
0
0
1
1
L
B
P
0
0
0
0
0
0
0
0
52
70
90
102
20
34
40
9
12
130
92
200
95
70
90
100
9
12
1
1
1
1
0
1
0
0
L
B
P
0
1
1
1
1
0
1
0
L
B
P
0
1
1
1
1
0
1
0
1
1
1
1
0
1
0
0
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
F
a
ce
R
ec
o
g
n
itio
n
Usi
n
g
C
o
m
p
leted
Lo
ca
l Tern
a
r
y
P
a
tter
n
(
C
LT
P
)
Textu
r
e
De
s
crip
to
r
(
Ta
h
a
H.
R
a
s
s
em)
1597
T
h
en
,
th
e
s
p
is
u
s
ed
to
b
u
ild
th
e
C
L
B
P
-
Si
g
n
(
C
LBP
_
S
)
,
w
h
er
ea
s
th
e
m
p
i
s
u
s
ed
to
b
u
ild
C
L
B
P
-
m
ag
n
it
u
d
e
(
C
LBP
_
M
)
.
T
h
e
C
LBP
_
S
an
d
C
LBP
_
M
ar
e
m
ath
e
m
atica
ll
y
d
escr
ib
ed
as f
o
llo
ws:
,
,
0
,
,
1
)
(
2
_
1
0
,
o
t
h
e
r
wi
s
e
i
i
s
i
i
s
S
CL
B
P
c
p
P
c
p
p
P
p
R
P
(
4
)
,
|
)
(
|
,
0
,
|
)
(
|
,
1
)
,
(
)
,
(
2
_
1
0
,
c
i
i
c
i
i
c
m
t
c
m
t
M
C
L
B
P
c
p
c
p
p
p
P
p
R
P
p
(
5
)
W
h
er
e
i
c
,
i
p
,
R
an
d
P
ar
e
d
ef
in
ed
b
ef
o
r
e
in
E
q
u
atio
n
(
1
)
,
w
h
i
le
c
d
en
o
tes th
e
m
ea
n
v
al
u
e
o
f
m
p
in
th
e
w
h
o
le
i
m
a
g
e.
T
h
e
C
LBP
_
S
is
eq
u
al
to
L
B
P
w
h
er
ea
s
t
h
e
C
LBP
_
M
m
ea
s
u
r
es th
e
lo
ca
l v
ar
ia
n
ce
o
f
m
ag
n
it
u
d
e.
Fu
r
t
h
er
m
o
r
e,
Gu
o
et
al.
[
1
2
]
u
s
ed
th
e
v
al
u
e
o
f
t
h
e
g
r
e
y
le
v
el
o
f
ea
ch
p
atter
n
to
co
n
s
tr
u
ct
a
n
e
w
o
p
er
ato
r
,
ca
lled
C
L
B
P
-
ce
n
tr
e
(
C
L
B
P
_
C
)
.
T
h
e
C
LBP
_
C
ca
n
b
e
m
a
th
e
m
atica
ll
y
d
escr
ib
ed
as f
o
llo
w
s
:
)
,
(
_
,
I
c
R
P
c
i
t
C
C
L
B
P
(
6
)
W
h
er
e
i
c
d
en
o
tes th
e
g
r
e
y
v
alu
e
o
f
th
e
ce
n
tr
e
p
ix
el
an
d
c
I
is
t
h
e
av
er
a
g
e
g
r
e
y
le
v
el
o
f
th
e
wh
o
le
i
m
a
g
e.
Gu
o
et
a
l.
[
1
2
]
co
m
b
in
ed
t
h
ei
r
o
p
er
ato
r
s
in
to
j
o
in
t
o
r
h
y
b
r
i
d
d
is
tr
ib
u
tio
n
s
a
n
d
ac
h
iev
ed
r
e
m
ar
k
ab
le
tex
t
u
r
e
class
i
f
icatio
n
ac
cu
r
ac
y
.
T
h
ey
co
m
b
in
ed
C
LBP
_
S
a
n
d
C
LBP
_
M
i
n
t
w
o
w
a
y
s
.
I
n
th
e
f
ir
s
t
w
a
y
,
t
h
e
y
co
n
ca
ten
ated
t
h
eir
h
is
to
g
r
a
m
to
b
u
ild
C
LBP
_
S
_
M
,
w
h
i
le
in
t
h
e
s
ec
o
n
d
w
a
y
t
h
e
y
ca
lc
u
lated
t
h
e
2
D
j
o
in
t
h
is
to
g
r
a
m
.
T
h
is
2
D
j
o
in
t
h
i
s
to
g
r
a
m
i
s
k
n
o
w
n
a
s
C
LBP
_
S
/M
.
T
h
e
C
LBP
_
C
also
co
m
b
in
ed
w
it
h
t
h
e
C
LBP
_
S
an
d
C
LBP
_
M
in
t
w
o
w
a
y
s
.
I
n
th
e
f
ir
s
t
w
a
y
,
b
o
th
o
f
t
h
e
m
ar
e
co
m
b
i
n
ed
as
3D
j
o
in
t
h
is
to
g
r
a
m
an
d
d
en
o
ted
as
C
LBP
_
S
/M/
C
.
I
n
th
e
s
ec
o
n
d
w
a
y
,
th
e
C
LBP
_
C
is
f
ir
s
t
co
m
b
in
ed
j
o
in
tl
y
w
it
h
th
e
C
LB
P
_
S
o
r
C
LBP
_
M
to
b
u
ild
2
D
j
o
in
t
h
is
to
g
r
a
m
d
e
n
o
ted
C
LBP
_
S
/C
o
r
C
LBP
_
M/
C
,
r
esp
ec
tiv
el
y
.
T
h
en
,
t
h
i
s
2
D
j
o
in
t
h
is
to
g
r
a
m
h
a
s
t
o
co
n
v
er
t
to
1D
h
is
to
g
r
a
m
a
n
d
h
as
to
b
e
co
n
ca
ten
ated
w
it
h
C
LBP
_
M
o
r
C
LBP
_
S
to
b
u
ild
th
e
f
in
al
h
i
s
to
g
r
a
m
th
at
d
en
o
ted
b
y
C
LBP
_
M_
S
/C
o
r
C
LBP
_
S
_
M/C
.
3.
CO
M
P
L
E
T
E
D
L
O
CAL
T
E
RNAR
Y
P
AT
T
E
RN
(
CL
T
P
)
I
n
C
L
T
P
[
1
7
]
,
l
o
ca
l
d
if
f
er
en
ce
o
f
th
e
i
m
ag
e
is
d
ec
o
m
p
o
s
ed
in
to
t
w
o
s
i
g
n
co
m
p
le
m
en
tar
y
co
m
p
o
n
e
n
t
s
an
d
t
w
o
m
a
g
n
it
u
d
e
co
m
p
le
m
en
tar
y
co
m
p
o
n
e
n
t
s
as f
o
llo
w
s
:
|
)
(
|
|,
)
(
|
))
(
(
)
)
,
(
(
t
i
i
m
t
i
i
m
t
i
i
s
s
t
i
i
s
s
c
p
l
o
w
e
r
p
c
p
u
p
p
e
r
p
c
p
l
o
w
e
r
p
c
p
u
p
p
e
r
p
(
7
)
W
h
er
e
i
c
,
an
d
i
p
ar
e
d
ef
in
ed
b
ef
o
r
e
in
(
1
)
w
h
ile
t
d
en
o
tes t
h
e
u
s
er
T
h
r
esh
o
ld
.
T
h
en
,
th
e
u
p
p
e
r
p
s
an
d
l
o
w
e
r
p
s
ar
e
u
s
ed
to
b
u
ild
th
e
u
p
p
e
r
R
P
S
C
L
T
P
,
_
an
d
l
o
w
e
r
R
P
S
C
L
T
P
,
_
,
r
esp
ec
tiv
el
y
,
a
s
f
o
llo
w
s
:
,
,
0
,
,
1
))
(
(
2
_
1
0
,
o
t
h
e
r
w
i
s
e
t
i
i
s
t
i
i
s
S
CL
T
P
c
p
u
p
p
e
r
P
c
p
p
P
p
u
p
p
e
r
R
P
(
8
)
,
,
0
,
,
1
))
(
(
2
_
1
0
,
o
t
h
e
r
wi
s
e
t
i
i
s
t
i
i
s
S
C
L
T
P
c
p
l
o
w
e
r
P
c
p
p
P
p
l
o
w
e
r
R
P
(
9
)
T
h
en
R
P
S
C
L
T
P
,
_
is
th
e
co
n
ca
te
n
atio
n
o
f
t
h
e
u
p
p
e
r
R
P
S
C
L
T
P
,
_
an
d
l
o
w
e
r
R
P
S
C
L
T
P
,
_
,
as f
o
llo
w
s
:
]
_
_
[
_
,
,
,
l
o
w
e
r
R
P
u
p
p
e
r
R
P
R
P
S
C
L
T
P
S
C
L
T
P
S
C
L
T
P
(
1
0
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
3
,
J
u
n
e
2
0
1
7
:
1
5
9
4
–
1
6
0
1
1598
Si
m
i
lar
to
R
P
S
C
L
T
P
,
_
,
th
e
R
P
M
C
L
T
P
,
_
is
b
u
il
t
u
s
in
g
t
h
e
t
w
o
m
a
g
n
it
u
d
e
co
m
p
le
m
en
tar
y
co
m
p
o
n
e
n
t
s
u
p
p
e
r
p
m
an
d
l
o
w
e
r
p
m
,
as f
o
llo
w
s
:
c
t
i
i
c
t
i
i
c
m
t
c
m
t
M
C
L
T
P
c
p
c
p
u
p
p
e
r
p
u
p
p
e
r
p
P
p
u
p
p
e
r
r
R
P
p
|
)
(
|
,
0
,
|
)
(
|
,
1
)
,
(
)
,
(
2
_
1
0
,
(
1
1
)
c
t
i
i
c
t
i
i
c
m
t
c
m
t
M
C
L
T
P
c
p
c
p
l
o
w
e
r
p
l
o
w
e
r
p
P
p
l
o
w
e
r
R
P
p
|
)
(
|
,
0
,
|
)
(
|
,
1
)
,
(
)
,
(
2
_
1
0
,
(
1
2
)
]
_
_
[
_
,
,
,
l
o
w
e
r
R
P
u
p
p
e
r
R
P
R
P
M
C
L
T
P
M
C
L
T
P
M
C
L
T
P
(
1
3
)
Mo
r
eo
v
er
,
th
e
u
p
p
e
r
R
P
C
C
L
T
P
,
_
an
d
l
o
w
e
r
R
P
C
C
L
T
P
,
_
ca
n
b
e
d
escr
ib
ed
m
at
h
e
m
atica
ll
y
as
f
o
llo
w
s
:
)
,
(
_
,
I
u
p
p
e
r
c
u
p
p
e
r
R
P
c
i
t
C
C
L
T
P
(
1
4
)
)
,
(
_
,
I
l
o
w
e
r
c
l
o
e
rr
R
P
c
i
t
C
C
L
T
P
(
1
5
)
W
h
er
e
,
t
i
i
c
u
p
p
e
r
c
t
i
i
c
l
o
w
e
r
c
an
d
I
c
is
th
e
av
er
a
g
e
g
r
e
y
lev
el
o
f
th
e
w
o
r
ld
i
m
a
g
e.
T
h
e
p
r
o
p
o
s
ed
C
L
T
P
o
p
e
r
ato
r
s
ar
e
co
m
b
in
ed
i
n
to
j
o
in
t
o
r
h
y
b
r
id
d
is
tr
ib
u
tio
n
s
to
b
u
ild
th
e
f
i
n
al
o
p
er
ato
r
h
is
to
g
r
a
m
li
k
e
t
h
e
C
L
B
P
an
d
C
L
B
C
[
1
2
]
,
[
1
3
]
.
I
n
th
e
C
L
T
P
,
th
e
o
p
er
ato
r
s
o
f
th
e
s
a
m
e
t
y
p
e
o
f
p
atter
n
;
i.e
.
,
t
h
e
u
p
p
er
a
n
d
th
e
lo
w
er
p
atter
n
,
ar
e
co
m
b
i
n
ed
f
ir
s
t
i
n
to
j
o
in
t
o
r
h
y
b
r
id
d
i
s
tr
ib
u
tio
n
s
.
T
h
en
,
t
h
ei
r
r
esu
lt
s
ar
e
co
n
ca
ten
ated
to
b
u
i
ld
th
e
f
i
n
al
o
p
er
ato
r
h
is
to
g
r
a
m
.
4.
E
XP
E
R
I
M
E
NT
S AN
D
DIS
CUSS
I
O
NS
I
n
th
is
s
ec
tio
n
,
s
er
ies
o
f
ex
p
er
i
m
e
n
ts
ar
e
p
er
f
o
r
m
ed
to
s
tu
d
y
an
d
in
v
es
tig
a
te
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
C
L
T
P
f
o
r
f
ac
e
r
ec
o
g
n
itio
n
t
ask
.
J
AFFE
[
1
9
]
an
d
FEI
[
2
0
]
s
tan
d
ar
d
d
atab
ases
ar
e
u
s
ed
in
t
h
i
s
s
t
u
d
y
.
E
m
p
ir
icall
y
,
t
h
e
t
h
r
es
h
o
ld
v
al
u
e
t
is
s
et
to
5
i
n
all
C
L
T
P
ex
p
er
i
m
en
ts
.
D
if
f
er
e
n
t d
atase
ts
wer
e
u
s
ed
in
o
r
d
er
to
f
i
n
d
th
e
s
u
itab
le
t
h
r
esh
o
ld
v
a
l
u
e
w
h
ic
h
w
il
l
b
e
u
s
ed
in
t
h
e
C
L
T
P
ev
alu
atio
n
ex
p
er
i
m
en
t
s
.
T
h
e
v
alu
es
w
er
e
r
an
g
ed
f
r
o
m
0
to
2
5
,
an
d
5
was
th
e
s
u
i
tab
le
t
h
r
es
h
o
ld
v
al
u
e
[
1
7
]
-
[
1
8
]
.
I
n
all
ex
p
er
i
m
en
t
s
,
th
e
L
B
P
,
C
L
B
P
,
an
d
C
L
T
P
ar
e
ex
tr
ac
ted
b
ased
o
n
th
r
ee
d
if
f
er
en
t
te
x
t
u
r
e
p
atter
n
s
,
n
a
m
el
y
,
(
P
=
8
an
d
R
=
1
)
,
(
P
=
1
6
an
d
R
=
2
)
,
an
d
(
P
=
2
4
an
d
R
=
3
)
.
4
.
1
.
Dis
s
i
m
ila
rit
y
M
e
a
s
uring
F
ra
m
ew
o
r
k
I
n
th
i
s
s
t
u
d
y
,
t
h
e
n
ea
r
es
t
n
ei
g
h
b
o
u
r
h
o
o
d
class
i
f
ier
as
w
el
l
as
th
e
c
h
i
-
s
q
u
ar
e
s
tatis
tic
i
s
u
s
ed
to
m
ea
s
u
r
e
th
e
d
is
s
i
m
ilar
it
y
o
f
t
h
e
h
is
to
g
r
a
m
s
.
E
q
u
atio
n
(
1
6
)
d
escr
ib
es
th
e
2
d
is
tan
ce
b
et
w
ee
n
t
w
o
h
is
to
g
r
a
m
s
i
h
H
an
d
i
k
K
w
h
er
e
)
,
.
.
.
3
,
2
,
1
(
B
i
.
B
i
i
i
i
i
k
h
k
h
K
H
i
t
y
D
i
s
s
i
m
i
l
a
r
1
2
)
(
)
,
(
2
(
1
6
)
4
.
2
.
E
x
peri
m
e
nta
l R
esu
lt
s
o
f
J
AF
F
E
da
t
a
s
et
T
h
e
J
A
FF
E
d
atab
ase
in
clu
d
es
1
0
class
es
an
d
to
tal
2
1
3
im
a
g
es.
E
ac
h
class
h
as
2
0
J
P
E
G
i
m
ag
e
s
o
f
d
if
f
er
e
n
t
J
ap
an
ese
f
e
m
ale
i
n
a
d
if
f
er
en
t
v
ie
w
o
f
f
ac
e
ex
p
r
es
s
io
n
w
h
ic
h
h
as
an
g
r
y
,
s
m
ile,
s
ad
,
w
o
r
r
y
,
n
er
v
o
u
s
,
n
eu
tr
al
a
n
d
etc.
T
h
ese
im
a
g
es
ar
e
g
r
ey
a
n
d
2
5
6
x
2
5
6
in
s
ize.
E
x
a
m
p
les
o
f
t
h
ese
i
m
ag
e
s
ar
e
s
h
o
w
n
i
n
Fi
g
u
r
e
4
.
T
a
b
le
1
s
h
o
w
s
t
h
e
a
v
er
ag
e
o
f
clas
s
i
f
icatio
n
r
es
u
lts
o
f
f
ac
e
r
ec
o
g
n
it
io
n
d
ata
s
et
o
f
1
0
0
r
an
d
o
m
s
p
lit
s
.
I
n
ea
ch
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
F
a
ce
R
ec
o
g
n
itio
n
Usi
n
g
C
o
m
p
leted
Lo
ca
l Tern
a
r
y
P
a
tter
n
(
C
LT
P
)
Textu
r
e
De
s
crip
to
r
(
Ta
h
a
H.
R
a
s
s
em)
1599
class
,
N
=
(
2
,
5
,
1
0
)
is
u
s
ed
a
s
tr
ain
in
g
i
m
a
g
es,
w
h
i
le
t
h
e
r
e
m
ai
n
in
g
i
m
a
g
es
ar
e
u
s
ed
as
test
i
n
g
i
m
a
g
e
s
.
T
h
e
b
est
class
i
f
icat
io
n
ac
c
u
r
ac
y
h
as
o
b
tain
ed
b
y
C
LT
P
_
S
/M/
C
3,
24
o
p
er
ato
r
,
w
h
ich
h
as
r
ea
c
h
e
d
u
p
9
9
.
3
8
%
w
h
ile
th
e
C
LBP
_
S
/M/
C
3,
24
h
as
ac
h
ie
v
ed
th
e
b
est
class
if
ica
tio
n
ac
c
u
r
ac
y
,
w
h
ich
h
as
r
ea
ch
ed
u
p
t
o
9
8
.
7
7
%.
Fro
m
th
e
T
ab
le
X,
all
C
L
T
P
d
escr
ip
to
r
s
o
u
tp
er
f
o
r
m
ed
th
e
C
L
B
P
d
escr
ip
to
r
s
in
ter
m
o
f
ac
c
u
r
ac
y
r
ate.
Fig
u
r
e
4
.
So
m
e
i
m
a
g
es
f
r
o
m
J
AFFE
Data
b
ase
T
ab
le
1
.
C
lass
if
icatio
n
r
ates (
%)
o
n
J
A
F
FE
Data
b
ase
D
e
scri
p
t
o
r
s
R
=
1
,
P
=
8
R
=
2
,
P
=
1
6
R
=
3
,
P
=
2
4
2
5
10
2
5
10
2
5
10
C
L
B
P
_
S
5
6
.
4
3
6
3
.
9
7
7
0
.
5
2
4
7
.
6
2
5
3
.
5
4
5
8
.
7
8
4
8
.
5
9
5
6
.
3
4
6
1
.
7
6
C
L
T
P
_
S
7
4
.
2
7
8
1
.
3
7
8
5
.
7
4
8
0
.
4
1
8
6
.
6
3
9
0
.
5
0
8
4
.
6
1
8
9
.
8
8
9
4
.
0
0
C
L
B
P
_
M
7
0
.
2
9
7
4
.
3
7
7
6
.
5
8
7
4
.
9
3
7
9
.
9
7
8
3
.
9
4
7
3
.
4
7
7
9
.
3
1
8
3
.
7
6
C
L
T
P
_
M
7
8
.
2
0
8
4
.
6
7
8
8
.
7
6
8
0
.
5
9
8
5
.
6
1
8
7
.
7
2
7
8
.
7
4
8
5
.
0
3
8
9
.
0
2
C
L
B
P
_
M
/
C
8
7
.
2
6
9
1
.
5
3
9
4
.
0
3
8
8
.
0
3
9
2
.
7
3
9
4
.
8
9
8
9
.
0
9
9
4
.
1
4
9
6
.
2
7
C
L
T
P
_
M
/
C
8
6
.
4
2
9
1
.
8
8
9
5
.
4
2
9
0
.
4
4
9
5
.
2
5
9
7
.
8
8
9
1
.
9
9
9
6
.
1
2
9
8
.
2
4
C
L
B
P
_
S
_
M
/
C
8
6
.
2
4
9
0
.
7
1
9
3
.
9
8
8
5
.
7
9
9
1
.
9
7
9
4
.
6
8
8
9
.
7
3
9
5
.
0
0
9
7
.
1
9
C
L
T
P
_
S
_
M
/
C
8
8
.
2
4
9
3
.
4
0
9
6
.
6
6
9
2
.
1
3
9
6
.
5
6
9
8
.
5
4
9
3
.
1
0
9
6
.
5
1
9
8
.
1
0
C
L
B
P
_
S
/
M
7
3
.
3
4
8
1
.
6
0
8
7
.
7
0
7
2
.
5
4
8
1
.
7
3
8
8
.
3
4
7
9
.
7
8
8
6
.
5
1
9
0
.
7
0
C
L
T
P
_
S
/
M
8
4
.
2
0
9
0
.
3
3
9
3
.
7
4
8
7
.
6
7
9
3
.
5
1
9
6
.
5
4
9
0
.
0
0
9
3
.
9
3
9
7
.
1
8
C
L
B
P
_
S
/
M
/
C
8
6
.
3
9
9
2
.
2
9
9
4
.
6
0
8
8
.
2
8
9
3
.
5
5
9
6
.
0
2
9
2
.
5
4
9
6
.
6
5
9
8
.
7
7
C
L
T
P
_
S
/
M
/
C
8
9
.
2
1
9
5
.
0
8
9
7
.
4
6
9
4
.
5
7
9
7
.
2
5
9
8
.
7
4
9
5
.
4
4
9
8
.
5
3
9
9
.
3
8
4
.
3
.
E
x
peri
m
e
nta
l R
esu
lt
s
o
f
F
E
I
F
a
ce
Da
t
a
ba
s
e
T
h
e
FEI
f
ac
e
d
atab
ase
i
s
o
n
e
o
f
t
h
e
s
ta
n
d
ar
d
f
ac
e
s
d
atab
ase
s
.
T
h
is
d
atab
ase
co
llected
in
c
lu
d
es
f
ac
es
f
o
r
2
0
0
B
r
az
ilian
p
er
s
o
n
s
ca
p
tu
r
ed
o
n
2
0
0
5
an
d
2
0
0
6
at
A
r
ti
f
icial
I
n
telli
g
e
n
ce
L
ab
o
r
a
to
r
y
o
f
FEI
in
São
B
er
n
ar
d
o
d
o
C
a
m
p
o
,
São
P
au
lo
,
B
r
az
il.
FEI
f
ac
e
d
a
tab
ase
h
a
s
a
2
8
0
0
f
ac
e
i
m
ag
e
i
n
to
t
al
f
r
o
m
t
h
e
s
t
u
d
en
t
an
d
s
ta
f
f
i
n
t
h
e
FEI
f
r
o
m
1
9
y
ea
r
s
o
ld
to
4
0
y
ea
r
s
o
ld
.
T
h
e
i
m
a
g
es
i
n
FEI
d
atab
ase
ar
e
o
r
g
an
i
s
ed
in
2
0
0
class
es
a
n
d
ea
ch
cla
s
s
co
n
ta
in
s
1
4
i
m
ag
e
s
f
o
r
th
e
s
a
m
e
p
er
s
o
n
in
d
if
f
er
e
n
t
f
ac
e
v
ie
w
,
r
o
ta
tio
n
o
f
al
m
o
s
t
1
8
0
d
eg
r
ee
s
an
d
w
it
h
d
if
f
er
en
t
f
ac
ial
ex
p
r
ess
io
n
.
T
h
e
s
ize
o
f
f
ac
es
i
m
a
g
e
is
6
4
0
x
4
8
0
p
ix
els.
Fig
u
r
e
x
s
h
o
ws
s
o
m
e
ex
a
m
p
les o
f
FEI
f
ac
e
d
a
tab
ase.
Fig
u
r
e
5
.
So
m
e
i
m
a
g
es
f
r
o
m
F
E
I
Face
Data
b
ase
T
ab
le
2
s
h
o
w
s
t
h
e
av
er
ag
e
o
f
class
i
f
icatio
n
r
esu
lt
s
o
f
f
ac
e
r
ec
o
g
n
itio
n
d
ataset
o
f
1
0
0
r
an
d
o
m
s
p
lit
s
.
I
n
ea
c
h
cla
s
s
,
N
=
(
2
,
5
,
1
0
)
is
u
s
ed
as
tr
ain
in
g
i
m
a
g
es,
w
h
ile
th
e
r
e
m
a
in
i
n
g
i
m
a
g
es
ar
e
u
s
ed
a
s
test
i
n
g
i
m
a
g
es.
T
h
e
b
est
clas
s
i
f
icati
o
n
ac
cu
r
ac
y
h
as
o
b
t
ai
n
ed
b
y
C
LT
P
_
S
/M/
C
2,
16
o
p
er
ato
r
,
w
h
ich
h
a
s
r
ea
ch
ed
u
p
8
5
.
2
2
%
w
h
ile
th
e
C
LBP
_
S
/M
/C
3,
24
h
as
ac
h
iev
ed
t
h
e
b
est
class
i
f
icatio
n
ac
cu
r
ac
y
,
w
h
ic
h
h
as
r
ea
ch
ed
u
p
to
7
6
.
1
5
%.
A
s
id
e
f
r
o
m
C
LT
P
_
M
o
p
er
ato
r
w
h
en
N
=
(
2
,
5
,
1
0
)
at
P
=
2
4
,
R
=
3
,
all
C
L
T
P
o
p
er
at
o
r
s
h
av
e
ac
h
iev
ed
h
ig
h
er
p
er
f
o
r
m
a
n
ce
t
h
an
C
L
B
P
o
p
er
at
o
r
s
f
o
r
all
N
n
u
m
b
er
s
o
f
tr
ai
n
i
n
g
i
m
a
g
es a
t
r
ad
iu
s
es 1
,
2
,
3.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
3
,
J
u
n
e
2
0
1
7
:
1
5
9
4
–
1
6
0
1
1600
T
ab
le
2
.
C
lass
if
icatio
n
r
ates (
%)
o
n
FEI
Face
Data
b
ase
F
E
I
d
a
t
a
b
a
se
R
=
1
,
P
=
8
R
=
2
,
P
=
1
6
R
=
3
,
P
=
2
4
2
5
10
2
5
10
2
5
10
C
L
B
P
_
S
1
2
.
1
3
1
6
.
9
0
2
1
.
0
3
9
.
0
9
1
1
.
9
8
1
5
.
1
5
1
2
.
1
8
1
6
.
2
5
2
0
.
1
2
C
L
T
P
_
S
3
0
.
5
5
3
8
.
5
3
4
6
.
6
6
4
3
.
1
7
5
4
.
8
2
6
6
.
7
4
4
7
.
1
6
5
9
.
3
4
7
2
.
5
8
C
L
B
P
_
M
1
8
.
9
8
2
6
.
4
7
3
3
.
2
9
2
2
.
8
7
3
2
.
9
3
4
1
.
0
0
2
6
.
9
6
3
6
.
9
0
4
5
.
1
6
C
L
T
P
_
M
2
5
.
1
4
3
2
.
8
2
4
2
.
8
1
2
6
.
8
1
3
5
.
0
3
4
7
.
2
5
2
6
.
3
4
3
4
.
6
3
4
5
.
1
9
C
L
B
P
_
M
/
C
3
3
.
4
6
4
6
.
1
7
5
5
.
6
1
3
5
.
2
2
4
9
.
7
6
5
9
.
6
6
4
0
.
9
7
5
3
.
8
7
6
1
.
1
6
C
L
T
P
_
M
/
C
4
3
.
2
1
5
4
.
2
0
6
8
.
0
4
4
5
.
9
3
5
6
.
8
7
7
1
.
1
0
4
2
.
7
8
5
4
.
4
8
6
8
.
9
6
C
L
B
P
_
S
_
M
/
C
3
2
.
8
2
4
4
.
1
6
5
3
.
6
4
3
7
.
2
4
4
9
.
3
0
5
9
.
3
1
4
4
.
3
3
5
6
.
9
2
6
6
.
4
7
C
L
T
P
_
S
_
M
/
C
4
7
.
8
1
6
0
.
3
6
7
6
.
8
4
5
1
.
1
9
6
3
.
5
7
7
7
.
2
2
4
8
.
4
5
6
1
.
3
8
7
6
.
7
1
C
L
B
P
_
S
/
M
2
3
.
4
7
3
4
.
1
5
4
2
.
3
1
2
9
.
6
8
4
2
.
1
1
5
0
.
5
2
4
0
.
5
8
5
5
.
4
0
6
4
.
2
2
C
L
T
P
_
S
/
M
4
3
.
5
8
5
7
.
5
6
7
1
.
9
1
4
8
.
1
7
6
2
.
2
8
7
6
.
2
0
4
6
.
8
3
6
0
.
2
6
7
6
.
5
2
C
L
B
P
_
S
/
M
/
C
3
7
.
0
7
5
0
.
9
7
6
0
.
2
9
4
7
.
1
0
6
0
.
7
0
7
0
.
2
3
5
5
.
0
2
6
8
.
8
0
7
6
.
1
5
C
L
T
P
_
S
/
M
/
C
5
5
.
0
7
6
8
.
2
2
8
2
.
0
5
5
8
.
3
6
7
1
.
4
6
8
5
.
2
2
5
6
.
7
2
7
0
.
8
2
8
4
.
3
2
5.
CO
NCLU
SI
O
N
S
I
n
t
h
is
p
ap
er
,
th
e
p
r
o
p
o
s
ed
C
o
m
p
leted
L
o
ca
l
T
er
n
ar
y
P
atte
r
n
(
C
L
T
P
)
tex
tu
r
e
d
escr
ip
to
r
ar
e
s
tu
d
ie
d
an
d
ev
al
u
ated
f
o
r
f
ac
e
r
ec
o
g
n
itio
n
tas
k
.
T
w
o
s
ta
n
d
ar
d
f
ac
e
d
atasets
ar
e
u
s
ed
in
t
h
e
ex
p
er
i
m
en
ts
i
n
th
i
s
s
t
u
d
y
w
h
ic
h
ar
e
J
A
FF
E
a
n
d
P
E
I
d
atasets
.
Di
f
f
er
en
t
n
u
m
b
er
s
o
f
tr
ain
in
g
i
m
a
g
es
w
it
h
a
d
i
f
f
er
en
t
s
ize
o
f
t
h
e
d
escr
ip
to
r
s
ar
e
u
s
ed
in
th
e
ex
p
er
i
m
e
n
ts
.
T
h
e
ex
p
er
i
m
e
n
tal
r
esu
lts
s
h
o
w
ed
th
e
s
u
p
er
io
r
it
y
o
f
th
e
p
r
o
p
o
s
ed
C
L
T
P
ag
ain
s
t
C
L
B
P
in
b
o
th
J
AFFE
an
d
P
E
I
d
atab
ases
.
T
h
is
is
d
u
e
to
th
e
p
r
o
p
er
ties
o
f
C
L
T
P
c
o
m
p
ar
ed
w
it
h
C
L
B
P
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
i
s
w
o
r
k
is
s
u
p
p
o
r
t
e
d
b
y
t
h
e
Un
iv
er
s
iti
Ma
la
y
s
ia
P
ah
a
n
g
(
UM
P
)
v
ia
R
esear
ch
Gr
an
t
UM
P
R
DU1
6
0
3
4
9
an
d
R
esear
ch
Gr
an
t U
MP
DR
U1
5
0
3
5
3
.
RE
F
E
R
E
NC
E
S
[1
]
W.
H.
Al
-
A
ra
sh
i,
e
t
a
l.
,
“
Op
ti
m
i
z
in
g
p
rin
c
ip
a
l
c
o
m
p
o
n
e
n
t
a
n
a
ly
si
s
p
e
rf
o
r
m
a
n
c
e
f
o
r
f
a
c
e
r
e
c
o
g
n
it
io
n
u
sin
g
g
e
n
e
ti
c
a
lg
o
rit
h
m
,”
Ne
u
ro
c
o
mp
u
ti
n
g
,
v
o
l.
1
2
8
,
p
p
.
4
1
5
-
4
2
0
,
2
0
1
4
.
[2
]
K.
Et
e
m
a
d
a
n
d
R.
Ch
e
ll
a
p
p
a
.
,
“
D
isc
rim
in
a
n
t
a
n
a
l
y
sis
f
o
r
re
c
o
g
n
it
io
n
o
f
h
u
m
a
n
f
a
c
e
i
m
a
g
e
s
,
”
J
o
u
rn
a
l
o
f
t
h
e
Op
ti
c
a
l
S
o
c
iety
o
f
Ame
ric
a
A
.
,
v
o
l.
1
4
,
p
p
.
1
7
2
4
–
1
7
3
3
,
1
9
9
7
.
[3
]
M
.
S
.
Ba
rtl
e
tt
,
e
t
a
l
.,
“
F
a
c
e
re
c
o
g
n
it
io
n
b
y
in
d
e
p
e
n
d
e
n
t
c
o
m
p
o
n
e
n
t
a
n
a
ly
sis
,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Ne
u
r
a
l
Ne
two
rk
s
,
v
o
l.
1
3
,
p
p
.
1
4
5
0
-
1
4
6
4
,
2
0
0
2
.
[4
]
T
.
A
h
o
n
e
n
,
e
t
a
l
.
,
“
F
a
c
e
De
sc
rip
ti
o
n
w
it
h
L
o
c
a
l
Bi
n
a
ry
P
a
tt
e
rn
s:
A
p
p
li
c
a
ti
o
n
to
F
a
c
e
Re
c
o
g
n
it
io
n
,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Pa
tt
e
rn
A
n
a
lys
is
a
n
d
M
a
c
h
i
n
e
In
telli
g
e
n
c
e
,
v
o
l.
2
8
,
p
p
.
2
0
3
7
-
2
0
4
1
,
2
0
0
6
.
[5
]
T
.
Oja
la,
e
t
a
l
.
,
“
M
u
lt
ires
o
lu
t
io
n
g
ra
y
-
s
c
a
le
a
n
d
ro
tati
o
n
in
v
a
rian
t
t
e
x
tu
re
c
las
si
f
ica
ti
o
n
w
it
h
l
o
c
a
l
b
i
n
a
ry
p
a
tt
e
rn
s,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Pa
t
ter
n
An
a
lys
is
a
n
d
M
a
c
h
in
e
In
telli
g
e
n
c
e
,
v
o
l.
2
4
,
p
p
.
9
7
1
–
9
8
7
,
2
0
0
2
.
[6
]
J.
X
iao
,
e
t
a
l
.
,
“
S
UN
d
a
tab
a
se
:
L
a
r
g
e
sc
a
le
sc
e
n
e
re
c
o
g
n
it
io
n
f
ro
m
a
b
b
e
y
to
z
o
o
,
”
in
2
0
1
0
IEE
E
Co
mp
u
ter
S
o
c
iet
y
Co
n
fer
e
n
c
e
o
n
Co
m
p
u
ter
Vi
si
o
n
a
n
d
Pa
tt
e
rn
Rec
o
g
n
it
io
n
,
p
p
.
3
4
8
5
–
3
4
9
2
,
2
0
1
0
.
[7
]
X
.
W
a
n
g
,
e
t
a
l
.
,
“
A
n
HO
G
-
L
B
P
h
u
m
a
n
d
e
tec
to
r
w
it
h
p
a
rti
a
l
o
c
c
lu
sio
n
h
a
n
d
li
n
g
,
”
in
1
2
t
h
IE
E
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
f
C
o
mp
u
ter
Vi
sio
n
,
p
p
.
3
2
–
3
9
,
2
0
0
9
[8
]
V
.
T
a
k
a
la
a
n
d
M
.
P
ietik
a
in
e
n
,
“
M
u
lt
i
-
o
b
jec
t
trac
k
in
g
u
si
n
g
c
o
lo
r
,
tex
tu
re
a
n
d
m
o
ti
o
n
,
”
in
Pro
c
e
e
d
i
n
g
s
o
f
th
e
2
0
0
7
IEE
E
Co
m
p
u
ter
S
o
c
iety
Co
n
fer
e
n
c
e
o
n
Co
m
p
u
ter
Vi
si
o
n
a
n
d
Pa
tt
e
rn
Rec
o
g
n
it
i
o
n
,
p
p
.
1
–
7
,
2
0
0
7
.
[9
]
S
.
M
o
o
re
a
n
d
R.
Bo
w
d
e
n
,
“
L
o
c
a
l
b
in
a
ry
p
a
tt
e
rn
s
f
o
r
m
u
lt
i
-
v
ie
w
fa
c
ial
e
x
p
re
ss
io
n
re
c
o
g
n
it
io
n
,
”
Co
mp
u
ter
Vi
si
o
n
a
n
d
Ima
g
e
Un
d
e
rs
ta
n
d
in
g
,
v
o
l
/i
ss
u
e
:
1
1
5
(
4
),
p
p
.
5
4
1
-
5
5
8
,
2
0
1
1
.
[1
0
]
J
.
Y.
Ch
o
i
,
e
t
a
l
.,
“
Co
l
o
r
lo
c
a
l
tex
tu
re
f
e
a
tu
re
s
f
o
r
c
o
lo
r
f
a
c
e
re
c
o
g
n
it
io
n
,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s o
n
Ima
g
e
Pro
c
e
ss
in
g
,
v
o
l.
2
1
,
p
p
.
1
3
6
6
–
1
3
8
0
,
2
0
1
2
.
[1
1
]
X
.
T
a
n
a
n
d
B.
T
rig
g
s,
“
En
h
a
n
c
e
d
lo
c
a
l
tex
tu
re
f
e
a
tu
re
s
e
ts
f
o
r
f
a
c
e
re
c
o
g
n
it
io
n
u
n
d
e
r
d
if
f
icu
lt
li
g
h
ti
n
g
c
o
n
d
i
ti
o
n
s
,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Pa
t
ter
n
An
a
lys
is
a
n
d
M
a
c
h
in
e
In
telli
g
e
n
c
e
,
v
o
l.
1
9
,
p
p
.
1
6
3
5
–
1
6
5
0
,
2
0
1
0
.
[1
2
]
Z.
G
u
o
,
e
t
a
l
.
,
“
A
c
o
m
p
lete
d
m
o
d
e
li
n
g
o
f
lo
c
a
l
b
in
a
ry
p
a
tt
e
rn
o
p
e
ra
to
r
f
o
r
tex
tu
re
c
las
sifica
ti
o
n
,
”
IEE
E
Tr
a
n
sa
c
ti
o
n
s
o
n
Ima
g
e
Pro
c
e
ss
in
g
,
v
o
l.
1
9
,
p
p
.
1
6
5
7
–
1
6
6
3
,
2
0
1
0
.
[1
3
]
Y.
Zh
a
o
,
e
t
a
l
.,
“
Co
m
p
lete
d
lo
c
a
l
b
in
a
ry
c
o
u
n
t
f
o
r
ro
tatio
n
in
v
a
rian
t
tex
tu
re
c
las
si
f
ica
ti
o
n
,”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Ima
g
e
Pro
c
e
ss
in
g
,
v
o
l.
2
1
,
pp.
4
4
9
2
–
4
4
9
7
,
2
0
1
2
.
[1
4
]
S
.
S
i
n
g
h
,
e
t
a
l
.
,
“
A
p
p
li
c
a
ti
o
n
o
f
Co
m
p
lete
L
o
c
a
l
Bin
a
r
y
P
a
tt
e
rn
M
e
th
o
d
f
o
r
f
a
c
ial
e
x
p
re
ss
io
n
re
c
o
g
n
it
io
n
,”
i
n
4
t
h
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
In
telli
g
e
n
t
Hu
ma
n
Co
m
p
u
ter
I
n
ter
a
c
ti
o
n
(
IHCI)
,
p
p
.
1
-
4
,
2
0
1
2
.
[1
5
]
J.
L
i,
e
t
a
l
.,
“
F
a
c
ial
e
x
p
re
ss
io
n
re
c
o
g
n
it
io
n
b
a
se
d
o
n
c
o
m
p
lete
d
lo
c
a
l
b
in
a
ry
p
a
tt
e
rn
a
n
d
S
RC
,”
i
n
9
t
h
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Na
t
u
ra
l
Co
mp
u
t
a
t
io
n
(
ICNC)
,
p
p
.
3
3
3
-
3
3
7
,
2
0
1
3
.
[1
6
]
S.
A.
A.
M
.
F
a
u
d
z
i
a
n
d
N.
Ya
h
y
a
,
“
Ev
a
lu
a
ti
o
n
o
f
L
BP
-
b
a
se
d
f
a
c
e
re
c
o
g
n
it
io
n
tec
h
n
iq
u
e
s
,”
i
n
5
th
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
In
telli
g
e
n
t
a
n
d
Ad
v
a
n
c
e
d
S
y
ste
ms
(
ICIAS
)
,
p
p
.
1
-
6,
2
0
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
F
a
ce
R
ec
o
g
n
itio
n
Usi
n
g
C
o
m
p
leted
Lo
ca
l Tern
a
r
y
P
a
tter
n
(
C
LT
P
)
Textu
r
e
De
s
crip
to
r
(
Ta
h
a
H.
R
a
s
s
em)
1601
[1
7
]
T
.
H.
Ra
ss
e
m
a
n
d
B.
E.
K
h
o
o
,
“
Co
m
p
lete
d
lo
c
a
l
tern
a
ry
p
a
tt
e
rn
f
o
r
ro
tatio
n
i
n
v
a
rian
t
tex
tu
re
c
las
sif
ica
ti
o
n
,”
T
h
e
S
c
ien
ti
fi
c
W
o
rl
d
J
o
u
rn
a
l
,
2
0
1
4
.
[1
8
]
Ra
ss
e
m
T
.
H
.
,
e
t
a
l
.,
“
P
e
rf
o
r
m
a
n
c
e
e
v
a
lu
a
ti
o
n
o
f
Co
m
p
lete
d
L
o
c
a
l
T
e
rn
a
r
y
P
a
tt
e
rn
s
(CL
T
P
)
f
o
r
m
e
d
ica
l,
sc
e
n
e
a
n
d
e
v
e
n
t
i
m
a
g
e
c
a
te
g
o
risa
ti
o
n
,”
i
n
4
th
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
f
S
o
ft
wa
re
En
g
in
e
e
rin
g
a
n
d
Co
mp
u
ter
S
y
ste
ms
(
ICS
ECS
)
,
p
p
.
3
3
-
3
8
,
2
0
1
5
.
[1
9
]
M
.
J.
Ly
o
n
s,
e
t
a
l.
,
“
A
u
to
m
a
ti
c
c
las
sif
ic
a
ti
o
n
o
f
sin
g
le
fa
c
ial
i
m
a
g
e
s
,”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Pa
tt
e
r
n
An
a
lys
is
a
n
d
M
a
c
h
in
e
In
telli
g
e
n
c
e
,
v
o
l
.
2
1
,
p
p
.
1
3
5
7
-
1
3
6
2
,
1
9
9
9
.
[2
0
]
C.
E.
T
h
o
m
a
z
a
n
d
G
.
A
.
G
irald
i,
“
A
n
e
w
ra
n
k
in
g
m
e
th
o
d
f
o
r
p
ri
n
c
ip
a
l
c
o
m
p
o
n
e
n
ts
a
n
a
ly
sis
a
n
d
it
s
a
p
p
li
c
a
ti
o
n
to
f
a
c
e
i
m
a
g
e
a
n
a
l
y
sis
,”
Ima
g
e
a
n
d
Vi
sio
n
Co
mp
u
ti
n
g
,
v
o
l.
2
8
,
p
p
.
9
0
2
-
9
1
3
,
2
0
1
0
.
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
T
a
h
a
H.
R
a
ss
e
m
re
c
e
iv
e
d
th
e
B.
S
c
.
d
e
g
re
e
in
Co
m
p
u
ter
En
g
in
e
e
rin
g
f
ro
m
Un
iv
e
r
sit
y
o
f
T
e
c
h
n
o
lo
g
y
,
Ba
g
h
d
a
d
,
Ira
q
in
2
0
0
1
,
M
.
T
e
c
h
in
C
o
m
p
u
ter
S
c
ien
c
e
f
ro
m
Un
iv
e
rsit
y
o
f
H
y
d
e
ra
b
a
d
,
Hy
d
e
ra
b
a
d
,
In
d
ia
i
n
2
0
0
7
,
a
n
d
P
h
D
in
th
e
S
c
h
o
o
l
o
f
El
e
c
tri
c
a
l
&
El
e
c
tro
n
ic
En
g
in
e
e
rin
g
(Co
m
p
u
ter E
n
g
in
e
e
ri
n
g
-
Im
a
g
e
P
ro
c
e
ss
in
g
)
f
ro
m
Un
iv
e
rsiti
S
a
in
s M
a
lay
sia
in
2
0
1
4
.
Cu
rre
n
tl
y
,
h
e
is
a
se
n
io
r
lec
tu
re
r
in
fa
c
u
lt
y
o
f
c
o
m
p
u
ter
s
y
ste
m
s
a
n
d
so
f
t
w
a
r
e
e
n
g
in
e
e
rin
g
,
Un
iv
e
rsiti
M
a
lay
sia
P
a
h
a
n
g
(UM
P
)
.
His res
e
a
rc
h
in
tere
st i
s
in
th
e
a
re
a
o
f
c
o
m
p
u
ter v
isio
n
,
im
a
g
e
p
ro
c
e
ss
in
g
,
a
n
d
d
ig
it
a
l
w
a
ter
m
a
r
k
in
g
.
He
w
a
s
a
n
a
ss
istan
t
lec
tu
re
r
in
Co
m
p
u
ter
En
g
in
e
e
rin
g
De
p
a
rtme
n
t,
S
c
h
o
o
l
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
in
e
e
rin
g
a
t
Ho
d
e
id
a
h
U
n
iv
e
rsity
,
Ho
d
e
id
a
h
,
Ye
m
e
n
Na
srin
M
.
M
a
k
b
o
l
re
c
e
iv
e
d
th
e
B.
S
c
.
d
e
g
re
e
in
Co
m
p
u
ter
S
c
ien
c
e
f
ro
m
Un
iv
e
rsit
y
o
f
T
e
c
h
n
o
lo
g
y
,
Ba
g
h
d
a
d
,
Ira
q
in
2
0
0
1
,
M
.
T
e
c
h
in
C
o
m
p
u
ter
S
c
ien
c
e
f
ro
m
Un
iv
e
rsit
y
o
f
H
y
d
e
ra
b
a
d
,
Hy
d
e
ra
b
a
d
,
In
d
ia
i
n
2
0
0
7
,
a
n
d
P
h
D
in
th
e
S
c
h
o
o
l
o
f
El
e
c
tri
c
a
l
&
El
e
c
tro
n
ic
En
g
in
e
e
r
in
g
(Co
m
p
u
ter
E
n
g
in
e
e
rin
g
-
Dig
it
a
l
S
ig
n
a
l
P
r
o
c
e
ss
in
g
)
f
r
o
m
Un
iv
e
rsiti
S
a
in
s
M
a
lay
si
a
in
2
0
1
5
.
Cu
rre
n
tl
y
,
h
e
is
a
p
o
std
o
c
to
ra
l
f
e
ll
o
w
at
S
c
h
o
o
l
o
f
El
e
c
tri
c
a
l
&
El
e
c
tro
n
ic
En
g
in
e
e
r
in
g
,
Un
iv
e
rsiti
S
a
in
s
M
a
lay
si
a
.
H
er
re
se
a
rc
h
in
tere
st
is
in
th
e
a
re
a
o
f
im
a
g
e
p
ro
c
e
ss
in
g
,
a
n
d
d
ig
it
a
l
w
a
t
e
r
m
a
r
k
in
g
.
Sh
e
w
a
s
a
n
a
ss
i
sta
n
t
lec
tu
re
r
in
Co
m
p
u
ter
S
c
i
e
n
c
e
De
p
a
rt
m
e
n
t,
S
c
h
o
o
l
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
E
n
g
in
e
e
ri
n
g
a
t
Ho
d
e
i
d
a
h
Un
iv
e
rsit
y
,
Ho
d
e
i
d
a
h
,
Ye
m
e
n
.
S
a
m
Yin
Ye
e
re
c
e
iv
e
d
th
e
B.
S
c
.
d
e
g
re
e
in
M
u
l
ti
m
e
d
ia
a
n
d
Gra
p
h
ics
,
Un
iv
e
rsiti
M
a
lay
sia
P
a
h
a
n
g
,
2
0
1
7
.
C
u
rre
n
tl
y
,
sh
e
w
il
l
sta
rt
h
e
r
m
a
ste
r
stu
d
y
u
n
d
e
r
Dr
T
a
h
a
H.
Ra
ss
e
m
su
p
e
rv
isio
n
.
He
r
re
se
a
rc
h
in
tere
st i
s in
t
h
e
a
re
a
o
f
th
e
im
a
g
e
p
ro
c
e
ss
in
g
,
a
n
d
tex
t
u
re
c
las
sif
ic
a
ti
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.