I
n
d
on
e
s
ian
Jou
r
n
al
o
f
E
lec
t
r
ica
l
E
n
gin
e
e
r
in
g
a
n
d
Com
p
u
t
e
r
S
c
ience
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
,
pp.
989
~
99
4
I
S
S
N:
2502
-
4752,
DO
I
:
10
.
11591/i
j
e
e
c
s
.
v
25
.i
2
.
pp
989
-
99
4
989
Jou
r
n
al
h
o
m
e
page
:
ht
tp:
//
ij
e
e
c
s
.
iaes
c
or
e
.
c
om
A
f
ast
an
d
n
on
-
t
r
ai
n
ab
le
f
ac
ia
l
r
e
c
ogn
ition
sy
st
e
m
f
or
sc
h
ool
s
K
az
e
e
m
Oyebod
e
1
,
K
i
n
gs
l
e
y
Chi
wui
k
e
Uk
aoh
a
2
1
D
e
s
ig
n a
nd N
e
w
M
e
d
ia
D
e
pa
r
tm
e
nt
, P
a
n
-
A
t
la
nt
ic
U
ni
ve
r
s
it
y
,
L
a
g
o
s
, N
ig
e
r
ia
2
D
e
pa
r
tm
e
nt
of
C
o
mpu
te
r
S
c
ie
n
c
e
, U
ni
ve
r
s
it
y
of
B
e
ni
n, B
e
ni
n,
N
ig
e
r
ia
Ar
t
ic
l
e
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
i
ve
d
S
e
p
1
,
2021
R
e
vi
s
e
d
No
v
6
,
2021
A
c
c
e
pt
e
d
De
c
2
,
2021
D
ee
p
l
e
ar
n
i
n
g
m
o
d
e
l
s
h
a
v
e
b
ee
n
at
t
h
e
fo
r
e
fro
n
t
o
f
fac
i
a
l
reco
g
n
i
t
i
o
n
b
e
c
a
u
s
e
t
h
ey
d
e
l
i
v
e
r
i
m
p
ro
v
e
d
c
l
as
s
i
fi
c
at
i
o
n
a
cc
u
ra
cy
o
v
e
r
t
rad
i
t
i
o
n
a
l
o
n
e
s
.
Re
g
ar
d
l
e
s
s
,
d
ee
p
l
e
arn
i
n
g
mo
d
el
s
re
q
u
i
r
e
a
n
e
x
t
en
s
i
v
e
d
at
as
e
t
fo
r
t
rai
n
i
n
g
.
T
o
s
i
g
n
i
fi
c
a
n
t
l
y
cu
t
d
o
w
n
o
n
i
t
s
t
rai
n
i
n
g
t
i
me
an
d
d
at
as
e
t
v
o
l
u
me
,
p
re
-
t
rai
n
ed
mo
d
el
s
,
h
av
e
b
een
u
s
ed
al
t
h
o
u
g
h
,
t
h
ey
ar
e
s
t
i
l
l
re
q
u
i
r
e
d
t
o
u
n
d
e
rg
o
t
h
e
u
s
u
a
l
t
rai
n
i
n
g
p
ro
ce
s
s
fo
r
c
u
s
t
o
m
faci
al
r
ec
o
g
n
i
t
i
o
n
t
as
k
s
.
T
h
i
s
re
s
e
ar
c
h
fo
cu
s
e
s
o
n
a
n
i
m
p
ro
v
ed
fa
c
i
a
l
r
ec
o
g
n
i
t
i
o
n
s
y
s
t
em
t
h
at
l
ac
k
s
t
h
e
t
rai
n
i
n
g
an
d
re
t
rai
n
i
n
g
r
e
q
u
i
remen
t
s
.
T
h
e
s
y
s
t
em
u
s
e
s
an
e
x
i
s
t
i
n
g
d
ee
p
l
e
ar
n
i
n
g
f
e
at
u
r
e
e
x
t
rac
t
i
o
n
mo
d
el
.
Fi
rs
t
,
a
u
s
e
r
s
t
an
d
s
b
e
fo
r
e
a
c
a
me
ra
-
en
ab
l
ed
s
y
s
t
em.
A
ft
e
r
t
h
at
,
t
h
e
u
s
e
r
s
u
p
p
l
i
e
s
a
u
n
i
q
u
e
i
d
en
t
i
fi
c
at
i
o
n
n
u
m
b
e
r
t
o
f
e
t
ch
a
c
o
rr
e
s
p
o
n
d
i
n
g
fa
ce
i
m
a
g
e
fro
m
t
h
e
d
at
ab
as
e
.
T
h
i
s
p
ro
ce
s
s
g
e
n
e
rat
e
s
t
w
o
face
f
e
at
u
r
e
v
ec
t
o
rs
.
O
n
e
fro
m
t
h
e
c
a
me
ra
a
n
d
t
h
at
re
t
ri
ev
e
d
fr
o
m
t
h
e
d
at
ab
as
e
.
T
h
e
co
s
i
n
e
d
i
s
t
an
ce
f
u
n
c
t
i
o
n
d
e
t
e
r
mi
n
e
s
t
h
e
s
i
mi
l
ar
i
t
y
v
a
l
u
e
o
f
t
h
e
s
e
v
ec
t
o
rs
.
W
h
en
t
h
e
c
o
s
i
n
e
d
i
s
t
an
ce
v
a
l
u
e
fa
l
l
s
b
el
o
w
a
s
e
t
t
h
re
s
h
o
l
d
,
t
h
e
face
i
s
reco
g
n
i
z
ed
an
d
a
cce
s
s
g
ran
t
e
d
.
I
f
t
h
e
c
o
s
i
n
e
d
i
s
t
an
ce
o
f
t
h
e
t
w
o
v
ec
t
o
rs
g
i
v
e
s
a
v
al
u
e
ab
o
v
e
t
h
i
s
t
h
re
s
h
o
l
d
,
acce
s
s
i
s
d
e
n
i
e
d
.
T
h
e
p
ro
p
o
s
e
d
mo
d
e
l
p
e
rf
o
r
m
s
s
at
i
s
fac
t
o
ri
l
y
o
n
p
u
b
l
i
c
l
y
a
v
ai
l
ab
l
e
d
at
as
e
t
s
.
K
e
y
w
o
r
d
s
:
De
e
p
l
e
a
r
ni
ng
F
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
F
e
a
t
ur
e
v
e
c
t
or
No
n
-
tr
a
i
na
bl
e
s
y
s
t
e
m
P
a
tt
e
r
n
m
a
t
c
hi
ng
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
cen
s
e.
C
or
r
e
s
pon
din
g
A
u
th
or
:
K
a
z
e
e
m
O
y
e
b
o
de
De
s
i
g
n
a
n
d
Ne
w
M
e
d
i
a
De
pa
r
t
m
e
n
t
,
P
a
n
-
A
t
l
a
n
t
i
c
Uni
ve
r
s
i
t
y
Km
52
L
e
kk
i
-
E
pe
E
x
p
y
,
105101,
L
a
go
s
,
Ni
ge
r
i
a
E
m
a
i
l
:
ka
z
e
e
m
k
z
@
g
m
a
il
.
c
o
m
1.
I
NT
RODU
C
T
I
ON
F
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
s
y
s
t
e
m
s
h
a
v
e
b
e
e
n
be
n
e
f
i
c
i
a
l
i
n
r
e
s
o
l
vi
ng
a
h
o
s
t
o
f
pr
o
bl
e
m
s
s
uc
h
a
s
i
de
n
t
i
t
y
v
e
r
i
f
i
c
a
t
i
o
n
[
1
]
,
[
2
]
,
f
a
c
i
a
l
e
x
p
r
e
s
s
i
o
n
i
de
n
t
i
f
i
c
a
t
i
o
n
[
3
]
,
f
a
c
i
a
l
e
m
ot
i
o
n
s
c
l
a
s
s
i
f
i
c
a
t
i
o
n
[
4
]
,
c
r
i
m
e
r
e
s
o
l
u
t
i
o
n
[
5
]
,
v
e
hi
c
l
e
s
e
c
ur
i
t
y
[
6]
,
[
7]
a
n
d
h
o
s
t
o
f
ot
h
e
r
s
.
T
h
e
r
a
t
e
o
f
k
i
d
n
a
pp
i
ng
s
i
n
N
i
ge
r
i
a
n
s
c
h
o
o
l
s
ha
s
i
nc
r
e
a
s
e
d
o
wi
n
g
to
a
l
a
c
k
o
f
s
e
c
ur
i
t
y
f
r
a
m
e
wo
r
ks
.
V
i
s
i
t
o
r
s
c
a
n
g
a
in
a
c
c
e
s
s
w
i
t
h
o
ut
pr
o
pe
r
c
h
e
c
ks
.
I
ns
e
c
ur
i
t
y
i
n
s
c
h
o
o
l
s
c
o
u
l
d
h
a
v
e
d
e
v
a
s
ta
t
i
n
g
e
f
f
e
c
t
s
on
t
h
e
gi
r
l
c
hi
l
d
e
duc
a
t
i
o
n
[
8
]
,
[
9
]
a
n
d
c
oul
d
i
m
pa
c
t
th
e
i
r
a
c
a
de
m
i
c
p
e
r
f
or
m
a
nc
e
s
[
1
0
]
.
A
c
c
o
r
d
i
n
g
t
o
a
s
o
ur
c
e
i
n
[
11]
,
a
b
o
u
t
13.
5
m
il
li
o
n
c
hil
dr
e
n
a
r
e
o
ut
o
f
s
c
h
o
o
l
o
w
i
n
g
t
o
i
ns
e
c
ur
i
t
y
.
F
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
s
y
s
t
e
m
s
c
o
ul
d
h
e
l
p
m
i
t
i
ga
t
e
t
hi
s
c
h
a
l
l
e
n
ge
.
On
e
o
f
s
u
c
h
i
s
t
h
e
l
o
c
a
l
bi
n
a
r
y
p
a
tt
e
r
n
(
L
B
P
)
[
12
]
,
[
1
3
]
.
T
h
e
L
B
P
u
s
e
s
t
h
e
b
i
n
a
r
y
r
e
p
r
e
s
e
n
ta
t
i
o
n
to
r
e
p
r
e
s
e
n
t
th
e
tex
tur
e
p
r
ope
r
t
i
e
s
o
f
f
a
c
e
i
m
a
ge
s
.
A
n
i
m
a
ge
i
s
s
ub
-
d
i
vi
de
d
i
n
to
c
e
l
l
s
(
s
a
y
20
by
2
0
pi
xe
l
s
)
t
h
e
n
e
a
c
h
p
i
x
e
l
'
s
n
e
i
gh
b
or
i
n
g
p
r
ope
r
t
i
e
s
a
r
e
e
x
a
m
i
n
e
d
.
A
c
o
m
pa
r
i
s
o
n
b
e
t
w
e
e
n
e
a
c
h
p
i
x
e
l
r
i
n
a
c
o
n
s
i
de
r
e
d
c
e
l
l
a
n
d
i
t
s
e
i
gh
t
s
u
r
r
oun
d
i
n
g
pi
x
e
l
s
i
s
c
a
r
r
i
e
d
ou
t.
W
h
e
r
e
r
i
n
ten
s
i
t
y
v
a
l
ue
i
s
g
r
e
a
te
r
t
h
a
n
a
s
u
r
r
oun
d
i
n
g
p
i
x
e
l
v
,
v
t
h
e
n
h
o
l
ds
a
b
i
n
a
r
y
v
a
l
ue
z
e
r
o
oth
e
r
wi
s
e
,
v
h
o
l
ds
v
a
l
u
e
o
n
e
.
T
h
e
bi
n
a
r
y
v
a
l
ue
s
o
f
t
h
e
8
-
pi
x
e
l
n
e
i
ghb
or
s
o
f
r
a
r
e
c
o
l
l
e
c
ted
a
n
d
tr
a
n
s
f
o
r
m
e
d
i
n
to
a
s
e
que
n
c
e
o
f
n
um
be
r
s
.
T
h
e
n
u
m
b
e
r
c
o
m
bi
n
a
t
i
o
n
s
f
r
o
m
i
n
d
i
vi
dua
l
c
e
l
l
s
ov
e
r
t
h
e
e
n
t
i
r
e
i
m
a
ge
a
r
e
p
r
o
j
e
c
t
e
d
i
n
to
a
hi
s
tog
r
a
m
o
f
f
e
a
t
u
r
e
s
.
Al
s
o,
a
tr
a
n
s
f
o
r
m
a
t
i
o
n
i
s
c
a
r
r
i
e
d
ou
t
on
th
e
f
a
c
e
to
b
e
r
e
c
ogn
i
z
e
d
(
g
)
i
n
to
a
hi
s
tog
r
a
m
o
f
f
e
a
tu
r
e
s
.
A
c
o
m
pa
r
i
s
o
n
i
s
m
a
de
b
e
tw
e
e
n
t
h
e
f
e
a
tu
r
e
hi
s
tog
r
a
m
o
f
t
h
e
tr
a
i
ni
ng
d
a
tas
e
t
a
n
d
t
h
a
t
o
f
g
.
T
h
e
E
uc
l
i
de
a
n
di
s
tan
c
e
s
o
f
g
to
v
a
r
i
o
us
hi
s
tog
r
a
m
s
o
f
t
h
e
tr
a
i
ni
n
g
da
tas
e
t
a
r
e
j
uxta
p
os
e
d.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
989
-
99
4
990
T
h
e
s
m
a
l
l
e
s
t
E
uc
l
i
de
a
n
d
i
s
t
a
n
c
e
to
g
i
s
c
o
n
s
i
de
r
e
d
th
e
r
e
c
ogni
z
e
d
f
a
c
e
[
12
]
.
T
h
e
p
r
i
n
c
i
pa
l
c
o
m
p
o
n
e
n
t
a
n
a
l
y
s
i
s
(
P
C
A
)
[
1
4
]
,
[
1
5]
i
s
a
n
ot
h
e
r
f
a
c
i
a
l
r
e
c
ogn
i
t
i
o
n
m
o
d
e
l
a
n
d
w
or
ks
by
m
a
xi
m
i
z
i
n
g
t
h
e
v
a
r
i
a
n
c
e
a
m
o
n
gs
t
c
l
a
s
s
e
s
o
f
d
a
ta
.
P
C
A
e
x
tr
a
c
t
s
uni
que
f
e
a
tu
r
e
s
f
r
o
m
a
c
o
l
l
e
c
t
i
o
n
o
f
i
m
a
ge
s
a
n
d
p
r
o
j
e
c
t
s
t
h
e
m
o
n
to
a
gi
v
e
n
f
a
c
e
s
pa
c
e
.
F
or
f
r
o
n
ta
l
f
a
c
e
r
e
c
ogni
t
i
o
n
,
t
h
e
P
C
A
h
a
s
t
h
e
a
dv
a
n
t
a
ge
o
f
i
t
s
r
a
pi
d
c
o
m
pu
ta
t
i
o
n
[
1
4
]
.
T
h
e
f
i
s
h
e
r
s
l
i
n
e
a
r
di
s
c
r
i
m
i
n
a
n
t
a
n
a
l
y
s
i
s
(
L
DA
)
[
16
]
,
[
17
]
i
s
a
n
ot
h
e
r
m
ode
l
t
h
a
t
s
h
a
r
e
s
a
s
i
m
i
l
a
r
p
r
o
s
pe
c
t
wi
t
h
t
h
e
P
C
A
f
or
f
a
c
i
a
l
r
e
c
ogn
i
t
i
o
n
.
I
t
m
a
xi
mi
z
e
s
t
h
e
di
s
tan
c
e
s
a
m
o
n
gs
t
d
a
t
a
c
l
us
te
r
s
whi
l
e
e
n
s
u
r
i
n
g
t
h
a
t
e
a
c
h
g
r
oup
i
s
t
i
gh
t
l
y
pa
c
ke
d
.
R
e
c
e
n
t
l
y
,
t
h
e
a
va
i
l
a
bil
i
t
y
o
f
c
h
e
a
p
y
e
t
po
we
r
f
u
l
c
o
m
put
i
n
g
c
a
pa
bil
i
t
y
h
a
s
m
a
d
e
de
e
p
l
e
a
r
ni
ng
m
o
de
l
s
a
t
t
r
a
c
t
i
v
e
f
o
r
s
o
l
vi
n
g
f
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
pr
o
bl
e
m
s
.
I
n
a
dd
i
t
i
o
n
,
i
t
s
a
c
c
ur
a
c
y
l
e
ve
l
s
i
g
nif
i
c
a
n
t
l
y
o
u
t
pe
r
f
o
r
m
s
t
h
e
m
o
de
l
s
pr
e
vi
o
us
ly
d
i
s
c
us
s
e
d.
How
e
v
e
r
,
t
h
e
y
h
a
v
e
s
im
il
a
r
c
h
a
r
a
c
t
e
r
i
s
t
i
c
s
-
t
h
e
y
b
o
t
h
un
de
r
go
t
r
a
i
ni
ng.
M
o
r
e
i
m
po
r
t
a
n
t
l
y
,
t
h
e
c
o
nv
o
l
ut
i
o
n
n
e
ur
a
l
n
e
t
wo
r
k
(
C
NN
)
[
18
]
,
[
19]
n
e
e
ds
m
o
r
e
t
r
a
i
ni
ng
da
t
a
s
e
t
s
f
o
r
i
t
s
m
o
de
l
t
o
pe
r
f
o
r
m
o
pt
i
m
a
ll
y
[
20]
.
I
t
s
h
e
a
vy
r
e
li
a
n
c
e
o
n
l
o
t
s
o
f
t
r
a
i
ni
ng
im
a
ge
s
h
a
s
a
l
s
o
be
e
n
s
o
l
ve
d
pa
r
t
l
y
by
i
m
a
ge
a
ug
m
e
n
t
a
t
i
o
n
w
i
t
h
t
h
e
a
i
m
o
f
i
m
pr
o
v
e
d
pe
r
f
o
r
m
a
n
c
e
[
21]
.
T
hi
s
de
ve
l
o
p
m
e
n
t
c
o
ul
d
b
e
pe
r
c
e
i
v
e
d
a
s
a
d
own
s
i
de
a
l
t
h
o
u
gh
,
p
r
e
-
tr
a
i
n
e
d
m
ode
l
s
m
i
t
i
ga
t
e
th
e
r
e
l
i
a
n
c
e
o
n
l
a
r
ge
tr
a
i
ni
n
g
da
tas
e
t
s
.
F
or
e
x
a
m
p
l
e
,
i
n
[
2
2
]
,
t
h
e
pr
e
-
t
r
a
i
n
e
d
m
o
de
l
i
s
us
e
d
w
i
t
h
a
f
a
c
e
c
l
a
s
s
if
i
c
a
t
i
o
n
l
a
y
e
r
.
T
hi
s
a
ppr
o
a
c
h
pr
o
vi
de
s
t
h
e
po
s
s
i
bil
i
t
y
o
f
us
i
n
g
o
nl
y
a
f
e
w
t
r
a
i
ni
ng
im
a
ge
s
i
n
a
de
p
l
o
y
e
d
e
nvi
r
o
nm
e
n
t
.
I
n
a
s
c
e
n
a
r
i
o
whe
r
e
t
h
e
f
a
c
e
s
t
o
b
e
r
e
c
o
gn
i
z
e
d
a
r
e
kn
o
wn
i
n
a
d
v
a
n
c
e
,
t
hi
s
a
ppr
o
a
c
h
wo
ul
d
wo
r
k
we
l
l.
Ho
we
v
e
r
,
i
n
a
n
e
nvi
r
o
nm
e
n
t
t
h
a
t
a
tt
r
a
c
t
s
a
l
l
k
i
nd
s
o
f
n
e
w
f
a
c
e
s
t
h
a
t
n
e
e
d
to
b
e
gr
a
n
t
e
d
a
c
c
e
s
s
i
n
t
o
a
f
a
c
i
li
t
y
,
t
hi
s
a
ppr
o
a
c
h
i
n
[
22]
m
a
y
n
o
t
b
e
a
ppr
o
pr
i
a
t
e
.
T
h
e
r
e
a
s
o
n
i
s
t
h
a
t
t
h
e
m
o
de
l
h
a
s
t
o
un
de
r
go
a
r
e
tr
a
i
ni
ng
p
r
o
c
e
s
s
o
n
n
e
w
f
a
c
e
s
.
T
hi
s
de
v
e
l
o
p
m
e
n
t
m
a
ke
s
i
t
s
de
p
l
o
y
m
e
n
t
im
pr
a
c
t
i
c
a
l
i
n
r
e
a
l
-
li
f
e
s
c
e
n
a
r
i
o
s
s
u
c
h
a
s
s
c
h
o
o
l
e
nvi
r
o
nm
e
n
t
s
wh
e
r
e
vi
s
i
t
o
r
s
r
e
qui
r
e
s
c
r
e
e
ni
n
g
i
n
r
e
a
l
-
t
i
m
e
t
o
ga
i
n
a
c
c
e
s
s
t
o
s
c
h
o
o
l
f
a
c
il
i
t
i
e
s
.
2.
T
HE
P
ROP
OS
E
D
M
E
T
HO
D
T
o
m
i
t
i
g
a
t
e
t
hi
s
c
h
a
ll
e
n
ge
o
f
t
r
a
i
ni
ng
a
n
d
r
e
t
r
a
i
ni
n
g
f
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
s
y
s
t
e
m
s
,
we
pr
o
po
s
e
a
m
o
de
l
a
n
c
h
o
r
e
d
o
n
a
C
NN
a
r
c
hi
t
e
c
t
ur
e
s
e
e
n
i
n
[
23]
.
T
h
i
s
a
r
c
hi
t
e
c
t
ur
e
i
s
t
r
a
i
n
e
d
o
n
2.
6
m
il
li
o
n
f
a
c
e
im
a
ge
s
.
On
e
a
dv
a
n
t
a
ge
o
f
t
hi
s
a
r
c
hi
t
e
c
t
ur
e
i
s
t
h
a
t
i
t
c
a
n
b
e
a
da
pt
e
d
to
ge
n
e
r
a
t
e
f
e
a
t
ur
e
v
e
c
t
o
r
s
o
f
f
a
c
e
s
a
t
i
t
s
o
u
tput
l
a
y
e
r
wh
e
n
a
f
a
c
e
im
a
ge
i
s
pa
s
s
e
d
i
n
t
o
i
t
s
i
nput
l
a
y
e
r
.
T
h
e
pr
o
po
s
e
d
wor
kf
l
o
w
h
a
s
t
h
r
e
e
s
t
a
ge
s
.
T
h
e
f
ir
s
t
i
s
t
h
e
f
a
c
i
a
l
r
e
g
i
s
t
r
a
t
i
o
n
s
t
a
ge
.
He
r
e
,
o
nl
y
a
s
i
n
g
l
e
p
h
o
to
o
f
a
s
t
ude
n
t
o
r
a
vi
s
i
t
o
r
i
s
c
a
p
t
ur
e
d
a
n
d
s
t
o
r
e
d
w
i
t
h
a
s
t
ude
n
t
o
r
vi
s
i
t
o
r
'
s
i
de
n
t
i
t
y
n
u
m
be
r
i
n
a
c
e
n
t
r
a
l
da
t
a
b
a
s
e
.
S
e
c
o
n
d
l
y
,
i
n
t
h
e
r
e
c
o
gni
t
i
o
n
s
t
a
ge
,
t
h
e
h
u
man
b
e
f
o
r
e
t
h
e
c
a
m
e
r
a
s
upp
li
e
s
t
h
e
i
de
n
t
i
f
i
c
a
t
i
o
n
n
u
m
be
r
.
T
hi
s
i
de
n
t
i
f
i
c
a
t
i
o
n
n
u
m
be
r
s
e
r
ve
s
t
o
r
e
t
r
i
e
v
e
a
m
a
t
c
h
e
d
f
a
c
e
s
to
r
e
d
i
n
t
h
e
da
t
a
b
a
s
e
.
T
hi
r
d
ly
,
a
f
e
a
t
ur
e
v
e
c
t
o
r
I
i
s
ge
n
e
r
a
t
e
d
f
r
o
m
t
h
e
f
a
c
e
s
r
e
t
r
i
e
v
e
d
vi
a
t
h
e
da
t
a
b
a
s
e
.
I
n
a
dd
i
t
i
o
n
,
a
n
o
t
h
e
r
f
e
a
t
ur
e
v
e
c
t
or
n
a
m
e
d
K
i
s
ge
ne
r
a
t
e
d
f
r
o
m
t
h
e
p
h
o
to
c
a
pt
u
r
e
d
vi
a
t
h
e
a
c
c
e
s
s
c
o
n
t
r
o
l
c
a
m
e
r
a
.
A
c
o
m
pa
r
i
s
o
n
i
s
m
a
de
b
e
t
we
e
n
t
h
e
s
e
f
e
a
t
ur
e
v
e
c
t
o
r
s
us
i
n
g
a
c
o
s
i
n
e
s
i
mi
l
a
r
i
t
y
m
e
a
s
ur
e
[
24]
.
A
th
r
e
s
h
o
l
d
v
a
l
ue
o
f
0.
4
c
l
a
s
s
if
i
e
s
t
h
e
i
n
put
ph
oto
p
r
o
vi
de
d
by
t
h
e
c
a
m
e
r
a
.
A
t
a
di
s
t
a
n
c
e
,
n
o
t
m
o
r
e
t
h
a
n
0.
4,
th
e
m
o
de
l
r
e
t
u
r
n
s
‘
t
h
e
f
a
c
e
pr
e
s
e
n
t
e
d
i
s
r
e
c
o
gni
z
e
d
’
.
I
n
a
dd
i
t
i
o
n
,
i
f
t
h
e
m
o
de
l
r
e
t
ur
n
s
a
v
a
l
ue
o
f
m
o
r
e
t
h
a
n
0.
4,
a
n
o
u
t
pu
t
'
t
h
e
f
a
c
e
pr
e
s
e
n
t
e
d
i
s
n
o
t
r
e
c
o
gni
z
e
d
'
i
s
ge
n
e
r
a
t
e
d.
F
i
gur
e
1
e
x
e
m
p
li
f
i
e
s
t
h
e
r
e
g
i
s
t
r
a
t
i
o
n
pr
o
c
e
s
s
.
F
i
gur
e
1.
F
a
c
e
c
a
pt
ur
i
n
g
ph
a
s
e
F
i
gur
e
2
de
t
a
i
l
s
t
h
e
f
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
us
e
c
a
s
e
.
T
h
i
s
pr
o
c
e
s
s
r
e
qu
i
r
e
s
t
h
a
t
t
h
e
us
e
r
e
n
t
e
r
s
a
n
i
de
n
t
i
f
i
c
a
t
i
o
n
n
u
m
be
r
i
n
t
o
t
h
e
s
y
s
t
e
m
.
Af
t
e
r
t
h
a
t
,
t
h
e
i
de
n
t
i
f
i
c
a
t
i
o
n
s
e
r
ve
s
t
o
r
e
t
r
i
e
v
e
a
m
a
t
c
h
e
d
f
a
c
e
i
n
t
h
e
da
t
a
b
a
s
e
.
T
h
e
f
e
a
t
ur
e
v
e
c
t
o
r
o
f
t
h
e
f
a
c
e
r
e
t
r
i
e
v
e
d
is
t
h
e
n
c
o
m
p
a
r
e
d
w
i
t
h
t
h
a
t
f
e
t
c
h
e
d
f
r
o
m
t
h
e
c
a
m
e
r
a
.
I
f
t
h
e
r
e
i
s
a
m
a
t
c
h
,
t
h
e
s
y
s
t
e
m
gr
a
n
t
s
a
c
c
e
s
s
i
f
o
t
h
e
r
w
i
s
e
,
a
c
c
e
s
s
i
s
a
ll
o
we
d.
T
hi
s
pr
o
c
e
s
s
i
s
f
a
s
t
a
n
d
pr
a
c
t
i
c
a
l
a
s
i
t
r
e
qu
i
r
e
s
n
o
tr
a
i
ni
ng
a
n
d
a
r
e
t
r
a
i
ni
ng
pr
o
c
e
s
s
.
Algo
r
i
t
hm
1
a
l
s
o
s
h
o
ws
t
h
e
i
nn
e
r
wo
r
k
i
n
gs
o
f
t
h
e
pr
o
f
f
e
r
e
d
m
o
de
l
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
A
fas
t
and
non
-
tr
ainabl
e
f
ac
ial
r
e
c
ognit
ion
s
y
s
tem
f
or
s
c
hools
…
(
K
a
z
e
e
m
Oy
e
bode
)
991
F
i
gur
e
3
pr
o
vi
de
s
f
ur
t
h
e
r
i
n
s
i
g
h
t
i
n
to
v
a
r
i
o
us
c
om
po
n
e
n
t
s
o
f
t
h
e
pr
o
f
f
e
r
e
d
a
l
go
r
i
t
hm
.
T
he
s
y
s
t
e
m
f
u
n
c
t
i
o
n
a
li
t
i
e
s
a
r
e
i
n
t
wo
pa
r
t
s
-
t
h
e
y
a
r
e
t
h
e
f
r
o
n
t
e
n
d
a
n
d
t
h
e
b
a
c
ke
n
d.
T
he
f
r
o
n
t
e
n
d
h
a
s
t
wo
uni
t
s
-
t
h
e
y
a
r
e
t
h
e
a
d
m
i
n
u
ni
t
a
n
d
t
h
e
us
e
r
u
ni
t
.
T
h
e
a
d
m
i
n
u
nit
e
n
a
bl
e
s
s
t
ude
n
t
s
/vi
s
i
t
o
r
s
r
e
g
i
s
t
r
a
t
i
o
n
w
hil
e
t
h
e
us
e
r
u
ni
t
f
a
c
il
i
t
a
t
e
s
t
h
e
a
ut
h
e
n
t
i
c
a
t
i
o
n
pr
o
c
e
s
s
.
T
h
e
b
a
c
ke
n
d
s
a
v
e
s
u
s
e
r
da
t
a
a
n
d
a
l
s
o
c
o
nv
e
r
t
s
f
a
c
e
s
i
n
t
o
f
e
a
t
ur
e
v
e
c
t
o
r
s
.
T
h
e
c
o
s
i
n
e
d
i
s
t
a
n
c
e
f
u
n
c
t
i
o
n
t
h
e
n
c
o
m
pa
r
e
s
t
h
e
s
e
ve
c
t
or
s
f
o
r
a
m
a
t
c
h
.
F
i
gur
e
2
.
F
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
p
h
a
s
e
Al
go
r
it
h
m
1.
A
f
a
s
t
f
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
a
c
c
e
s
s
c
o
n
t
r
o
l
s
y
s
t
e
m
f
o
r
s
c
h
o
o
l
s
1.
Inputs
:
F
, it represents the presented face before a camera
2.
Outputs
:
O
represents system output
‘face recognized’
or
‘face not recognized’
3.
Precondition
:
A
replica
of
Face
F
must
be
registered
in
the
database
with
a
n
identification number
N
4.
Begin:
5.
Initialize
6.
Retrieve face
from the camera and assign to
F.
7.
Convert
F
into a single feature vector
I
8.
The user also enters the identification number and assigned to
N
9.
Retrieve face
V
using
N
from the database.
10.
Convert
V
into a single feature vector
K
11.
G = cosine_distance (
I,K
)
12.
If
G >0.4
13.
O = “face not recognized”
14.
Else
15.
O = ‘face recognized’
16.
return
F
i
gur
e
3
.
T
h
e
a
r
c
hi
t
e
c
t
ur
e
o
f
t
h
e
pr
o
p
o
s
e
d
a
l
go
r
i
t
hm
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
989
-
99
4
992
3.
RE
S
E
AR
CH
M
E
T
HO
D
T
h
e
pr
o
po
s
e
d
m
o
de
l
i
s
t
e
s
t
e
d
o
n
t
w
o
publ
i
c
ly
a
v
a
il
a
bl
e
f
r
o
n
t
a
l
f
a
c
e
da
t
a
s
e
t
s
us
i
ng
a
n
i
7
l
a
pt
o
p
pr
o
c
e
s
s
o
r
wi
t
h
16GB
o
f
r
a
n
do
m
a
c
c
e
s
s
m
e
m
o
r
y
(
R
A
M
)
.
T
wo
publ
i
c
ly
a
v
a
il
a
bl
e
da
t
a
s
e
t
s
(
F
a
c
e
95
a
n
d
f
a
c
e
96)
[
25]
,
pr
o
vi
de
d
by
t
he
U
ni
ve
r
s
i
t
y
o
f
E
s
s
e
x
a
r
e
us
e
d
f
o
r
t
h
e
e
va
l
u
a
t
i
o
n
p
ha
s
e
.
F
a
c
e
95
h
a
s
72
uni
que
f
a
c
e
s
(
72
pe
r
s
o
n
s
)
,
a
n
d
e
a
c
h
f
a
c
e
h
a
s
20
f
r
o
n
t
a
l
im
a
ge
s
.
T
h
e
r
e
F
a
c
e
95
ha
s
a
t
ot
a
l
o
f
1,
440
f
a
c
e
s
.
O
f
t
h
e
s
e
i
m
a
ge
s
,
o
nl
y
72
im
a
ge
s
w
e
r
e
he
l
d
ba
c
k
(
o
n
e
p
i
c
t
u
r
e
pe
r
pe
r
s
o
n
)
a
s
i
m
a
ge
s
s
t
o
r
e
d
i
n
t
h
e
da
t
a
b
a
s
e
.
F
o
r
f
a
c
e
96
da
t
a
s
e
t,
i
t
h
a
s
151
u
ni
que
f
a
c
e
s
(
151
pe
r
s
o
n
s
)
.
E
a
c
h
u
ni
que
f
a
c
e
h
a
s
20
s
a
m
p
l
e
s
.
T
he
r
e
f
o
r
e
,
F
a
c
e
96
ha
s
a
tot
a
l
o
f
3020
f
a
c
e
s
.
O
f
t
h
e
s
e
im
a
ge
s
,
o
nl
y
1
51
im
a
ge
s
we
r
e
up
l
o
a
de
d
i
n
t
o
t
h
e
da
t
a
b
a
s
e
.
T
h
e
a
c
c
ur
a
c
y
m
e
t
r
i
c
i
n
1
e
v
a
l
ua
t
e
s
t
h
e
pr
o
p
o
s
e
d
m
o
de
l
.
A
(
%
)
=
.
.
(
1)
4.
RE
S
UL
T
S
A
ND
D
IS
CU
S
S
I
ON
T
a
bl
e
s
1
a
n
d
2
s
h
o
w
t
h
e
pe
r
f
o
r
m
a
n
c
e
s
o
f
t
h
e
pr
o
po
s
e
d
m
o
de
l
.
W
e
o
b
s
e
r
v
e
d
t
h
a
t
i
t
pe
r
f
o
r
m
e
d
b
e
t
t
e
r
t
h
a
n
a
ll
c
o
ns
i
de
r
e
d
m
o
de
l
s
.
Al
t
h
o
ugh
i
t
o
nl
y
h
a
d
a
r
e
f
e
r
e
n
c
e
im
a
ge
pe
r
pe
r
s
o
n
i
n
t
e
r
m
s
o
f
f
a
c
e
f
e
a
t
ur
e
v
e
c
t
or
c
o
m
pa
r
i
s
o
n
,
i
t
pe
r
f
o
r
m
e
d
b
e
t
t
e
r
t
h
a
n
o
t
h
e
r
s
.
T
hi
s
pe
r
f
o
r
m
a
nc
e
i
s
e
x
pe
c
t
e
d
s
i
n
c
e
t
h
e
de
v
e
l
o
p
m
e
nt
o
f
t
h
e
o
f
f
e
r
e
d
m
o
de
l
s
i
t
s
o
n
a
r
o
b
us
t
f
e
a
t
ur
e
e
x
t
r
a
c
t
i
o
n
m
o
de
l
.
Ot
h
e
r
m
o
de
l
s
a
r
e
s
e
n
s
i
t
i
ve
to
n
o
i
s
e
,
f
o
r
e
x
a
m
p
l
e
,
P
C
A
a
n
d
L
D
A
.
T
h
e
l
o
c
a
l
bi
na
r
y
p
a
tt
e
r
n
(
L
B
P
)
h
a
s
l
e
s
s
l
i
g
h
t
s
e
n
s
i
t
i
v
i
t
y
,
c
o
n
s
e
que
n
t
l
y
pe
r
f
o
r
m
s
b
e
t
t
e
r
.
T
o
f
ur
t
h
e
r
de
m
o
ns
t
r
a
t
e
t
h
e
r
o
b
us
t
n
e
s
s
o
f
t
h
e
pr
o
p
o
s
e
d
a
r
c
hi
t
e
c
t
ur
e
,
we
c
o
m
pa
r
e
a
p
i
c
t
ur
e
t
a
ke
n
in
2017
(
a
s
t
h
e
da
t
a
b
a
s
e
f
a
c
e
)
w
i
t
h
t
h
a
t
e
x
t
r
a
c
t
e
d
f
r
om
t
h
e
we
b
c
a
m
i
n
2021.
F
i
gur
e
4
de
p
i
c
t
s
t
h
e
ph
o
to
t
a
ke
n
i
n
2017
us
e
d
a
s
t
h
e
r
e
f
e
r
e
n
c
e
im
a
ge
s
t
o
r
e
d
i
n
t
h
e
da
tab
a
s
e
by
t
h
e
a
d
mi
n
i
s
t
r
a
tor
,
a
n
d
a
n
i
de
n
t
i
f
i
c
a
t
i
o
n
nu
m
b
e
r
o
f
011075
i
s
a
t
t
a
c
h
e
d
to
i
t
.
T
h
i
s
n
u
m
be
r
c
o
ul
d
a
l
s
o
s
e
r
v
e
a
s
a
s
t
ude
n
t
'
s
m
a
t
r
i
c
u
l
a
t
i
o
n
n
u
m
be
r
.
T
hi
s
pr
o
c
e
s
s
i
s
a
o
n
e
-
t
i
m
e
r
e
g
i
s
t
r
a
t
i
o
n
pr
o
c
e
s
s
.
T
h
e
n
e
x
t
i
s
t
h
e
r
e
c
o
gni
t
i
o
n
s
t
a
ge
s
e
e
n
i
n
F
i
gur
e
5,
w
h
e
r
e
t
h
e
s
t
ude
n
t
wo
ul
d
e
n
t
e
r
t
h
e
m
a
t
r
i
c
u
l
a
t
i
o
n
n
u
m
be
r
.
Af
t
e
r
t
h
a
t
,
t
h
e
f
a
c
e
f
e
a
t
ur
e
v
e
c
t
o
r
,
e
x
t
r
a
c
t
e
d
f
r
o
m
t
h
e
c
a
m
e
r
a
a
s
s
h
o
wn
in
F
i
gur
e
6,
i
s
c
o
m
pa
r
e
d
t
o
t
h
a
t
s
tor
e
d
i
n
t
h
e
da
t
a
ba
s
e
,
us
i
n
g
a
c
o
s
i
ne
f
u
nc
t
i
o
n
.
F
i
na
l
ly
,
i
f
t
h
e
c
o
s
i
ne
d
i
s
t
a
nc
e
f
u
n
c
t
i
o
n
g
i
v
e
s
a
v
a
l
ue
l
e
s
s
t
h
a
n
0.
4
a
s
s
h
o
w
n
i
n
F
i
gur
e
7,
t
h
e
f
a
c
e
i
s
i
n
d
e
e
d
r
e
c
o
gni
z
e
d.
T
hi
s
de
ve
l
o
p
m
e
n
t
a
l
s
o
s
h
o
ws
t
h
a
t
a
n
o
l
d
p
i
c
t
ur
e
o
f
a
s
t
ude
n
t
s
to
r
e
d
i
n
t
h
e
f
i
r
s
t
y
e
a
r
i
n
s
c
h
o
o
l
c
a
n
b
e
r
e
l
e
v
a
n
t
i
n
gr
a
n
t
i
ng
a
c
c
e
s
s
f
o
r
s
e
v
e
r
a
l
y
e
a
r
s
.
I
n
a
dd
i
t
i
o
n
,
i
t
s
h
o
ws
t
h
a
t
di
s
t
i
n
c
t
f
a
c
i
a
l
f
e
a
t
ur
e
s
do
n
ot
f
a
de
o
v
e
r
t
i
m
e
.
T
h
e
f
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
p
h
a
s
e
l
a
s
t
s
f
o
r
o
nl
y
f
i
ve
s
e
c
o
n
d
s
.
F
i
gur
e
8
s
h
o
ws
w
h
a
t
m
i
g
h
t
ha
ppe
n
i
f
a
n
o
t
h
e
r
pe
r
s
o
n
's
i
de
n
t
i
f
i
c
a
t
i
o
n
i
s
e
n
t
e
r
e
d
a
s
s
e
e
n
i
n
F
i
gu
r
e
9.
T
h
e
s
y
s
t
e
m
w
i
ll
d
e
ny
a
c
c
e
s
s
a
s
a
c
o
s
i
n
e
d
i
s
t
a
n
c
e
r
e
t
ur
n
s
a
v
a
l
ue
m
o
r
e
t
h
a
n
0.
4.
T
a
bl
e
1.
M
o
de
l
s
’
pe
r
f
o
r
m
a
n
c
e
o
n
f
a
c
e
95
M
o
de
l
A
(
%
)
P
C
A
80.3
L
D
A
92.7
L
B
P
89.8
P
r
o
p
o
s
e
d
99.7
T
a
bl
e
2
.
M
o
de
l
s
’
pe
r
f
o
r
m
a
n
c
e
s
o
n
f
a
c
e
96
M
o
de
l
A
(
%
)
P
C
A
62.1
L
D
A
77.5
L
B
P
82
P
r
o
p
o
s
e
d
98.6
F
i
gur
e
4.
R
e
f
e
r
e
n
c
e
i
m
a
ge
s
t
o
r
e
d
i
n
t
h
e
da
t
a
b
a
s
e
by
t
h
e
a
d
m
i
n
i
s
t
r
a
to
r
(
t
a
ke
n
i
n
2017)
.
A
n
i
de
n
t
i
t
y
n
u
m
be
r
o
f
011075
i
s
a
tt
a
c
h
e
d
to
t
hi
s
i
m
a
ge
F
i
gur
e
5.
T
h
e
us
e
r
e
n
t
e
r
s
t
h
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
n
u
m
be
r
i
n
t
h
e
s
y
s
t
e
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
A
fas
t
and
non
-
tr
ainabl
e
f
ac
ial
r
e
c
ognit
ion
s
y
s
tem
f
or
s
c
hools
…
(
K
a
z
e
e
m
Oy
e
bode
)
993
F
i
gur
e
6.
I
m
a
ge
c
a
pt
ur
e
d
by
t
h
e
s
y
s
t
e
m
vi
a
t
h
e
we
b
c
a
m
F
i
gur
e
7.
T
h
e
c
o
s
i
n
e
d
i
s
t
a
n
c
e
f
u
nc
t
i
o
n
c
o
m
pa
r
e
s
f
e
a
t
ur
e
v
e
c
t
or
s
a
n
d
r
e
t
u
r
n
s
a
v
a
l
ue
o
f
0.
3168.
T
hi
s
va
l
ue
i
s
l
e
s
s
t
h
a
n
t
h
e
s
e
t
t
h
r
e
s
h
o
l
d
v
a
l
u
e
o
f
0.
4
F
i
gur
e
8.
T
h
e
c
o
s
i
n
e
d
i
s
t
a
n
c
e
f
u
nc
t
i
o
n
c
o
m
pa
r
e
s
f
e
a
t
ur
e
v
e
c
t
or
s
a
n
d
r
e
t
u
r
n
s
a
v
a
l
ue
o
f
0.
6986.
T
hi
s
i
s
m
o
r
e
t
h
a
n
t
h
e
s
e
t
t
h
r
e
s
h
o
l
d
v
a
l
ue
o
f
0.
4
F
i
gur
e
9.
T
h
e
us
e
r
e
n
t
e
r
s
a
w
r
o
n
g
i
de
n
t
i
t
y
n
u
m
be
r
o
f
011076
5.
CONC
L
USI
ON
F
a
c
i
a
l
r
e
c
o
gni
t
i
o
n
s
y
s
t
e
m
s
a
r
e
c
r
uc
i
a
l
f
o
r
i
de
n
t
i
f
i
c
a
t
i
o
n
pur
po
s
e
s
a
n
d
c
a
n
p
l
a
y
a
c
r
i
t
i
c
a
l
r
o
l
e
i
n
gr
a
n
t
i
n
g
a
c
c
e
s
s
to
f
a
c
il
i
t
i
e
s
,
e
s
pe
c
i
a
ll
y
i
n
s
c
h
o
ol
s
.
E
xi
s
t
i
n
g
m
o
de
l
s
s
u
c
h
a
s
de
e
p
l
e
a
r
ni
ng
o
f
t
e
n
r
e
qui
r
e
t
r
a
i
ni
ng
a
n
d
r
e
t
r
a
i
ni
ng.
T
hi
s
a
ppr
o
a
c
h
m
a
y
wo
r
k
i
n
e
nvi
r
o
nm
e
n
t
s
w
h
e
r
e
t
h
e
f
a
c
e
s
to
b
e
r
e
c
o
gn
i
z
e
d
a
r
e
kn
o
wn
i
n
a
d
v
a
n
c
e
.
W
h
e
r
e
f
a
c
e
s
t
o
b
e
r
e
c
o
gni
z
e
d
a
r
e
n
o
t
kn
o
wn
i
n
a
d
v
a
n
c
e
,
t
h
e
pr
o
po
s
e
d
m
o
de
l
c
a
n
be
r
e
l
e
v
a
n
t
,
e
s
pe
c
i
a
ll
y
i
n
s
c
h
o
o
l
s
.
T
h
e
m
o
de
l
l
e
v
e
r
a
ge
s
i
de
n
t
i
t
i
e
s
a
s
s
i
g
ne
d
to
pe
r
s
o
n
s
f
o
r
t
h
e
r
e
tr
i
e
v
a
l
o
f
s
t
o
r
e
d
f
a
c
e
s
.
T
hi
s
de
v
e
l
o
p
m
e
n
t
m
a
ke
s
i
t
p
o
s
s
i
bl
e
t
o
c
o
m
pa
r
e
t
h
e
f
e
a
t
ur
e
v
e
c
t
or
o
f
t
h
e
r
e
t
r
i
e
v
e
d
f
a
c
e
t
o
t
h
a
t
e
x
tr
a
c
t
e
d
f
r
o
m
t
h
e
we
b
c
a
m
o
r
c
a
m
e
r
a
.
T
h
e
pr
o
f
f
e
r
e
d
s
o
l
ut
i
o
n
i
s
t
e
s
t
e
d
o
n
t
w
o
f
a
c
e
da
t
a
s
e
t
s
w
i
t
h
pr
o
m
i
s
i
ng
r
e
s
u
l
t
s
.
W
e
h
o
pe
t
h
a
t
i
t
w
il
l
b
e
i
ns
t
a
l
l
e
d
f
o
r
w
i
de
s
pr
e
a
d
us
e
i
n
N
i
ge
r
i
a
a
n
d
b
e
y
o
n
d.
Ho
we
ve
r
,
f
i
na
nc
i
ng
i
s
s
t
i
ll
a
bi
g
c
h
a
ll
e
n
ge
a
s
t
h
e
i
ni
t
i
a
l
s
e
t
-
up
c
o
s
t
s
o
f
t
h
e
c
o
m
po
n
e
n
t
s
c
o
ul
d
be
hi
g
h
.
RE
F
E
R
E
NC
E
S
[
1]
B
.
R
.
I
l
y
a
s
,
B
.
M
o
ha
mm
e
d,
M
.
K
ha
l
e
d
,
a
nd
K
.
M
il
o
ud,
"
E
nha
nc
e
d
F
a
c
e
R
e
c
o
gni
ti
o
n
S
y
s
t
e
m
B
a
s
e
d
o
n
D
e
e
p
C
N
N
,
"
in
E
nha
nc
e
d
F
ac
e
R
e
c
ogni
ti
on
Sy
s
te
m
B
as
e
d
on
D
e
e
p
C
N
N
6t
h
I
nt
e
r
na
ti
onal
C
onf
e
r
e
nc
e
on
I
m
age
and
Si
gnal
P
r
oc
e
s
s
in
g
and
t
he
ir
A
ppl
ic
at
io
ns
(
I
S
P
A
)
, 24
-
25 N
ov
. 2019, pp. 1
-
6, d
o
i:
10.1109/I
S
P
A
48434.2019.8966797
.
[
2]
L
.
Y
ua
n,
Z
.
Q
u,
Y
.
Z
ha
o
,
H
.
Z
ha
ng
,
a
nd
Q
.
N
ia
n,
"
A
c
o
n
vo
lu
ti
o
na
l
n
e
ur
a
l
n
e
tw
o
r
k
ba
s
e
d
o
n
T
e
ns
or
F
l
o
w
f
o
r
f
a
c
e
r
e
c
o
gn
it
i
o
n,
"
2017,"
in
I
E
E
E
2nd
A
dv
anc
e
d
I
nf
or
m
at
io
n
T
e
c
hnol
ogy
,
E
le
c
t
r
oni
c
and
A
ut
om
at
io
n
C
ont
r
ol
C
onf
e
r
e
nc
e
(
I
A
E
A
C
)
,
C
ho
ngq
i
ng,
C
hi
na
. 25
-
26 M
a
r
c
h 2017, pp. 525
-
529, d
o
i:
10.1109
/I
A
E
A
C
.2017.8054070
.
[
3]
Y
.
W
a
ng,
Y
.
L
i,
Y
.
S
o
ng
,
a
nd
X
.
R
o
ng,
"
T
he
A
ppl
i
c
a
ti
o
n
of
a
H
y
br
id
T
r
a
ns
f
e
r
A
lg
o
r
it
h
m
B
a
s
e
d
o
n
a
C
o
n
vo
lu
ti
o
na
l
N
e
ur
a
l
N
e
tw
o
r
k
M
o
de
l
a
nd
a
n
I
mpr
ove
d
C
o
n
vo
lu
ti
o
n
R
e
s
tr
i
c
te
d
B
o
l
tz
ma
nn
M
a
c
hi
n
e
M
o
de
l
in
F
a
c
ia
l
E
x
pr
e
s
s
io
n
R
e
c
o
gni
ti
o
n,
"
I
E
E
E
A
c
c
e
s
s
,
v
o
l.
7, pp. 184599
-
184610, 2019, d
o
i:
10.1109
/AC
C
E
S
S
.2019.2961161
.
[
4]
A
.
A
ta
na
s
s
o
v
a
nd
D
. P
il
e
v
,
"
P
r
e
-
tr
a
in
e
d D
e
e
p
L
e
a
r
ni
ng
M
o
d
e
ls
f
o
r
F
a
c
ia
l
E
m
o
ti
o
ns
R
e
c
o
gni
ti
o
n,
"
i
n
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
A
ut
om
at
ic
s
and I
nf
or
m
at
ic
s
(
I
C
A
I
)
, V
a
r
na
B
ul
ga
r
ia
, 2020
, pp. 1
-
6, do
i
:
10.1109/I
C
A
I
50593.2020.9311334
.
[
5]
B
,
O
r
e
ll
a
na
,
L
,
Á
l
v
a
r
e
z
,
a
nd
J
.
A
r
ma
s
-
A
gui
r
r
e
,
"
F
a
c
e
R
e
c
o
gni
ti
o
n
f
or
C
r
im
in
a
l
I
d
e
nt
i
f
i
c
a
ti
o
n,
"
in
P
r
oc
e
e
di
ngs
of
th
e
6t
h
B
r
az
il
ia
n T
e
c
hnol
ogy
Sy
m
pos
iu
m
(
B
T
Sy
m
’
20)
. B
T
Sy
m
2020. Smar
t
I
nnov
at
io
n, Sy
s
te
m
s
and
T
e
c
hnol
ogi
e
s
, 2020.
[
6]
A
.
A
.
E
ln
ga
r
a
nd
M
.
K
a
y
e
d,
"
O
p
e
n
C
omput
e
r
S
c
ie
n
c
e
,
"
V
e
hi
c
le
Se
c
u
r
it
y
Sy
s
te
m
s
us
in
g
F
ac
e
R
e
c
ogni
ti
on
bas
e
d
on
I
nt
e
r
n
e
t
of
T
hi
ngs
,
vo
l.
10, n
o
. 1, pp. 17
-
29, 2020, d
o
i
:
10.1515/c
omp
-
202
0
-
0003
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
989
-
99
4
994
[
7]
R
.
F
.
R
a
hma
t,
M
.
P
.
L
oi
,
S
.
F
a
z
a
,
D
.
A
r
is
a
ndi
,
a
nd
R
.
B
udi
a
r
to
,
"
F
a
c
ia
l
R
e
c
o
gni
ti
o
n
f
or
C
a
r
S
e
c
ur
it
y
S
y
s
te
m
U
s
in
g
F
is
h
e
r
f
a
c
e
M
e
th
o
d,
"
i
n
J
our
nal
of
P
hy
s
ic
s
:
C
onf
e
r
e
nc
e
Se
r
i
e
s
, v
o
l.
1235,
no
. 1
,
2019, d
o
i:
10.1088/1742
-
6596/1235/
1/
012119.
[
8]
O
.
S
a
nni
,
"
E
f
f
e
c
ts
of
in
s
e
c
ur
it
y
a
nd
c
ha
ll
e
nge
s
o
n
f
e
ma
l
e
s
’
e
d
u
c
a
ti
o
n
in
N
ig
e
r
ia
,
"
A
f
r
ic
an
J
our
nal
f
o
r
th
e
P
s
y
c
hol
ogi
c
al
St
udy
of
Soc
ia
l
I
s
s
ue
s
,
v
o
l.
18, n
o
. 3, 2015.
[
9]
U
.
B
a
s
s
e
y
,
"
I
ns
e
c
ur
it
y
a
nd
g
ir
l
-
c
hi
ld
e
du
c
a
ti
o
n
in
N
ig
e
r
ia
,
"
E
ur
ope
an
J
our
nal
of
E
duc
at
io
n
St
udi
e
s
,
vo
l.
2,
n
o
.
11,
2
016,
do
i:
10.46827/
e
j
e
s
.
v
0i
0.405
.
[
10]
M
.
O
ju
kw
u
,
"
E
f
f
e
c
t
of
I
ns
e
c
u
r
it
y
of
S
c
h
oo
l
E
n
v
ir
o
nm
e
nt
o
n
th
e
A
c
a
de
m
ic
P
e
r
f
or
ma
nc
e
of
S
e
c
o
nda
r
y
S
c
h
oo
l
S
tu
d
e
nt
s
in
I
mo
S
ta
te
,
"
I
nt
e
r
nat
io
nal
J
our
nal
of
E
duc
at
io
n &
L
it
e
r
ac
y
St
udi
e
s
,
vo
l.
1, n
o
. 5, 2017, d
o
i:
10.7575/aia
c
.i
j
e
ls
.v
.5n.1p.20
.
[
11]
T
.
C
a
bl
e
,
"
T
h
e
C
a
bl
e
,
"
[
O
nl
in
e
]
.
A
v
a
il
a
bl
e
:
ht
tp
s
:/
/ww
w
.t
he
c
a
bl
e
.ng/
in
s
e
c
u
r
it
y
-
is
-
c
r
ip
p
li
ng
-
ni
g
e
r
ia
s
-
e
duc
a
ti
o
n
-
s
y
s
t
e
m
-
h
e
r
e
s
-
w
ha
t
-
we
-
mus
t
-
d
o
(
a
c
c
e
s
s
e
d A
ug
. 27,
2021
)
.
[
12]
S
.
Z
hu
a
nd
J
.
F
e
ng,
"
A
N
ove
l
F
a
c
ia
l
A
ppe
a
r
a
nc
e
D
e
s
c
r
ip
t
o
r
B
a
s
e
d
o
n
L
o
c
a
l
B
in
a
r
y
P
a
tt
e
r
n,
"
in
C
hi
ne
s
e
C
on
f
e
r
e
nc
e
on
P
at
t
e
r
n
R
e
c
ogni
ti
on
, B
e
ij
in
g, C
hi
na
, 22
-
24 O
c
t.
2008, pp. 1
-
4,
d
o
i:
10.1
109/
C
C
P
R
.2008.58.
[
13]
N
.
N
.
F
a
ti
ha
h,
G
.
A
r
i
y
a
nt
o
,
A
.
J
.
L
a
ti
pa
h
,
a
nd
D
.
M
.
P
a
nge
s
tu
t
y
,
"
F
a
c
e
R
e
c
o
gni
ti
o
n
U
s
in
g
L
oc
a
l
B
in
a
r
y
P
a
tt
e
r
n
a
nd
N
e
a
r
e
s
t
N
e
ig
hb
o
ur
C
la
s
s
if
i
c
a
ti
o
n,
"
in
I
nt
e
r
n
at
io
nal
Sy
m
pos
iu
m
on
A
dv
anc
e
d
I
nt
e
ll
ig
e
nt
I
nf
or
m
at
ic
s
(
SA
I
N
)
,
I
ndo
n
e
s
ia
,
29
-
30
A
ug.
2018,
pp. 142
-
147, do
i
:
10.1109/S
A
I
N
.2018.8673375.
[
14]
R
.
K
a
ur
a
nd
E
.
H
im
a
ns
hi
,
"
F
a
c
e
r
e
c
o
gni
ti
o
n
us
in
g
P
r
in
c
ip
a
l
C
o
mp
o
n
e
nt
A
na
l
y
s
is
,
"
in
I
E
E
E
I
nt
e
r
nat
io
nal
A
dv
anc
e
C
om
pu
ti
ng
C
onf
e
r
e
nc
e
(
I
A
C
C
)
, B
a
ngl
o
r
e
, 2015, pp. 585
-
589, d
o
i:
10.1109
/I
A
D
C
C
.2015.7154774.
[
15]
W
.
D
e
ng,
J
.
H
u,
J
.
L
u
,
a
nd
J
.
G
u
o
,
"
T
r
a
ns
f
o
r
m
-
I
n
v
a
r
ia
nt
P
C
A
:
A
U
ni
f
i
e
d
A
ppr
o
a
c
h
t
o
F
u
ll
y
A
ut
o
ma
ti
c
F
a
c
e
A
li
gnm
e
nt
,
R
e
pr
e
s
e
nt
a
ti
o
n,
a
nd
R
e
c
o
gni
ti
o
n,
"
I
E
E
E
T
r
ans
ac
ti
ons
on
P
at
te
r
n
A
nal
y
s
is
and
M
ac
hi
ne
I
nt
e
ll
ig
e
nc
e
,
vo
l.
36,
n
o.
6,
pp. 1275
-
1284, 2014, do
i
:
10.1109/
T
P
A
M
I
.2013.194
.
[
16]
M
.
B
o
b
e
r
,
K
.
K
uc
ha
r
s
ki
,
a
nd
W
S
ka
r
be
k,
“
F
a
c
e
R
e
c
o
gni
ti
o
n
b
y
F
is
he
r
a
nd
S
c
a
tt
e
r
L
in
e
a
r
D
is
c
r
im
in
a
nt
A
na
l
y
s
is
,
”
C
om
put
e
r
A
nal
y
s
is
of
I
m
age
s
and
P
at
te
r
ns
,
10t
h
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
,
G
r
o
ni
ng
e
n,
T
h
e
N
e
th
e
r
la
nds
,
A
ugus
t
25
-
27,
2003
,
do
i:
10.1007/978
-
3
-
540
-
45179
-
2_78
.
[
17]
W
.
Y
unz
hu
a
nd
C
.
Y
unl
i,
"
A
n
e
w
f
e
a
tu
r
e
e
x
tr
a
c
ti
o
n
a
lg
or
it
hm
ba
s
e
d
o
n
f
is
h
e
r
li
n
e
a
r
di
s
c
r
im
in
a
nt
a
na
l
y
s
is
,
"
in
th
e
3r
d
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
C
ont
r
ol
,
A
ut
om
at
io
n
and
R
obot
ic
s
(
I
C
C
A
R
)
,
24
-
26
A
pr
il
,
N
a
goy
a
,
J
a
pa
n
2017,
pp.
414
-
417,
do
i:
10.1109/I
C
C
A
R
.2017.7942729.
[
18]
S
.
A
lb
a
w
i,
T
.
A
.
M
o
ha
mm
e
d
,
a
nd
S
.
A
l
-
a
Z
a
w
i,
"
U
nd
e
r
s
ta
n
di
ng
of
a
c
o
n
vo
lu
ti
o
na
l
ne
u
r
a
l
ne
tw
o
r
k,
"
in
2017
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
E
ngi
ne
e
r
in
g
and
T
e
c
hnol
ogy
(
I
C
E
T
)
,
A
nt
a
l
y
a
,
T
ur
k
e
y
,
21
-
23
A
ug.
2017
pp.
1
-
6,
do
i:
10.1109/I
C
E
ng
T
e
c
hn
ol
.2017.8308186.
[
19]
T
.
Z
e
bi
n, P
. J
. S
c
ul
l
y
, N
. P
e
e
k, A
. J
. C
a
s
s
o
n
,
a
nd K
. B
. O
z
a
n
y
a
n, "
D
e
s
ig
n a
nd I
mpl
e
m
e
nt
a
ti
o
n
of
a
C
o
n
vo
lu
ti
o
na
l
N
e
ur
a
l
N
e
t
w
o
r
k
o
n a
n E
dg
e
C
o
mput
in
g S
ma
r
tp
h
o
ne
f
or
H
uma
n A
c
t
i
v
it
y
R
e
c
o
g
ni
ti
o
n,
"
I
E
E
E
A
c
c
e
s
s
,
v
o
l.
7, pp. 33509
-
133520, 2019.
[
20]
D
.
W
a
ng,
H
.
Y
u,
D
.
W
a
ng
,
a
nd
G
.
L
i,
"
F
a
c
e
R
e
c
o
gni
ti
o
n
S
ys
te
m
B
a
s
e
d
o
n
C
N
N
,
"
in
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
C
om
p
ut
e
r
I
nf
or
m
at
io
n
and
B
ig
D
at
a
A
ppl
ic
at
io
ns
(
C
I
B
D
A
)
,
G
ui
y
a
ng,
C
hi
na
,
17
-
19
A
pr
il
2020,
pp.
470
-
473,
do
i:
10.1109/C
I
B
D
A
50819.2020.00111.
[
21]
L.
P
e
ng
,
S
.
B
a
oy
e
,
a
nd
X
L
in
,
"
H
uma
n
f
a
c
e
r
e
c
o
gni
t
i
o
n
ba
s
e
d
o
n
c
o
n
vo
lu
t
i
o
na
l
ne
ur
a
l
ne
tw
or
k
a
nd
a
ugme
nt
e
d
da
ta
s
e
t,
"
Sy
s
te
m
s
Sc
ie
nc
e
&
C
ont
r
ol
E
ngi
ne
e
r
in
g,
v
ol
. 9, n
o
.
2, pp. 29
-
37, 2021,
do
i
:
10.1080/21642583.2
020.1836526.
[
22]
R
.
M
.
P
r
a
ka
s
h,
N
.
T
h
e
nm
oe
z
hi
,
a
nd
M
.
G
a
y
a
th
r
i,
"
F
a
c
e
R
e
c
ogni
ti
o
n
w
it
h
C
o
n
vo
lu
ti
o
na
l
N
e
ur
a
l
N
e
tw
o
r
k
a
nd
T
r
a
ns
f
e
r
L
e
a
r
ni
ng
I
nt
e
r
na
ti
o
na
l
C
o
n
f
e
r
e
n
c
e
o
n
S
ma
r
t
S
y
s
t
e
ms
a
nd
I
n
ve
nt
i
ve
T
e
c
hno
l
o
g
y
(
I
C
S
S
I
T
)
,
"
i
n
I
nt
e
r
nat
io
nal
C
onf
e
r
e
n
c
e
on
Sm
ar
t
Sy
s
t
e
m
s
and I
nv
e
nt
iv
e
T
e
c
hnol
ogy
(
I
C
SSI
T
)
,
T
ir
un
e
l
ve
li
, I
nd
ia
, 27
-
29 N
ov
. 2019,
pp. 861
-
864, d
o
i
:
10.1109/I
C
S
S
I
T
46314.2019.89878
99
.
[
23]
M.
P
.
O
mka
r
,
A
.
V
e
da
ld
i
,
a
nd
A
.
Z
is
s
e
r
ma
n,
"
D
e
e
p
F
a
c
e
R
e
c
o
gni
ti
o
n,
"
in
B
r
it
is
h
M
ac
hi
ne
V
is
io
n
C
on
f
e
r
e
nc
e
,
20
15.
do
i:
10.5244/C
.29.41
.
[
24]
S
.
O
nt
a
ñón,
"
A
n
ove
r
v
i
e
w
of
d
is
ta
nc
e
a
nd
s
im
il
a
r
it
y
f
un
c
ti
o
n
s
f
o
r
s
tr
u
c
tu
r
e
d
da
ta
,
"
A
r
ti
f
I
nt
e
ll
,
vo
l.
53,
p
p.
5309
-
5351,
20
20,
do
i:
d
o
i.
or
g/
10.1007/s
10462
-
020
-
09821
-
w.
[
25]
S
.
L
ib
o
r
,
"
F
a
c
e
R
e
c
o
gni
ti
o
n
D
a
ta
ba
s
e
,
"
[
O
nl
in
e
]
.
A
v
a
il
a
bl
e
:
ht
tp
s
:/
/c
s
pe
r
s
o
n.kku.a
c
.t
h/
c
ha
kc
ha
i/
f
a
c
e
D
B
w
e
b/
f
a
c
e
_da
ta
s
e
t.
h
tm
l
(a
c
c
e
s
s
e
d J
a
n
. 2,
2018
)
.
B
I
OG
RA
P
HI
E
S
OF
AU
T
HO
RS
Ka
z
ee
m
Oy
ebo
de
rece
i
v
e
d
t
h
e
Ph
.
D
d
eg
r
ee
i
n
i
m
a
g
e
p
ro
ce
s
s
i
n
g
an
d
co
m
p
u
t
er
v
i
s
i
o
n
fro
m
t
h
e
U
n
i
v
e
rs
i
t
y
o
f
K
w
aZ
u
l
u
-
N
at
al
,
So
u
t
h
A
fri
c
a
,
i
n
2
0
1
7
.
H
e
c
u
rr
en
t
l
y
l
ec
t
u
r
e
s
at
Pan
-
A
t
l
an
t
i
c
U
n
i
v
e
rs
i
t
y
,
L
ag
o
s
.
N
i
g
e
r
i
a.
H
e
c
an
b
e
co
n
t
ac
t
ed
at
em
ai
l
:
k
az
eem
k
z
@
g
m
a
i
l
.
c
o
m
.
Ki
n
g
s
l
ey
C
h
i
w
u
i
k
e
U
k
a
o
h
a
i
s
an
A
s
s
o
ci
at
e
Pr
o
f
e
s
s
o
r
o
f
Co
m
p
u
t
e
r
S
c
i
en
ce
an
d
t
h
e
D
e
p
u
t
y
D
i
r
ec
t
o
r
(
I
CT
),
C
en
t
re
fo
r
D
i
s
t
an
ce
L
e
ar
n
i
n
g
at
t
h
e
U
n
i
v
e
rs
i
t
y
o
f
B
e
n
i
n
,
N
i
g
e
r
i
a.
H
e
c
an
b
e
co
n
t
a
c
t
ed
at
em
a
i
l
:
k
i
n
g
s
l
ey
.
u
k
ao
h
a@
u
n
i
b
en
.
e
d
u
.
Evaluation Warning : The document was created with Spire.PDF for Python.