I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
,
p
p
.
5
0
0
8
~
5
0
1
6
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
14
.i
6
.
p
p
5
0
0
8
-
5
0
1
6
5008
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Recog
nition o
f
In
do
nesia
n sig
n lang
ua
g
e using
deep
lea
rning
:
co
nv
o
lutiona
l neural network
-
ba
se
d
a
ppro
a
ch
O
liv
ia
K
em
bu
a
n
1,
2
,
H
a
ry
a
nt
o
1,
3
,
M
o
cha
ma
d B
ruri
T
riy
o
no
1
1
D
o
c
t
o
r
a
l
P
r
o
g
r
a
m i
n
Te
c
h
n
o
l
o
g
y
a
n
d
V
o
c
a
t
i
o
n
a
l
E
d
u
c
a
t
i
o
n
,
U
n
i
v
e
r
si
t
a
s N
e
g
e
r
i
Y
o
g
y
a
k
a
r
t
a
,
Y
o
g
y
a
k
a
r
t
a
,
I
n
d
o
n
e
si
a
2
I
n
f
o
r
mat
i
c
s E
n
g
i
n
e
e
r
i
n
g
S
t
u
d
y
P
r
o
g
r
a
m
,
F
a
c
u
l
t
y
o
f
En
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
s
i
t
y
o
f
M
a
n
a
d
o
,
M
a
n
a
d
o
,
I
n
d
o
n
e
s
i
a
3
Ed
u
c
a
t
i
o
n
a
l
R
e
se
a
r
c
h
a
n
d
E
v
a
l
u
a
t
i
o
n
,
U
n
i
v
e
r
si
t
a
s N
e
g
e
r
i
Y
o
g
y
a
k
a
r
t
a
,
Y
o
g
y
a
k
a
r
t
a
,
I
n
d
o
n
e
s
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
28
,
2
0
2
4
R
ev
is
ed
Sep
6
,
2
0
2
5
Acc
ep
ted
Oct
16
,
2
0
2
5
Th
is
stu
d
y
f
o
c
u
se
s
o
n
d
e
v
e
lo
p
in
g
a
n
a
u
t
o
m
a
ti
c
In
d
o
n
e
sia
n
sig
n
lan
g
u
a
g
e
(S
IBI)
re
c
o
g
n
it
io
n
sy
ste
m
u
sin
g
a
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
r
k
(CNN
).
S
ig
n
lan
g
u
a
g
e
is
e
ss
e
n
ti
a
l
f
o
r
c
o
m
m
u
n
ica
ti
o
n
a
m
o
n
g
d
e
a
f
a
n
d
h
a
rd
-
of
-
h
e
a
rin
g
in
d
iv
i
d
u
a
ls,
a
n
d
a
u
t
o
m
a
ti
c
re
c
o
g
n
i
ti
o
n
h
e
lp
s
imp
r
o
v
e
a
c
c
e
ss
ib
il
it
y
a
n
d
i
n
c
lu
siv
i
ty
.
CNN
s
a
re
c
h
o
se
n
fo
r
t
h
e
ir
a
b
i
li
ty
to
lea
rn
ima
g
e
fe
a
tu
re
s
a
u
to
m
a
ti
c
a
ll
y
,
e
li
m
in
a
ti
n
g
m
a
n
u
a
l
e
x
trac
ti
o
n
a
n
d
imp
ro
v
i
n
g
c
la
ss
ifi
c
a
ti
o
n
a
c
c
u
ra
c
y
.
Th
e
S
IBI
d
a
tas
e
t
u
se
d
c
o
n
tain
s
5
,
2
8
0
ima
g
e
s
o
f
2
6
lette
rs,
d
iv
i
d
e
d
in
to
train
i
n
g
a
n
d
v
a
li
d
a
ti
o
n
se
ts.
In
e
a
rly
train
i
n
g
,
t
h
e
m
o
d
e
l
a
c
h
iev
e
d
l
o
w
a
c
c
u
ra
c
y
(3
.
6
3
%
train
in
g
,
3
.
3
3
%
v
a
li
d
a
ti
o
n
),
b
u
t
a
fter
fiv
e
e
p
o
c
h
s,
it
sig
n
ifi
c
a
n
t
ly
im
p
ro
v
e
d
t
o
9
7
.
5
8
%
fo
r
trai
n
in
g
a
n
d
1
0
0
%
f
o
r
v
a
l
id
a
t
io
n
.
K
ey
w
o
r
d
s
:
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
Dee
p
lear
n
in
g
I
m
ag
e
r
ec
o
g
n
itio
n
Neu
r
al
n
etwo
r
k
Sig
n
lan
g
u
a
g
e
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Oliv
ia
Kem
b
u
an
Do
cto
r
al
Pro
g
r
a
m
in
T
ec
h
n
o
lo
g
y
an
d
Vo
ca
tio
n
al
E
d
u
ca
tio
n
,
Un
iv
er
s
itas
Neg
er
i Y
o
g
y
ak
ar
t
a
Yo
g
y
ak
ar
ta,
I
n
d
o
n
esia
E
m
ail:
o
liv
iak
em
b
u
a
n
.
2
0
2
3
@
s
tu
d
en
t.u
n
y
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
I
m
ag
e
r
ec
o
g
n
itio
n
r
ef
er
s
to
th
e
p
r
o
ce
s
s
o
f
id
e
n
tify
in
g
an
d
c
ateg
o
r
izin
g
o
b
jects
with
in
an
im
ag
e.
T
h
is
tech
n
o
lo
g
y
s
u
p
p
o
r
ts
a
wid
e
r
an
g
e
o
f
ap
p
licatio
n
s
,
in
clu
d
in
g
f
ac
ial
r
ec
o
g
n
itio
n
,
au
t
o
n
o
m
o
u
s
v
eh
icles,
m
ed
ical
d
iag
n
o
s
tics
,
an
d
r
etail
an
al
y
tics
[
1
]
‒
[
6
]
.
T
h
e
f
ield
h
as
ad
v
an
ce
d
s
ig
n
if
ican
tly
d
u
e
to
in
cr
ea
s
in
g
co
m
p
u
tatio
n
al
p
o
wer
,
th
e
av
ailab
ilit
y
o
f
ex
te
n
s
iv
e
d
atasets
,
an
d
b
r
ea
k
th
r
o
u
g
h
s
in
m
ac
h
in
e
lear
n
in
g
.
T
h
is
tech
n
iq
u
e,
u
s
ed
in
co
m
p
u
ter
v
is
io
n
an
d
im
ag
e
p
r
o
ce
s
s
in
g
,
h
as
ev
o
lv
e
d
f
r
o
m
tr
ad
itio
n
al
m
ac
h
in
e
lea
r
n
in
g
m
eth
o
d
s
to
s
o
p
h
is
ticated
d
ee
p
lear
n
in
g
a
p
p
r
o
ac
h
es.
Var
io
u
s
m
eth
o
d
s
an
d
a
p
p
r
o
ac
h
es
h
a
v
e
b
ee
n
d
e
v
elo
p
ed
f
o
r
im
ag
e
class
if
icatio
n
,
r
an
g
in
g
f
r
o
m
tr
ad
itio
n
al
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es
to
a
d
v
an
ce
d
d
ee
p
lear
n
in
g
m
o
d
els.
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NNs
)
ar
e
th
e
co
r
n
er
s
to
n
e
o
f
m
o
d
er
n
im
a
g
e
r
ec
o
g
n
itio
n
s
y
s
tem
s
.
A
C
NN
is
a
ty
p
e
o
f
d
ee
p
lea
r
n
in
g
m
o
d
el
s
p
ec
if
ically
d
esi
g
n
ed
f
o
r
an
aly
zin
g
s
tr
u
ctu
r
ed
g
r
id
d
ata
s
u
ch
as
im
ag
es
[
7
]
‒
[
9
]
.
A
C
NN
is
a
m
ath
em
atica
l
c
o
n
s
tr
u
ct
th
at
g
e
n
er
ally
c
o
n
s
is
ts
o
f
t
h
r
ee
ty
p
es
o
f
lay
er
s
(
o
r
b
u
ild
i
n
g
b
lo
c
k
s
)
:
co
n
v
o
l
u
tio
n
,
p
o
o
lin
g
,
an
d
f
u
lly
co
n
n
e
cted
lay
er
s
.
T
h
e
f
ir
s
t
two
lay
er
s
,
th
e
co
n
v
o
lu
tio
n
an
d
p
o
o
lin
g
lay
er
s
,
p
er
f
o
r
m
f
ea
tu
r
e
ex
tr
ac
tio
n
,
w
h
ile
th
e
th
ir
d
lay
er
,
th
e
f
u
lly
c
o
n
n
ec
t
ed
lay
er
,
m
ap
s
th
e
ex
tr
ac
ted
f
ea
tu
r
es
in
to
t
h
e
r
es
u
lt,
s
u
ch
as
class
if
icatio
n
.
A
c
o
n
v
o
lu
ti
o
n
lay
e
r
p
la
y
s
an
im
p
o
r
tan
t
p
a
r
t
in
C
NN,
wh
ich
is
co
n
s
tr
u
cted
o
f
a
s
tac
k
o
f
m
at
h
em
atica
l
o
p
er
atio
n
s
,
s
u
ch
as
co
n
v
o
lu
tio
n
,
a
s
p
ec
ia
lized
s
o
r
t
o
f
lin
e
ar
o
p
er
atio
n
[
7
]
.
C
NNs
ar
e
p
ar
ticu
lar
ly
ef
f
ec
tiv
e
f
o
r
task
s
lik
e
im
ag
e
class
if
icatio
n
,
o
b
ject
d
etec
tio
n
,
an
d
im
ag
e
s
eg
m
en
tatio
n
d
u
e
to
t
h
eir
a
b
ilit
y
to
lear
n
s
p
atial
h
ier
ar
ch
ies
o
f
f
ea
tu
r
es
au
to
m
atica
lly
an
d
ad
ap
tiv
el
y
[
10]
‒
[
1
2
]
.
T
h
e
m
o
tiv
atio
n
f
o
r
u
s
in
g
C
NNs
in
im
ag
e
class
if
icatio
n
s
tem
s
f
r
o
m
t
h
eir
ab
ilit
y
to
a
u
to
m
atica
lly
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
R
ec
o
g
n
itio
n
o
f I
n
d
o
n
esia
n
s
ig
n
la
n
g
u
a
g
e
u
s
in
g
d
ee
p
lea
r
n
in
g
…
(
Olivia
K
emb
u
a
n
)
5009
an
d
ad
a
p
tiv
ely
lear
n
s
p
atial
h
ier
ar
ch
ies
o
f
f
ea
tu
r
es
th
r
o
u
g
h
b
ac
k
p
r
o
p
ag
atio
n
.
T
h
is
r
e
d
u
ce
s
th
e
n
ee
d
f
o
r
m
an
u
al
f
ea
tu
r
e
en
g
i
n
ee
r
in
g
an
d
s
ig
n
if
ican
tly
im
p
r
o
v
es c
lass
if
icatio
n
ac
cu
r
ac
y
.
I
n
p
ar
allel
with
ad
v
an
ce
m
en
ts
in
im
ag
e
r
ec
o
g
n
itio
n
,
s
ig
n
lan
g
u
ag
e
r
ec
o
g
n
itio
n
(
SLR)
h
as
em
er
g
ed
as
a
cr
itical
ap
p
licatio
n
o
f
d
ee
p
lear
n
in
g
,
aim
in
g
to
b
r
i
d
g
e
co
m
m
u
n
icatio
n
g
a
p
s
f
o
r
th
e
d
ea
f
an
d
h
ar
d
-
of
-
h
ea
r
in
g
c
o
m
m
u
n
ities
.
I
n
I
n
d
o
n
esia,
I
n
d
o
n
esian
s
ig
n
la
n
g
u
ag
e
s
y
s
tem
(
SIBI)
s
er
v
es
a
s
th
e
f
o
r
m
al
s
ig
n
lan
g
u
ag
e
u
s
ed
in
ed
u
ca
tio
n
al
an
d
g
o
v
e
r
n
m
en
tal
c
o
n
tex
ts
.
Desp
ite
its
s
tan
d
ar
d
ized
s
tatu
s
,
r
esear
ch
o
n
SIBI
r
ec
o
g
n
itio
n
r
em
ain
s
lim
ited
,
esp
ec
ially
in
ter
m
s
o
f
p
u
b
licly
av
ailab
le
d
atasets
an
d
d
ee
p
lear
n
in
g
m
o
d
els
tailo
r
ed
to
its
u
n
iq
u
e
lin
g
u
is
tic
ch
ar
ac
ter
is
tics
.
R
ec
en
t
s
tu
d
ies
h
av
e
b
e
g
u
n
to
ex
p
lo
r
e
th
e
a
p
p
licatio
n
o
f
d
ee
p
lear
n
in
g
tech
n
iq
u
es,
s
u
c
h
as C
NNs
an
d
h
y
b
r
id
m
o
d
els
f
o
r
r
ec
o
g
n
izi
n
g
b
o
th
s
tatic
an
d
d
y
n
am
ic
S
I
B
I
s
ig
n
s
.
Ho
wev
er
,
th
ese
ef
f
o
r
ts
ar
e
r
elativ
ely
m
o
d
est
in
s
ca
le
a
n
d
s
co
p
e.
I
n
co
n
tr
ast,
s
ig
n
lan
g
u
ag
es
s
u
c
h
as
Am
er
ican
s
ig
n
lan
g
u
ag
e
(
ASL)
[
1
3
]
,
I
n
d
ian
s
ig
n
lan
g
u
a
g
e
(
I
SL)
,
B
r
itis
h
s
ig
n
lan
g
u
ag
e
(
B
SL)
,
an
d
b
a
h
a
s
a
is
ya
r
a
t
I
n
d
o
n
esia
(
B
I
SIN
DO)
[
1
4
]
,
[
1
5
]
h
av
e
b
ee
n
s
tu
d
ied
m
o
r
e
ex
ten
s
iv
ely
.
ASL
h
as
r
ec
eiv
e
d
s
ig
n
if
ican
t
atten
tio
n
,
s
u
p
p
o
r
te
d
b
y
lar
g
e
-
s
ca
le
d
atasets
an
d
th
e
ad
o
p
tio
n
o
f
ad
v
a
n
ce
d
a
r
c
h
itectu
r
es
in
clu
d
in
g
C
NNs,
lo
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
(
L
S
T
M)
n
etwo
r
k
s
,
a
n
d
T
r
an
s
f
o
r
m
e
r
-
b
ased
m
o
d
els
f
o
r
b
o
th
is
o
lated
an
d
c
o
n
tin
u
o
u
s
s
ig
n
r
ec
o
g
n
itio
n
.
T
o
co
n
tex
tu
alize
th
e
cu
r
r
en
t
s
tu
d
y
,
T
ab
le
1
co
m
p
a
r
es
r
ec
en
t
ef
f
o
r
ts
ac
r
o
s
s
v
ar
io
u
s
s
ig
n
lan
g
u
ag
es,
s
u
m
m
ar
izin
g
th
e
k
e
y
co
n
tr
ib
u
tio
n
s
an
d
h
ig
h
lig
h
tin
g
th
e
n
o
v
elty
o
f
th
i
s
wo
r
k
in
ad
v
an
cin
g
SIBI
-
b
ased
r
ec
o
g
n
itio
n
.
T
ab
le
1
.
C
o
m
p
a
r
is
o
n
o
f
r
elate
d
s
tu
d
ies in
SLR
R
e
f
e
r
e
n
c
e
M
o
d
e
l
/
t
e
c
h
n
i
q
u
e
La
n
g
u
a
g
e
/
d
a
t
a
se
t
Ta
sk
t
y
p
e
D
a
t
a
s
e
t
p
r
o
p
e
r
t
i
e
s
A
c
c
u
r
a
c
y
N
o
t
e
s
[
1
4
]
C
N
N
+
LST
M
B
I
S
I
N
D
O
,
V
i
d
e
o
R
e
c
o
g
n
i
t
i
o
n
(
st
a
t
i
c
)
1
0
B
I
S
I
N
D
O
si
g
n
s
(
2
l
e
t
t
e
r
s
+
8
w
o
r
d
s)
,
7
2
0
p
v
i
d
e
o
u
se
d
f
o
r
t
e
st
i
n
g
C
N
N
:
9
6
%
a
c
c
u
r
a
c
y
/
1
8
%
l
o
ss
LSTM
:
8
6
%
a
c
c
u
r
a
c
y
/
4
1
%
l
o
ss
C
N
N
+
LST
M
:
9
6
%
a
c
c
u
r
a
c
y
/
1
7
%
l
o
ss
M
e
t
r
i
c
s
:
a
c
c
u
r
a
c
y
,
l
o
ss
,
w
o
r
d
e
r
r
o
r
r
a
t
e
(
W
ER
)
,
c
h
a
r
a
c
t
e
r
e
r
r
o
r
r
a
t
e
(
C
ER
[
1
5
]
H
i
d
d
e
n
M
a
r
k
o
v
mo
d
e
l
(
H
M
M
)
w
i
t
h
G
a
u
ss
i
a
n
d
e
n
si
t
i
e
s
B
I
S
I
N
D
O
D
a
t
a
a
c
q
u
i
si
t
i
o
n
u
si
n
g
M
i
c
r
o
s
o
f
t
K
i
n
e
c
t
X
b
o
x
(
sk
e
l
e
t
o
n
t
r
a
c
k
i
n
g
)
2
5
B
I
S
I
N
D
O
r
o
o
t
w
o
r
d
s
A
c
c
u
r
a
c
y
r
a
n
g
e
s
f
r
o
m
5
6
%
t
o
7
6
%
D
a
t
a
l
a
b
e
l
e
d
p
e
r
f
r
a
me
-
Tr
a
i
n
i
n
g
/
t
e
s
t
i
n
g
sp
l
i
t
u
si
n
g
K
-
F
o
l
d
(
K
=
1
0
)
[
1
6
]
3D
-
C
N
N
,
b
i
d
i
r
e
c
t
i
o
n
a
l
r
e
c
u
r
r
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
(
B
i
-
R
N
N
)
,
G
R
U
,
S
o
f
t
M
a
x
,
C
T
C
l
o
ss
S
I
B
I
S
e
q
u
e
n
c
e
-
to
-
seq
u
e
n
c
e
r
e
c
o
g
n
i
t
i
o
n
t
a
s
k
.
3
,
0
0
6
o
r
i
g
i
n
a
l
v
i
d
e
o
s
o
f
3
0
sen
t
e
n
c
e
s
i
n
S
I
B
I
A
v
e
r
a
g
e
W
E
R
a
c
r
o
ss
mo
d
e
l
s
:
8
8
.
7
9
%
C
o
m
b
i
n
e
d
3
D
-
C
N
N
(
f
o
r
sp
a
t
i
a
l
-
t
e
m
p
o
r
a
l
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
)
a
n
d
B
i
-
R
N
N
(
f
o
r
se
q
u
e
n
c
e
mo
d
e
l
i
n
g
)
[
1
7
]
C
N
N
A
S
L
S
t
a
t
i
c
s
i
g
n
l
a
n
g
u
a
g
e
a
l
p
h
a
b
e
t
r
e
c
o
g
n
i
t
i
o
n
-
1
.
P
u
b
l
i
c
D
a
t
a
s
e
t
1
:
5
2
,
0
0
0
i
m
a
g
e
s
2
.
P
u
b
l
i
c
D
a
t
a
s
e
t
2
:
6
2
,
4
0
0
i
m
a
g
e
s
3
.
C
u
st
o
m
A
S
LA
D
a
t
a
s
e
t
:
1
0
4
,
0
0
0
i
ma
g
e
s
D
a
t
a
s
e
t
1
[
18
]
:
a
c
c
u
r
a
c
y
=
9
9
.
4
1
%,
l
o
ss
=
0
.
0
2
0
4
D
a
t
a
s
e
t
2
[
19
]
:
a
c
c
u
r
a
c
y
=
9
9
.
4
8
%,
l
o
ss
=
0
.
0
2
1
0
A
S
LA
(
c
u
s
t
o
m
d
a
t
a
se
t
)
:
a
c
c
u
r
a
c
y
=
9
9
.
3
8
%
,
l
o
ss
=
0
.
0
2
5
0
C
a
p
t
u
r
e
d
w
i
t
h
l
a
p
t
o
p
/
smar
t
p
h
o
n
e
c
a
mer
a
s
Th
i
s
w
o
r
k
C
u
s
t
o
m C
N
N
S
I
B
I
,
2
6
l
e
t
t
e
r
s
(
st
a
t
i
c
)
S
t
a
t
i
c
S
i
g
n
r
e
c
o
g
n
i
t
i
o
n
5
,
2
8
0
i
m
a
g
e
s
o
f
2
6
l
e
t
t
e
r
s
9
7
.
5
8
%
f
o
r
t
r
a
i
n
i
n
g
a
n
d
1
0
0
%
f
o
r
v
a
l
i
d
a
t
i
o
n
.
N
e
w
d
a
t
a
se
t
;
h
i
g
h
a
c
c
u
r
a
c
y
a
f
t
e
r
t
r
a
i
n
i
n
g
A
s
ig
n
if
ican
t
p
o
r
tio
n
o
f
SL
R
r
esear
ch
h
as
u
tili
ze
d
co
m
p
u
ter
v
is
io
n
wh
ile
t
h
e
m
aj
o
r
ity
o
f
SLR
r
esear
ch
em
p
l
o
y
s
v
is
io
n
-
b
ased
m
eth
o
d
s
u
s
in
g
R
GB
im
ag
es
o
r
v
id
eo
s
,
r
ec
en
t
ad
v
an
ce
m
en
ts
h
av
e
also
in
tr
o
d
u
ce
d
s
en
s
o
r
-
b
ased
ap
p
r
o
ac
h
es.
T
h
ese
lev
er
ag
e
to
o
ls
s
u
ch
as
th
e
leap
m
o
tio
n
co
n
tr
o
ller
(
L
MC)
o
r
wea
r
ab
le
g
lo
v
es
to
ca
p
tu
r
e
f
in
e
-
g
r
ain
e
d
,
th
r
ee
-
d
im
en
s
io
n
al
m
o
tio
n
d
ata.
Su
c
h
s
y
s
tem
s
o
f
f
er
b
en
ef
its
in
clu
d
in
g
h
ig
h
tem
p
o
r
al
r
eso
lu
tio
n
,
d
e
p
th
s
en
s
in
g
,
a
n
d
r
e
al
-
tim
e
f
ee
d
b
ac
k
,
m
ak
in
g
th
em
well
-
s
u
ited
f
o
r
d
y
n
am
ic
g
estu
r
e
r
ec
o
g
n
itio
n
an
d
em
b
ed
d
e
d
d
ep
lo
y
m
en
t
s
.
So
m
e
s
tu
d
ies,
f
o
r
in
s
tan
c
e,
ap
p
lied
ex
tr
em
e
lear
n
in
g
m
ac
h
in
es
(
E
L
M)
an
d
m
eta
-
lear
n
in
g
tech
n
i
q
u
es
t
o
r
ec
o
g
n
ize
two
-
h
a
n
d
ed
T
u
r
k
is
h
s
ig
n
la
n
g
u
a
g
e
(
T
SL)
g
estu
r
es
u
s
in
g
leap
m
o
t
io
n
[
20
]
,
[
21
]
.
Oth
er
s
h
a
v
e
f
o
cu
s
ed
o
n
o
p
tim
izin
g
SLR
s
y
s
tem
s
f
o
r
lo
w
-
p
o
wer
ed
g
e
d
ev
ices,
d
e
m
o
n
s
tr
atin
g
th
e
p
o
te
n
tial
f
o
r
p
o
r
tab
le
a
n
d
ef
f
icien
t
s
en
s
o
r
-
d
r
i
v
en
SLR
[
22
]
.
Desp
ite
th
eir
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
5
0
0
8
-
5
0
1
6
5010
s
tr
en
g
th
s
,
s
en
s
o
r
-
b
ased
s
y
s
tem
s
o
f
ten
r
ely
o
n
s
p
ec
ialized
h
ar
d
war
e,
lim
itin
g
ac
ce
s
s
ib
ilit
y
in
e
d
u
ca
tio
n
al
o
r
r
eso
u
r
ce
-
co
n
s
tr
ain
ed
s
ettin
g
s
.
B
y
co
n
tr
ast,
th
e
p
r
esen
t
s
t
u
d
y
p
r
o
p
o
s
es
a
v
is
io
n
-
b
ased
C
NN
m
o
d
el
th
at
o
p
er
ates
s
o
lely
o
n
R
GB
im
ag
es,
elim
in
atin
g
th
e
n
ee
d
f
o
r
e
x
ter
n
al
s
en
s
o
r
s
.
T
h
is
a
p
p
r
o
ac
h
o
f
f
er
s
h
ig
h
class
if
icatio
n
ac
cu
r
ac
y
wh
ile
r
em
ain
in
g
co
s
t
-
ef
f
ec
tiv
e
an
d
s
ca
lab
le
m
ak
in
g
it
p
a
r
ticu
lar
ly
s
u
itab
le
f
o
r
d
ep
lo
y
m
e
n
t in
s
ch
o
o
ls
,
p
u
b
lic
in
s
titu
tio
n
s
,
an
d
in
clu
s
iv
e
co
m
m
u
n
icatio
n
e
n
v
ir
o
n
m
en
ts
ac
r
o
s
s
I
n
d
o
n
esia.
T
h
e
p
u
r
p
o
s
e
o
f
th
is
r
esear
ch
is
to
d
e
v
elo
p
th
e
I
n
d
o
n
esia
s
ig
n
lan
g
u
ag
e
SIBI
im
ag
e
r
ec
o
g
n
itio
n
s
y
s
tem
b
y
u
s
in
g
C
NN
ar
ch
itectu
r
e.
Sig
n
lan
g
u
a
g
e
is
a
v
ital
co
m
m
u
n
icatio
n
m
eth
o
d
f
o
r
th
e
d
ea
f
a
n
d
h
ar
d
-
of
-
h
ea
r
in
g
co
m
m
u
n
ity
[
1
8
]
,
[
23
]
,
[
24
]
.
Au
to
m
atic
SLR
s
y
s
tem
s
ca
n
f
ac
ilit
ate
s
ea
m
le
s
s
co
m
m
u
n
icatio
n
,
en
h
an
cin
g
ac
ce
s
s
ib
ilit
y
an
d
in
clu
s
iv
ity
[
1
9
]
,
[2
5
]
,
[
2
6
]
.
T
h
e
s
tu
d
y
in
tr
o
d
u
ce
s
a
C
NN
ar
ch
i
tectu
r
e
s
p
ec
if
ically
o
p
tim
ized
f
o
r
r
ec
o
g
n
izin
g
SI
B
I
.
Un
lik
e
g
e
n
er
ic
C
NN
m
o
d
els,
th
e
p
r
o
p
o
s
ed
a
r
ch
itectu
r
e
is
f
in
e
-
t
u
n
ed
to
h
an
d
le
th
e
u
n
i
q
u
e
ch
ar
a
cter
is
tics
o
f
SIBI
s
ig
n
s
,
en
s
u
r
in
g
h
ig
h
er
r
ec
o
g
n
itio
n
ac
c
u
r
ac
y
a
n
d
r
o
b
u
s
tn
ess
.
T
h
e
p
r
im
ar
y
co
n
tr
i
b
u
tio
n
s
o
f
t
h
is
wo
r
k
in
clu
d
e
th
e
d
esig
n
o
f
an
ef
f
icien
t
C
NN
ar
ch
itectu
r
e
tailo
r
ed
f
o
r
SLR
s
y
s
tem
s
an
d
th
e
cr
ea
tio
n
o
f
a
r
o
b
u
s
t d
ataset
f
o
r
tr
ai
n
in
g
a
n
d
ev
alu
atin
g
th
e
m
o
d
el.
2.
M
E
T
H
O
D
T
h
is
s
ec
tio
n
will
b
e
d
escr
ib
e
d
th
e
m
eth
o
d
t
o
co
llect,
p
r
ep
r
o
ce
s
s
,
an
d
p
r
o
ce
s
s
d
ata
th
at
we
u
s
ed
.
T
h
is
s
y
s
te
m
p
u
r
p
o
s
e
d
is
u
s
i
n
g
C
N
N
t
h
a
t
c
o
n
s
is
ts
o
f
m
u
ltip
le
lay
er
s
,
in
clu
d
in
g
co
n
v
o
lu
tio
n
al
la
y
er
s
,
p
o
o
lin
g
lay
e
r
s
,
an
d
f
u
lly
co
n
n
ec
ted
lay
er
s
.
T
h
e
d
i
ag
r
a
m
o
f
t
h
is
r
es
ea
r
c
h
m
eth
o
d
o
l
o
g
y
is
s
h
o
wn
in
Fi
g
u
r
e
1
.
All
ex
p
e
r
im
en
ts
wer
e
co
n
d
u
cted
lo
ca
lly
o
n
a
Ma
cBo
o
k
Air
(
2
0
2
0
)
e
q
u
ip
p
ed
with
a
1
.
1
GHz
Qu
ad
-
C
o
r
e
I
n
tel
C
o
r
e
i5
p
r
o
ce
s
s
o
r
,
8
GB
L
PDD
R
4
X
R
AM
,
an
d
I
n
tel
I
r
is
P
lu
s
in
t
eg
r
ated
g
r
ap
h
ics.
T
h
e
d
ev
elo
p
m
en
t
en
v
ir
o
n
m
en
t
in
clu
d
ed
Py
th
o
n
3
.
9
a
n
d
T
en
s
o
r
Flo
w
2
.
1
0
,
alo
n
g
with
s
u
p
p
o
r
tin
g
lib
r
ar
ies
s
u
ch
as
Ker
as,
Nu
m
Py
,
Op
en
C
V,
an
d
Ma
tp
lo
tlib
.
Mo
d
el
tr
ain
in
g
an
d
ev
al
u
atio
n
wer
e
p
e
r
f
o
r
m
ed
in
a
J
u
p
y
ter
N
o
teb
o
o
k
e
n
v
ir
o
n
m
en
t
with
o
u
t
GPU
ac
ce
ler
atio
n
.
As
s
u
ch
,
co
m
p
u
tatio
n
al
tim
e
v
a
r
ied
b
ase
d
o
n
b
ac
k
g
r
o
u
n
d
p
r
o
ce
s
s
es
an
d
s
y
s
tem
lo
ad
,
an
d
p
r
ec
is
e
b
en
ch
m
a
r
k
in
g
was n
o
t
th
e
f
o
cu
s
o
f
th
is
s
tu
d
y
.
T
h
e
c
o
n
v
o
l
u
ti
o
n
la
y
e
r
,
a
c
r
u
ci
al
c
o
m
p
o
n
e
n
t
o
f
t
h
e
C
NN
,
a
p
p
lies
a
c
o
n
v
o
l
u
t
io
n
o
p
e
r
ati
o
n
t
o
t
h
e
o
u
tp
u
t
o
f
t
h
e
p
r
ec
e
d
i
n
g
la
y
e
r
.
T
h
is
p
r
o
ce
s
s
f
o
r
m
s
t
h
e
c
o
r
e
m
e
ch
a
n
is
m
o
f
t
h
e
C
NN
,
en
a
b
li
n
g
i
t
t
o
l
ea
r
n
an
d
e
x
t
r
ac
t
ess
e
n
ti
al
f
ea
tu
r
es
f
r
o
m
i
n
p
u
t
d
ata
.
C
o
n
v
o
l
u
ti
o
n
,
i
n
t
h
is
c
o
n
te
x
t
,
i
n
v
o
l
v
es
r
ep
ea
te
d
l
y
a
p
p
l
y
i
n
g
a
s
et
o
f
lea
r
n
a
b
l
e
f
ilt
e
r
s
t
o
c
ap
tu
r
e
s
p
a
tia
l
p
a
tte
r
n
s
,
s
u
c
h
as
e
d
g
es,
s
h
a
p
es
,
a
n
d
te
x
t
u
r
es
,
wh
ic
h
a
r
e
c
r
iti
ca
l
f
o
r
class
if
ica
ti
o
n
an
d
r
e
co
g
n
it
io
n
tas
k
s
.
T
h
is
ite
r
a
ti
v
e
p
r
o
ce
s
s
is
ill
u
s
t
r
a
te
d
in
Fi
g
u
r
e
2
.
Fig
u
r
e
1
.
Diag
r
a
m
f
o
r
r
esear
ch
m
eth
o
d
o
lo
g
y
Fig
u
r
e
2
.
C
o
n
v
o
lu
tio
n
al
n
eu
r
a
l n
etwo
r
k
s
tr
u
ctu
r
e
2
.
1
.
Da
t
a
c
o
llect
io
n
Data
co
llectio
n
was
ca
r
r
ied
o
u
t
to
o
b
tain
t
h
e
in
f
o
r
m
atio
n
n
ee
d
ed
to
ac
h
iev
e
th
e
o
b
jecti
v
es
o
f
th
e
s
tu
d
y
.
T
h
is
p
h
ase
b
eg
in
s
b
y
d
o
wn
lo
ad
in
g
t
h
e
SIBI
d
ataset
f
r
o
m
Kag
g
le
an
d
s
to
r
in
g
it
in
th
e
lo
ca
l
d
ir
ec
to
r
y
.
T
h
is
r
esear
ch
u
tili
ze
s
th
e
SIB
I
d
ataset
o
f
I
n
d
o
n
esian
s
ig
n
lan
g
u
ag
e
.
SIBI
is
u
s
ed
b
ec
au
s
e
n
ea
r
ly
all
f
o
r
m
al
ed
u
ca
tio
n
al
i
n
s
titu
tio
n
s
th
at
i
m
p
lem
en
t
s
ig
n
lan
g
u
ag
e
u
tili
ze
th
is
f
o
r
m
o
f
s
ig
n
lan
g
u
a
g
e
[
2
7
]
.
T
h
e
SIBI
d
ataset
co
n
tain
s
5
,
2
8
0
im
a
g
es
o
f
s
tatic
p
o
s
e
I
n
d
o
n
esian
s
ig
n
lan
g
u
a
g
e
ac
r
o
s
s
twen
ty
-
s
ix
(
2
6
)
ca
teg
o
r
ies
o
f
alp
h
ab
ets.
E
x
am
p
le
im
ag
es
f
r
o
m
th
e
d
ataset
ar
e
s
h
o
wn
i
n
Fig
u
r
e
3
.
T
h
e
d
ataset,
wh
ich
is
av
ailab
le
o
n
Kag
g
le,
h
as
a
s
ize
o
f
2
.
7
GB
.
T
h
er
e
ar
e
ap
p
r
o
x
im
ately
1
0
2
to
1
0
4
im
ag
es
p
er
alp
h
ab
et
c
h
ar
ac
ter
in
s
tan
d
a
r
d
R
GB
f
o
r
m
at.
A
wh
ite
p
ap
er
i
s
p
lace
d
b
eh
in
d
th
e
h
an
d
s
ig
n
as
a
b
ac
k
g
r
o
u
n
d
.
Am
o
n
g
th
e
5
,
2
8
0
im
ag
es,
we
u
s
ed
3
,
6
9
6
im
ag
es
(
7
0
%)
f
o
r
th
e
tr
ain
i
n
g
s
et,
a
n
d
1
,
5
8
4
im
ag
es
(
3
0
%)
f
o
r
th
e
v
alid
ati
o
n
s
et.
T
h
e
tr
ain
i
n
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
R
ec
o
g
n
itio
n
o
f I
n
d
o
n
esia
n
s
ig
n
la
n
g
u
a
g
e
u
s
in
g
d
ee
p
lea
r
n
in
g
…
(
Olivia
K
emb
u
a
n
)
5011
d
ataset
is
u
s
ed
to
tr
ain
th
e
m
o
d
el
wh
ile
th
e
v
alid
atio
n
d
ataset
is
u
s
ed
to
m
o
n
ito
r
t
h
e
wo
r
k
in
g
o
f
th
e
m
o
d
e
l
wh
ich
is
n
o
t
u
s
ed
d
u
r
in
g
th
e
tr
ain
in
g
d
ata.
T
h
e
v
alid
atio
n
d
ataset
also
h
elp
s
to
ch
ec
k
w
h
eth
er
th
e
m
o
d
el
is
o
v
er
f
itti
n
g
o
r
n
o
t.
Fig
u
r
e
3
.
E
x
am
p
le
o
f
d
atas
et
SIBI
s
ig
n
lan
g
u
ag
e
2
.
2
.
Da
t
a
prepro
ce
s
s
ing
I
m
ag
e
p
r
o
ce
s
s
in
g
is
a
m
eth
o
d
th
at
co
n
v
er
ts
an
im
ag
e
in
t
o
a
co
m
p
u
ter
alg
o
r
ith
m
co
m
p
u
t
atio
n
,
th
is
im
ag
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
e
is
u
s
ed
f
o
r
co
m
p
u
ter
s
to
f
in
d
o
u
t
th
e
in
f
o
r
m
atio
n
in
an
im
ag
e
th
at
h
as
b
ee
n
d
o
n
e
f
ea
tu
r
e
ex
tr
ac
tio
n
p
r
o
ce
s
s
.
T
h
e
s
tep
s
o
f
th
e
im
ag
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
e
ar
e
as f
o
llo
ws
[
2
8
]
:
‒
I
n
p
u
t im
a
g
es th
at
h
av
e
b
ee
n
ta
k
en
th
r
o
u
g
h
th
e
s
ca
n
n
e
r
p
r
o
ce
s
s
o
r
th
r
o
u
g
h
th
e
p
h
o
to
p
r
o
ce
s
s
d
ir
ec
tly
.
‒
Af
ter
in
s
er
tin
g
th
e
im
ag
e,
th
e
im
ag
e
an
aly
s
is
an
d
m
an
ip
u
latio
n
p
r
o
ce
s
s
will
b
e
ca
r
r
i
ed
o
u
t,
wh
ic
h
co
n
s
is
ts
o
f
im
p
r
o
v
in
g
im
ag
e
q
u
ality
,
co
m
p
r
ess
in
g
im
ag
e
d
at
a,
an
d
d
esig
n
in
g
im
a
g
e
p
atter
n
s
.
‒
Af
ter
th
e
im
ag
e
p
r
e
-
p
r
o
ce
s
s
in
g
p
r
o
ce
s
s
,
th
e
im
ag
e
d
ata
will b
e
co
n
v
e
r
ted
b
ac
k
in
t
o
an
im
a
g
e
th
at
will b
e
u
s
ed
in
th
e
class
if
icatio
n
p
r
o
c
ess
.
Data
p
r
ep
r
o
ce
s
s
in
g
is
a
cr
u
cial
s
tep
in
p
r
ep
ar
in
g
y
o
u
r
d
ataset
f
o
r
tr
ain
in
g
a
C
NN.
Pro
p
er
p
r
ep
r
o
ce
s
s
in
g
h
elp
s
im
p
r
o
v
e
th
e
q
u
ality
an
d
p
er
f
o
r
m
an
ce
o
f
y
o
u
r
m
o
d
el.
T
h
is
p
h
ase
b
eg
in
b
y
d
o
wn
lo
ad
in
g
th
e
SIBI
d
ataset
f
r
o
m
Kag
g
le
an
d
s
to
r
e
it
in
t
h
e
lo
ca
l
d
i
r
ec
to
r
y
.
Af
te
r
ex
tr
ac
tin
g
th
e
co
n
ten
ts
,
we
ass
ig
n
v
ar
iab
les
with
th
e
f
ile
p
ath
f
o
r
tr
ain
in
g
d
ataset
an
d
v
alid
atio
n
d
ata
s
et.
W
e
as
s
ig
n
v
ar
iab
les
with
th
e
f
ile
p
ath
f
o
r
th
e
tr
ain
in
g
d
ataset
an
d
v
a
lid
atio
n
d
ata
s
et
af
ter
ex
tr
ac
ti
n
g
th
e
co
n
ten
ts
.
Af
te
r
estab
lis
h
in
g
a
7
0
%
tr
ain
i
n
g
d
ataset
an
d
a
3
0
%
v
alid
atio
n
d
ataset
r
atio
,
we
s
av
e
th
e
p
h
o
to
g
r
ap
h
s
in
v
ar
io
u
s
f
o
ld
er
s
.
T
h
e
C
NN
m
o
d
el
is
tr
ain
ed
u
s
in
g
th
e
tr
ain
d
atas
et.
Af
ter
m
ak
in
g
a
s
et
o
f
p
r
ed
ictio
n
s
,
th
e
m
o
d
el
was
ev
alu
ated
u
s
in
g
t
h
e
v
alid
atio
n
d
ataset.
T
h
e
n
ex
t
s
t
ag
e
is
to
u
s
e
th
e
I
m
ag
eDa
taG
en
er
ato
r
class
th
at
tf
.
k
er
as
p
r
o
v
id
es
to
d
ec
o
d
e
th
e
co
n
ten
ts
o
f
th
ese
s
ig
n
la
n
g
u
a
g
e
im
ag
es
an
d
tr
a
n
s
f
o
r
m
th
e
m
in
to
f
lo
atin
g
p
o
in
t
ten
s
o
r
s
.
Data
th
at
h
as
b
ee
n
d
iv
id
ed
in
t
o
tr
ain
in
g
d
ata
an
d
v
alid
atio
n
d
ata
is
th
en
p
r
ep
r
o
c
ess
ed
s
u
ch
as r
escale,
r
o
tatio
n
,
an
d
f
lip
.
T
h
e
p
u
r
p
o
s
e
o
f
th
is
f
lip
p
in
g
p
r
o
ce
s
s
is
to
m
ak
e
p
ad
d
in
g
ea
s
i
er
wh
en
it
is
r
u
n
n
in
g
in
ea
ch
p
r
o
ce
s
s
.
At
th
e
n
ex
t
p
r
ep
r
o
ce
s
s
in
g
s
tag
e,
th
e
r
o
tatio
n
p
r
o
ce
s
s
is
ca
r
r
ied
o
u
t
wh
er
e
th
e
f
ac
e
im
ag
e
will
b
e
r
o
tated
f
r
o
m
th
e
lef
t
o
r
f
r
o
m
th
e
r
ig
h
t
.
T
h
e
s
ca
l
in
g
p
r
o
ce
s
s
s
tag
e
will
also
b
e
ap
p
lied
in
t
h
e
tr
ain
in
g
s
et,
with
th
e
aim
t
h
at
later
n
eu
r
al
n
etwo
r
k
s
ca
n
lear
n
th
e
f
ea
tu
r
es o
f
th
e
o
r
ig
in
al
s
ca
le.
2
.
3
.
Cre
a
t
e
a
nd
t
ra
in t
he
m
o
del
T
h
e
s
tr
u
ctu
r
e
o
f
C
NN
u
s
ed
i
n
th
is
r
esear
ch
s
h
o
wn
in
Fig
u
r
e
4
.
Po
o
lin
g
lay
er
s
ar
e
r
esp
o
n
s
ib
le
f
o
r
r
ed
u
cin
g
th
e
d
im
e
n
s
io
n
ality
o
f
f
ea
tu
r
e
m
ap
s
,
s
p
ec
if
ically
th
e
h
eig
h
t
an
d
wid
th
,
p
r
eser
v
in
g
th
e
d
ep
th
[
2
9
]
.
Ma
x
p
o
o
lin
g
o
u
tp
u
ts
th
e
m
ax
im
u
m
v
alu
e
o
f
th
e
elem
en
ts
in
th
e
p
o
r
tio
n
o
f
th
e
im
ag
e
co
v
er
ed
b
y
th
e
f
ilter
.
Ma
x
p
o
o
lin
g
is
b
etter
at
ex
tr
ac
tin
g
d
o
m
in
an
t f
ea
t
u
r
es a
n
d
th
e
r
ef
o
r
e
,
co
n
s
id
er
ed
m
o
r
e
p
er
f
o
r
m
an
t
[
3
0
]
.
Du
r
in
g
th
e
tr
ain
in
g
p
r
o
ce
s
s
,
th
e
C
NN
is
tr
ain
ed
o
n
a
d
ataset
an
d
th
en
ev
alu
ated
o
n
a
s
ep
ar
ate
v
a
lid
atio
n
d
ataset
to
m
o
n
ito
r
its
p
er
f
o
r
m
an
ce
.
At
th
is
s
tag
e,
a
co
n
v
o
lu
tio
n
o
p
er
ati
o
n
is
p
er
f
o
r
m
e
d
b
etwe
en
th
e
i
n
p
u
t m
atr
ix
an
d
th
e
f
ilter
m
atr
ix
.
T
h
ese
f
ilter
s
wi
ll
b
e
s
h
if
te
d
r
e
p
ea
ted
ly
o
v
e
r
th
e
en
tire
im
ag
e
ar
ea
,
p
r
o
d
u
c
in
g
a
f
ea
tu
r
e
m
a
p
m
atr
ix
as o
u
tp
u
t.
T
h
is
f
ea
tu
r
e
m
ap
m
atr
ix
ca
n
b
e
ca
lc
u
lated
u
s
in
g
th
e
(
1
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
5
0
0
8
-
5
0
1
6
5012
=
(
−
+
2
+
1
)
(
1
)
W
h
e
r
e
is
f
e
at
u
r
e
m
a
p
s
iz
e
;
is
m
at
r
i
x
i
n
p
u
t
s
i
ze
;
p
is
p
a
d
d
i
n
g
’
s
s
iz
e
;
a
n
d
s
is
s
tr
id
e
.
In
(
2
)
is
th
e
c
o
n
v
u
ls
i
o
n
o
p
e
r
ati
o
n
f
o
r
m
u
la
o
f
C
NN:
[
]
,
=
(
∑
∑
[
−
,
−
]
[
,
]
+
)
(
2
)
W
h
e
r
e
[
]
is
f
ea
tu
r
e
m
a
p
m
a
tr
ix
I
;
N
is
in
p
u
t
i
m
a
g
e
m
at
r
i
x
;
F
is
co
n
v
o
lu
ti
o
n
f
ilt
e
r
m
at
r
i
x
;
bF
is
b
ias
v
al
u
e
i
n
f
ilt
e
r
;
j,
k
is
p
i
x
el
p
o
s
iti
o
n
in
t
h
e
i
n
p
u
t
im
ag
e
m
a
tr
ix
;
an
d
m,
n
is
p
i
x
el
p
o
s
it
io
n
i
n
th
e
c
o
n
v
o
l
u
ti
o
n
f
i
lte
r
m
a
tr
ix
.
Af
ter
th
e
co
n
v
o
lu
tio
n
p
r
o
ce
s
s
is
co
m
p
lete,
th
e
n
ex
t
s
tep
is
t
o
ap
p
ly
a
n
ac
tiv
atio
n
f
u
n
ctio
n
u
s
in
g
th
e
r
ec
tifie
d
lin
ea
r
u
n
it
(
R
eL
u
)
.
T
h
e
R
eL
U
lay
er
,
o
r
r
ec
tifie
d
lin
ea
r
u
n
it
lay
er
,
ca
n
b
e
th
o
u
g
h
t o
f
as
a
th
r
esh
o
ld
i
n
g
p
r
o
ce
s
s
o
r
ac
tiv
atio
n
f
u
n
ctio
n
in
ar
tific
ial
n
eu
r
al
n
etwo
r
k
s
[
3
1
]
,
[
3
2
]
.
E
ac
h
p
ix
el
in
t
h
e
f
ea
tu
r
e
m
ap
will
b
e
in
p
u
t
to
th
e
R
eL
U
f
u
n
ctio
n
,
w
h
er
e
p
ix
els
with
v
alu
es
less
th
an
0
will
b
e
co
n
v
e
r
ted
to
0
.
T
h
e
f
o
r
m
u
la
u
s
ed
f
o
r
R
eL
U
is
f
(
x
)
=
m
ax
(
0
,
x
)
.
Fig
u
r
e
4
.
Stru
ctu
r
e
o
f
co
n
v
C
NN
2
.
4
.
E
v
a
lua
t
i
o
n
T
h
is
p
h
ase
is
to
ev
alu
ate
th
e
ac
cu
r
ac
y
o
f
a
C
NN
o
n
b
o
th
t
h
e
tr
ain
in
g
an
d
v
alid
atio
n
d
at
asets
u
s
in
g
T
en
s
o
r
Flo
w/Ker
as
an
d
Py
T
o
r
ch
.
T
h
is
p
r
o
ce
s
s
in
v
o
lv
es
tr
ain
in
g
th
e
m
o
d
el,
ev
alu
atin
g
it
o
n
b
o
th
d
atasets
,
an
d
o
p
tio
n
ally
p
lo
ttin
g
t
h
e
ac
cu
r
a
cy
v
alu
es
to
v
is
u
alize
th
e
m
o
d
el’
s
p
er
f
o
r
m
an
ce
o
v
e
r
tim
e.
W
h
en
ev
alu
atin
g
a
C
NN,
ac
cu
r
ac
y
is
a
k
ey
m
etr
i
c
th
at
in
d
icate
s
th
e
p
r
o
p
o
r
tio
n
o
f
co
r
r
ec
t
p
r
e
d
ictio
n
s
m
ad
e
b
y
th
e
m
o
d
el
o
u
t
o
f
all
p
r
ed
ictio
n
s
.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
is
s
ec
tio
n
d
escr
ib
es
th
e
r
e
s
u
lts
o
b
tain
ed
f
r
o
m
th
e
im
p
l
em
en
tatio
n
an
d
d
is
cu
s
s
io
n
o
f
th
e
im
ag
e
r
ec
o
g
n
itio
n
o
f
s
ig
n
lan
g
u
ag
e
u
s
in
g
th
e
C
NN
.
T
h
e
r
esu
lts
s
h
o
wn
ar
e
th
e
r
esu
lts
o
f
p
r
ep
r
o
ce
s
s
in
g
,
d
ata
m
o
d
elin
g
a
n
d
tr
ain
i
n
g
,
a
n
d
ev
alu
atio
n
.
T
h
e
r
esu
lts
ar
e
d
iv
id
ed
in
to
ac
cu
r
ac
y
test
in
g
an
d
d
ata
lo
s
s
test
in
g
.
3
.
1
.
T
ra
ini
ng
a
nd
v
a
lid
a
t
io
n
a
cc
ura
cy
T
h
e
m
o
d
el
is
co
m
p
o
s
ed
o
f
f
o
u
r
co
n
v
o
lu
tio
n
b
lo
c
k
s
,
as
s
u
m
m
ar
ized
in
Fig
u
r
e
4
,
ea
ch
o
f
wh
ich
h
as
a
m
ax
p
o
o
l
lay
er
an
d
is
tr
ig
g
er
ed
b
y
a
R
eL
u
ac
tiv
atio
n
f
u
n
ctio
n
.
T
h
e
m
o
d
el
b
e
g
in
s
with
th
e
f
ir
s
t
co
n
v
o
lu
tio
n
al
lay
er
(
co
n
v
2
d
)
,
wh
ich
em
p
l
o
y
s
3
2
f
ilter
s
to
ex
tr
ac
t
in
itial
f
ea
tu
r
es
f
r
o
m
th
e
in
p
u
t
im
ag
es.
T
h
is
lay
er
p
r
o
d
u
ce
s
f
ea
tu
r
e
m
ap
s
with
d
im
en
s
io
n
s
o
f
1
4
8
×
1
4
8
an
d
3
2
c
h
an
n
els.
Nex
t,
th
e
m
o
d
el
in
clu
d
es
a
s
ec
o
n
d
co
n
v
o
lu
tio
n
al
lay
er
(
co
n
v
2
d
_
1
)
with
6
4
f
ilt
er
s
,
wh
ich
ex
tr
ac
ts
m
o
r
e
co
m
p
lex
f
ea
tu
r
es.
T
h
e
s
p
atial
d
im
en
s
io
n
s
ar
e
s
lig
h
tly
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
R
ec
o
g
n
itio
n
o
f I
n
d
o
n
esia
n
s
ig
n
la
n
g
u
a
g
e
u
s
in
g
d
ee
p
lea
r
n
in
g
…
(
Olivia
K
emb
u
a
n
)
5013
r
ed
u
ce
d
to
7
2
×
7
2
.
A
s
ec
o
n
d
m
ax
p
o
o
lin
g
o
p
er
atio
n
is
ap
p
lied
th
r
o
u
g
h
th
e
m
ax
_
p
o
o
lin
g
2
d
_
1
lay
er
,
f
u
r
th
er
r
ed
u
cin
g
th
e
d
im
e
n
s
io
n
s
to
3
6
×
3
6
wh
ile
m
ain
tain
in
g
6
4
ch
a
n
n
els.
T
h
e
ar
ch
itectu
r
e
th
e
n
in
c
o
r
p
o
r
ates
a
th
ir
d
c
o
n
v
o
lu
tio
n
al
lay
e
r
(
co
n
v
2
d
_
2
)
with
1
2
8
f
ilter
s
,
p
r
o
d
u
cin
g
f
ea
tu
r
e
m
ap
s
with
d
im
en
s
io
n
s
o
f
3
4
×
3
4
an
d
a
g
r
ea
ter
d
e
p
th
.
T
h
is
s
tep
is
f
o
llo
wed
b
y
a
th
ir
d
m
ax
p
o
o
lin
g
o
p
er
atio
n
,
w
h
ich
r
ed
u
ce
s
th
e
s
p
atial
d
im
en
s
io
n
s
to
1
7
×
1
7
.
T
h
e
f
in
al
co
n
v
o
lu
tio
n
al
lay
er
(
co
n
v
2
d
_
3
)
u
tili
ze
s
2
5
6
f
ilter
s
to
ex
tr
ac
t
h
i
g
h
ly
d
etailed
f
ea
tu
r
es.
T
h
e
s
p
ati
al
d
im
en
s
io
n
s
ar
e
r
ed
u
ce
d
t
o
1
5
×
1
5
,
f
o
llo
we
d
b
y
a
f
in
al
p
o
o
lin
g
o
p
e
r
atio
n
(
m
ax
_
p
o
o
lin
g
2
d
_
3
)
,
wh
ich
o
u
t
p
u
ts
f
ea
tu
r
e
m
a
p
s
with
d
im
en
s
io
n
s
o
f
7
×
7
an
d
a
d
ep
th
o
f
2
5
6
.
T
o
p
r
ev
en
t
o
v
er
f
itti
n
g
,
a
d
r
o
p
o
u
t
lay
er
is
ap
p
lied
,
wh
ich
r
an
d
o
m
ly
d
ea
ctiv
ates
s
o
m
e
n
eu
r
o
n
s
d
u
r
in
g
tr
ain
in
g
.
Dr
o
p
o
u
t
is
a
C
NN
r
eg
u
lar
izatio
n
tech
n
iq
u
e
th
at
r
eso
lv
es
n
eu
r
o
n
al
in
ter
d
ep
en
d
en
cy
.
Ov
er
f
itti
n
g
o
f
th
e
d
ata
is
a
r
esu
lt
o
f
th
is
in
t
er
d
ep
en
d
en
cy
.
Po
o
r
p
r
e
d
ictio
n
s
in
a
d
ataset
ca
n
b
e
ca
u
s
ed
b
y
o
v
er
f
itti
n
g
[
3
3
]
.
Af
ter
war
d
,
th
e
th
r
ee
-
d
im
en
s
io
n
al
f
ea
tu
r
e
m
ap
s
a
r
e
f
latten
e
d
th
r
o
u
g
h
th
e
f
latten
lay
er
,
co
n
v
er
tin
g
t
h
em
in
t
o
a
one
-
d
im
e
n
s
io
n
al
v
ec
to
r
o
f
1
2
,
5
4
4
f
ea
t
u
r
es.
T
h
e
m
o
d
el
th
e
n
co
n
n
ec
ts
to
th
e
f
ir
s
t d
en
s
e
lay
er
,
co
m
p
r
is
in
g
1
,
0
2
4
n
eu
r
o
n
s
,
wh
ich
s
er
v
es
as a
b
r
id
g
e
b
etwe
en
th
e
ex
tr
ac
te
d
f
ea
tu
r
e
s
an
d
th
e
f
i
n
al
o
u
t
p
u
t.
L
astl
y
,
th
e
s
ec
o
n
d
d
en
s
e
lay
er
ac
ts
as
th
e
o
u
tp
u
t
lay
er
with
2
6
n
eu
r
o
n
s
,
r
ep
r
esen
tin
g
th
e
2
6
alp
h
ab
et
class
es
f
o
r
c
lass
if
icatio
n
.
I
n
to
tal,
th
e
m
o
d
el
h
as
1
3
,
2
5
9
,
2
3
4
tr
ain
ab
le
p
ar
am
eter
s
,
with
th
e
m
ajo
r
ity
co
n
ce
n
t
r
ated
in
th
e
d
en
s
e
lay
er
s
.
T
h
is
ar
ch
itectu
r
e
is
d
esig
n
ed
to
ef
f
icien
tly
class
if
y
alp
h
ab
ets
with
h
ig
h
ac
cu
r
a
cy
b
y
co
m
b
in
in
g
s
p
atial
f
ea
tu
r
e
e
x
tr
ac
tio
n
with
d
ee
p
lear
n
in
g
.
T
h
e
f
u
n
ctio
n
o
f
p
o
o
lin
g
lay
e
r
s
is
to
r
ed
u
ce
th
e
d
im
en
s
io
n
ality
o
f
f
ea
tu
r
e
m
ap
s
,
m
ea
n
in
g
th
at
th
e
d
ep
th
is
p
r
eser
v
ed
wh
ile
th
e
h
eig
h
t
an
d
b
r
ea
d
th
a
r
e
r
ed
u
ce
d
[
3
1
]
.
Ma
x
p
o
o
lin
g
p
r
o
d
u
ce
s
th
e
m
a
x
im
u
m
v
alu
e
with
in
ea
ch
r
eg
io
n
o
f
th
e
im
a
g
e
en
c
o
m
p
ass
ed
b
y
th
e
f
ilter
.
Ma
x
p
o
o
lin
g
is
th
o
u
g
h
t
to
b
e
m
o
r
e
ef
f
icien
t
s
in
ce
it
is
m
o
r
e
ef
f
ec
tiv
e
at
ex
tr
ac
tin
g
d
o
m
in
atin
g
f
ea
tu
r
es
[
3
4
]
.
T
h
e
f
i
n
a
l
f
e
a
tu
r
e
m
a
p
p
i
n
g
s
a
r
e
c
o
n
v
er
t
e
d
in
t
o
a
s
i
n
g
l
e
1
D
v
e
c
t
o
r
u
s
in
g
th
e
m
o
d
e
l
's
"
F
l
a
t
t
e
n
"
l
ay
e
r
.
A
f
t
e
r
c
er
t
a
i
n
co
n
v
o
lu
t
i
o
n
a
l
/m
a
x
p
o
o
l
l
a
y
er
s
,
t
h
e
f
l
a
t
t
en
i
n
g
s
t
e
p
i
s
r
eq
u
i
r
e
d
in
o
r
d
e
r
to
e
m
p
l
o
y
f
u
l
l
y
l
in
k
ed
l
a
y
e
r
s
[
2
0
]
.
W
e
u
s
e
d
th
e
s
o
f
t
m
a
x
a
c
t
iv
a
t
i
o
n
f
u
n
c
t
io
n
in
th
e
f
i
n
a
l
l
a
y
er
.
S
o
f
t
M
a
x
ac
t
iv
a
t
i
o
n
o
r
So
f
t
M
ax
c
l
a
s
s
i
f
i
e
r
i
s
a
n
o
t
h
er
f
o
r
m
o
f
lo
g
i
s
t
i
c
r
eg
r
e
s
s
i
o
n
a
lg
o
r
i
t
h
m
th
a
t
w
e
c
an
u
s
e
t
o
c
l
a
s
s
i
f
y
m
o
r
e
t
h
an
t
w
o
c
la
s
s
e
s
.
E
a
c
h
c
l
a
s
s
's
o
u
t
p
u
t
in
S
o
f
t
M
a
x
i
s
co
n
s
tr
a
i
n
ed
to
a
v
a
l
u
e
b
e
t
w
e
en
0
an
d
1
.
T
h
i
s
i
n
d
i
ca
t
e
s
t
h
e
l
i
k
e
l
ih
o
o
d
t
h
a
t
a
n
in
p
u
t
i
s
a
m
e
m
b
e
r
o
f
a
s
p
e
c
i
f
i
c
c
la
s
s
.
U
s
i
n
g
a
b
a
tc
h
s
i
z
e
o
f
1
0
a
n
d
f
i
v
e
e
p
o
c
h
s
o
f
d
a
t
a,
t
h
e
C
N
N
w
a
s
t
r
a
in
e
d
f
o
r
1
0
0
s
t
ep
s
p
e
r
e
p
o
c
h
.
W
h
en
th
e
w
h
o
l
e
d
a
t
as
e
t
r
u
n
s
th
r
o
u
g
h
th
e
t
r
a
in
i
n
g
d
a
t
a
s
e
t
,
i
t
i
s
c
a
l
l
e
d
a
n
ep
o
ch
.
T
h
e
m
o
d
el
is
ev
alu
ated
o
n
th
e
test
d
ataset
to
ch
ec
k
th
e
ac
cu
r
ac
y
.
T
h
e
tr
ai
n
in
g
an
d
v
alid
atio
n
ac
cu
r
ac
y
a
n
d
l
o
s
s
ar
e
p
lo
tted
f
o
r
v
is
u
aliza
tio
n
.
T
h
e
m
o
d
e
l
t
r
ain
ed
f
o
r
t
h
e
d
ataset
h
a
d
in
it
ial
tr
ain
in
g
s
et
an
d
v
alid
atio
n
s
et
ac
cu
r
ac
y
o
f
3
.
6
3
%
an
d
3
.
3
3
%
in
ep
o
c
h
-
1
.
T
h
e
v
alid
atio
n
ac
cu
r
ac
y
en
d
e
d
u
p
af
ter
5
ep
o
c
h
s
with
9
7
.
5
8
%
ac
c
u
r
ac
y
f
o
r
tr
a
in
in
g
s
et,
an
d
1
0
0
%
ac
cu
r
ac
y
f
o
r
v
alid
atio
n
s
et.
T
h
e
ac
c
u
r
ac
y
f
o
r
t
r
ain
in
g
s
et
an
d
v
alid
atio
n
s
et
ca
n
b
e
s
h
o
wn
in
Fig
u
r
e
5
.
3
.
2
.
T
ra
ini
ng
a
nd
v
a
lid
a
t
io
n
lo
s
s
T
h
e
m
o
d
el
tr
ain
e
d
f
o
r
th
e
d
ataset
h
ad
in
itial
tr
ain
in
g
s
et
an
d
v
alid
atio
n
s
et
lo
s
s
o
f
3
3
.
7
8
%
an
d
3
2
.
4
7
%
in
e
p
o
ch
-
1
.
T
h
e
v
ali
d
atio
n
ac
c
u
r
ac
y
en
d
e
d
u
p
af
t
er
5
ep
o
c
h
s
with
8
.
7
8
%
l
o
s
s
f
o
r
tr
ai
n
in
g
s
et,
an
d
0
.
6
2
% lo
s
s
f
o
r
v
alid
atio
n
s
et.
T
h
e
lo
s
s
f
o
r
tr
ai
n
i
n
g
s
et
a
n
d
v
a
lid
ati
o
n
s
et
ca
n
b
e
s
h
o
w
n
i
n
Fi
g
u
r
e
6
.
Fig
u
r
e
5
.
T
r
ain
in
g
a
n
d
v
alid
atio
n
ac
cu
r
ac
y
g
r
ap
h
ic
Fig
u
r
e
6
.
T
r
ain
in
g
a
n
d
v
alid
atio
n
lo
s
s
4.
CO
NCLU
SI
O
N
T
h
e
in
itial
tr
ain
in
g
s
et
an
d
v
a
lid
atio
n
s
et
ac
cu
r
ac
y
f
o
r
th
e
m
o
d
el
tr
ain
ed
o
n
th
e
d
ataset
was
3
.
6
3
%
an
d
3
.
3
3
%
in
e
p
o
ch
-
1
.
Af
ter
f
iv
e
ep
o
c
h
s
,
th
e
v
alid
atio
n
ac
c
u
r
ac
y
was
9
7
.
5
8
%
f
o
r
th
e
tr
ain
in
g
s
et
an
d
1
0
0
%
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
5
0
0
8
-
5
0
1
6
5014
f
o
r
th
e
v
alid
atio
n
s
et.
T
h
e
m
o
d
el
tr
ain
e
d
f
o
r
th
e
d
ataset
h
ad
in
itial
tr
ain
in
g
s
et
an
d
v
al
id
atio
n
s
et
lo
s
s
o
f
3
3
.
7
8
% a
n
d
3
2
.
4
7
% in
ep
o
ch
-
1
.
T
h
e
v
alid
atio
n
ac
cu
r
ac
y
en
d
ed
u
p
a
f
ter
5
ep
o
ch
s
with
8
.
7
8
% lo
s
s
f
o
r
tr
ain
in
g
s
et,
an
d
0
.
6
2
%
lo
s
s
f
o
r
v
alid
atio
n
s
et.
T
h
e
m
o
d
el
p
er
f
o
r
m
s
b
etter
in
test
in
g
wh
en
th
er
e
is
a
g
r
ea
ter
s
u
p
p
ly
o
f
tr
ain
in
g
d
ata.
Selectin
g
th
e
tr
a
in
in
g
d
ata'
s
b
atch
s
ize
an
d
ep
o
ch
co
u
n
t
is
a
cr
u
cial
s
tep
in
th
i
s
s
tu
d
y
.
T
h
is
wo
r
k
p
r
esen
ts
a
p
r
e
d
ictiv
e
m
o
d
el
t
h
at
is
tr
ain
ed
e
x
clu
s
iv
ely
to
r
ec
o
g
n
ize
SIBI.
T
h
e
m
o
d
el
ca
n
b
e
im
p
r
o
v
ed
in
th
e
f
u
tu
r
e
an
d
s
till
ca
n
b
e
t
r
ain
ed
to
r
ec
o
g
n
ize
m
o
r
e
ch
a
r
ac
ter
s
an
d
e
v
en
f
o
r
an
o
th
er
lan
g
u
ag
e
.
T
h
e
d
ataset
as
a
n
in
p
u
t
f
o
r
th
is
m
o
d
el
with
a
lo
t
o
f
v
a
r
iatio
n
s
an
d
ca
n
b
e
ef
f
e
ctiv
ely
u
s
ed
to
tr
ai
n
th
e
p
r
o
p
o
s
ed
m
o
d
el
in
o
r
d
e
r
to
in
cr
ea
s
e
its
ef
f
icien
cy
a
n
d
ac
cu
r
ac
y
as
well.
So
m
e
t
y
p
es
o
f
SIBI
s
ig
n
la
n
g
u
a
g
e
ch
ar
ac
ter
s
r
eq
u
ir
e
m
o
v
em
en
t,
t
h
er
ef
o
r
e
f
o
r
f
u
r
th
er
d
ev
el
o
p
m
en
t
a
s
y
s
tem
th
at
is
ab
le
to
r
ec
o
g
n
ize
n
o
t
o
n
ly
im
ag
es
b
u
t
als
o
v
id
eo
s
is
n
ee
d
ed
.
I
n
f
u
t
u
r
e
wo
r
k
,
co
m
p
u
tatio
n
al
tim
e
ca
n
b
e
b
en
ch
m
ar
k
ed
m
o
r
e
r
ig
o
r
o
u
s
ly
o
n
a
s
tan
d
ar
d
ized
GPU
s
etu
p
,
en
ab
lin
g
f
ai
r
er
c
o
m
p
ar
is
o
n
s
ac
r
o
s
s
d
if
f
er
en
t
m
o
d
els
an
d
d
atasets
.
Ho
wev
er
,
th
e
cu
r
r
en
t
r
esu
lts
co
n
f
ir
m
t
h
e
m
o
d
el’
s
p
o
ten
tial
f
o
r
ef
f
icien
t
d
ep
l
o
y
m
en
t
in
a
s
s
is
tiv
e
tech
n
o
lo
g
ies,
p
a
r
ticu
lar
ly
in
ed
u
ca
tio
n
al
an
d
co
m
m
u
n
icatio
n
to
o
ls
f
o
r
th
e
d
ea
f
a
n
d
h
a
r
d
-
of
-
h
ea
r
in
g
c
o
m
m
u
n
ity
.
ACK
NO
WL
E
DG
M
E
N
T
S
T
h
e
I
n
d
o
n
esia
E
n
d
o
wm
en
t
Fu
n
d
s
f
o
r
E
d
u
ca
tio
n
(
L
PDP)
o
f
t
h
e
R
ep
u
b
lic
o
f
I
n
d
o
n
esia,
t
h
e
C
en
ter
f
o
r
E
d
u
ca
tio
n
al
Fin
an
cial
Ser
v
ic
es
(
PUSLAP
DI
K)
,
an
d
th
e
Min
is
tr
y
o
f
E
d
u
ca
tio
n
,
C
u
lt
u
r
e,
R
esear
ch
,
an
d
T
ec
h
n
o
lo
g
y
(
Kem
en
d
ik
b
u
d
r
is
tek
)
ar
e
ac
k
n
o
wled
g
ed
b
y
t
h
e
au
th
o
r
s
f
o
r
p
r
o
v
id
i
n
g
I
n
d
o
n
esian
E
d
u
ca
tio
n
Sch
o
lar
s
h
ip
s
(
B
PI)
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
was f
u
n
d
ed
b
y
th
e
I
n
d
o
n
esia E
n
d
o
wm
en
t Fu
n
d
s
f
o
r
E
d
u
ca
tio
n
(
L
PDP)
o
f
th
e
R
ep
u
b
lic
o
f
I
n
d
o
n
esia u
n
d
er
th
e
I
n
d
o
n
e
s
ian
E
d
u
ca
tio
n
Sch
o
lar
s
h
ip
(
B
PI)
p
r
o
g
r
am
[
Gr
a
n
t N
o
.
2
0
2
3
2
9
1
1
3
2
9
5
]
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Oliv
ia
Kem
b
u
an
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Har
y
an
to
✓
✓
✓
✓
✓
Mo
ch
am
ad
B
r
u
r
i
T
r
iy
o
n
o
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
I
NF
O
RM
E
D
CO
NS
E
N
T
W
e
h
av
e
o
b
tain
ed
in
f
o
r
m
ed
c
o
n
s
en
t f
r
o
m
all
in
d
iv
id
u
als in
c
lu
d
ed
in
t
h
is
s
tu
d
y
.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
th
e
f
in
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
p
ar
tly
av
ailab
le
f
r
o
m
Kag
g
le
at
h
ttp
s
://www.
k
ag
g
le.
co
m
/d
atasets
/alv
in
b
in
tan
g
/s
ib
i
-
d
ataset?r
eso
u
r
ce
=d
o
wn
lo
a
d
.
R
estrictio
n
s
ap
p
ly
t
o
th
e
av
ailab
ilit
y
o
f
t
h
ese
d
ata,
w
h
ich
wer
e
u
s
ed
u
n
d
er
licen
s
e
f
o
r
th
is
s
tu
d
y
.
I
n
ad
d
itio
n
,
th
is
s
tu
d
y
also
e
m
p
lo
y
e
d
cu
s
to
m
ized
d
ata
g
e
n
er
ated
b
y
th
e
au
th
o
r
s
.
T
h
ese
cu
s
to
m
ized
d
ata
ar
e
av
ailab
le
f
r
o
m
th
e
au
th
o
r
s
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est,
with
th
e
p
er
m
is
s
io
n
o
f
Kag
g
le.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
R
ec
o
g
n
itio
n
o
f I
n
d
o
n
esia
n
s
ig
n
la
n
g
u
a
g
e
u
s
in
g
d
ee
p
lea
r
n
in
g
…
(
Olivia
K
emb
u
a
n
)
5015
RE
F
E
R
E
NC
E
S
[
1
]
E.
A
l
b
a
l
a
w
i
e
t
a
l
.
,
“
I
n
t
e
g
r
a
t
e
d
a
p
p
r
o
a
c
h
o
f
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
w
i
t
h
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
f
o
r
c
l
a
ssi
f
i
c
a
t
i
o
n
a
n
d
d
i
a
g
n
o
si
s
o
f
b
r
a
i
n
t
u
m
o
r
,
”
B
MC
Me
d
i
c
a
l
I
m
a
g
i
n
g
,
v
o
l
.
2
4
,
n
o
.
1
,
p
p
.
1
–
1
5
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
8
6
/
s1
2
8
8
0
-
0
2
4
-
0
1
2
6
1
-
0.
[
2
]
H
.
A
o
u
a
n
i
a
n
d
Y
.
B
.
A
y
e
d
,
“
D
e
e
p
f
a
c
i
a
l
e
x
p
r
e
ssi
o
n
d
e
t
e
c
t
i
o
n
u
si
n
g
V
i
o
l
a
a
n
d
Jo
n
e
s
a
l
g
o
r
i
t
h
m,
C
N
N
-
M
LP
a
n
d
C
N
N
-
S
V
M
,
”
S
o
c
i
a
l
N
e
t
w
o
r
k
A
n
a
l
y
si
s
a
n
d
M
i
n
i
n
g
,
v
o
l
.
1
4
,
n
o
.
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
3
2
7
8
-
024
-
0
1
2
3
1
-
y.
[
3
]
L.
W
a
n
g
,
X
.
K
a
n
g
,
F
.
D
i
n
g
,
S
.
N
a
k
a
g
a
w
a
,
a
n
d
F
.
R
e
n
,
“
A
j
o
i
n
t
l
o
c
a
l
s
p
a
t
i
a
l
a
n
d
g
l
o
b
a
l
t
e
mp
o
r
a
l
C
N
N
-
Tr
a
n
sf
o
r
mer
f
o
r
d
y
n
a
m
i
c
f
a
c
i
a
l
e
x
p
r
e
ss
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
A
p
p
l
i
e
d
S
o
f
t
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
6
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
s
o
c
.
2
0
2
4
.
1
1
1
6
8
0
.
[
4
]
M
.
D
.
Ze
i
l
e
r
a
n
d
R
.
F
e
r
g
u
s,
“
V
i
s
u
a
l
i
z
i
n
g
a
n
d
u
n
d
e
r
s
t
a
n
d
i
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
t
w
o
r
k
s,
”
i
n
C
o
m
p
u
t
e
r
Vi
s
i
o
n
-
E
C
C
V
2
0
1
4
,
2
0
1
4
,
p
p
.
8
1
8
-
8
3
3
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
3
1
9
-
1
0
5
9
0
-
1
_
5
3
.
[
5
]
C
.
Q
i
n
g
z
h
e
n
g
,
T.
Q
i
n
g
,
Z.
M
u
c
h
a
o
,
a
n
d
M
.
Lu
y
a
o
,
“
C
N
N
-
b
a
s
e
d
g
e
st
u
r
e
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
r
a
w
n
u
m
e
r
i
c
a
l
g
r
a
y
-
sca
l
e
i
m
a
g
e
s
o
f
su
r
f
a
c
e
e
l
e
c
t
r
o
my
o
g
r
a
p
h
y
,
”
B
i
o
m
e
d
i
c
a
l
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
a
n
d
C
o
n
t
r
o
l
,
v
o
l
.
1
0
1
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
b
sp
c
.
2
0
2
4
.
1
0
7
1
7
6
.
[
6
]
N
.
Th
a
k
u
r
,
P
.
K
u
m
a
r
,
a
n
d
A
.
K
u
mar,
“
M
u
l
t
i
l
e
v
e
l
se
ma
n
t
i
c
s
e
g
m
e
n
t
a
t
i
o
n
a
n
d
o
p
t
i
ma
l
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
b
a
s
e
d
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
(
O
p
-
C
N
N
)
f
o
r
b
r
e
a
st
c
a
n
c
e
r
i
d
e
n
t
i
f
i
c
a
t
i
o
n
a
n
d
c
l
a
ssi
f
i
c
a
t
i
o
n
u
si
n
g
mamm
o
g
r
a
m
i
m
a
g
e
s,
”
B
i
o
m
e
d
i
c
a
l
S
i
g
n
a
l
Pr
o
c
e
ss
i
n
g
a
n
d
C
o
n
t
ro
l
,
v
o
l
.
1
0
3
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
b
s
p
c
.
2
0
2
4
.
1
0
7
3
7
4
.
[
7
]
A
.
P
a
t
i
l
a
n
d
M
.
R
a
n
e
,
“
C
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
:
A
n
o
v
e
r
v
i
e
w
a
n
d
i
t
s
a
p
p
l
i
c
a
t
i
o
n
s
i
n
p
a
t
t
e
r
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
S
m
a
r
t
I
n
n
o
v
a
t
i
o
n
,
S
y
st
e
m
s
a
n
d
T
e
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
1
9
5
,
p
p
.
2
1
–
3
0
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
981
-
15
-
7
0
7
8
-
0
_
3
.
[
8
]
D
.
G
e
r
t
s
v
o
l
f
,
M
.
H
o
r
v
a
t
,
D
.
A
sl
a
m,
A
.
K
h
a
d
e
mi
,
a
n
d
U
.
B
e
r
a
r
d
i
,
“
A
U
-
n
e
t
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
a
p
p
l
i
c
a
t
i
o
n
f
o
r
i
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
e
n
e
r
g
y
l
o
ss
i
n
i
n
f
r
a
r
e
d
t
h
e
r
m
o
g
r
a
p
h
i
c
i
m
a
g
e
s,”
Ap
p
l
i
e
d
E
n
e
r
g
y
,
v
o
l
.
3
6
0
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
p
e
n
e
r
g
y
.
2
0
2
4
.
1
2
2
6
9
6
.
[
9
]
S
.
M
o
h
a
p
a
t
r
a
,
P
.
S
.
Je
j
i
,
G
.
K
.
P
a
t
i
,
M
.
M
i
s
h
r
a
,
a
n
d
T.
S
w
a
r
n
k
a
r
,
“
C
o
mp
a
r
a
t
i
v
e
e
x
p
l
o
r
a
t
i
o
n
o
f
d
e
e
p
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
u
si
n
g
r
e
a
l
-
t
i
me
e
n
d
o
sc
o
p
y
i
mag
e
s,”
Bi
o
m
e
d
i
c
a
l
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
8
,
p
p
.
1
–
1
6
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
b
mt
.
2
0
2
4
.
0
9
.
0
0
3
.
[
1
0
]
G
.
X
i
e
,
L.
W
a
n
g
,
R
.
A
.
W
i
l
l
i
a
ms,
Y
.
Li
,
P
.
Zh
a
n
g
,
a
n
d
S
.
G
u
,
“
S
e
g
me
n
t
a
t
i
o
n
o
f
w
o
o
d
C
T
i
ma
g
e
s
f
o
r
i
n
t
e
r
n
a
l
d
e
f
e
c
t
s
d
e
t
e
c
t
i
o
n
b
a
s
e
d
o
n
C
N
N
:
A
c
o
m
p
a
r
a
t
i
v
e
st
u
d
y
,
”
C
o
m
p
u
t
e
rs
a
n
d
E
l
e
c
t
ro
n
i
c
s
i
n
A
g
ri
c
u
l
t
u
r
e
,
v
o
l
.
2
2
4
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
a
g
.
2
0
2
4
.
1
0
9
2
4
4
.
[
1
1
]
L.
S
c
h
n
e
i
d
e
r
,
A
.
K
r
a
so
w
s
k
i
,
V
.
P
i
t
c
h
i
k
a
,
L.
B
o
m
b
e
c
k
,
F
.
S
c
h
w
e
n
d
i
c
k
e
,
a
n
d
M
.
B
ü
t
t
n
e
r
,
“
A
ssessm
e
n
t
o
f
C
N
N
s,
t
r
a
n
sf
o
r
mers
,
a
n
d
h
y
b
r
i
d
a
r
c
h
i
t
e
c
t
u
r
e
s
i
n
d
e
n
t
a
l
i
ma
g
e
s
e
g
m
e
n
t
a
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
D
e
n
t
i
st
r
y
,
v
o
l
.
1
5
6
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
d
e
n
t
.
2
0
2
5
.
1
0
5
6
6
8
.
[
1
2
]
S
.
Li
a
n
d
C
.
H
u
a
n
g
,
“
U
si
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
o
r
i
ma
g
e
s
e
m
a
n
t
i
c
s
e
g
me
n
t
a
t
i
o
n
a
n
d
o
b
j
e
c
t
d
e
t
e
c
t
i
o
n
,
”
S
y
s
t
e
m
s
a
n
d
S
o
f
t
C
o
m
p
u
t
i
n
g
,
v
o
l
.
6
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
sas
c
.
2
0
2
4
.
2
0
0
1
7
2
.
[
1
3
]
K
.
B
a
n
t
u
p
a
l
l
i
a
n
d
Y
.
X
i
e
,
“
A
meri
c
a
n
s
i
g
n
l
a
n
g
u
a
g
e
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
c
o
mp
u
t
e
r
v
i
si
o
n
,
”
i
n
2
0
1
8
I
EE
E
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
B
i
g
D
a
t
a
(
Bi
g
D
a
t
a
)
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
B
i
g
D
a
t
a
.
2
0
1
8
.
8
6
2
2
1
4
1
.
[
1
4
]
A
.
A
l
j
a
b
a
r
a
n
d
S
u
h
a
r
j
i
t
o
,
“
B
I
S
I
N
D
O
(
B
a
h
a
sa
I
sy
a
r
a
t
I
n
d
o
n
e
s
i
a
)
s
i
g
n
l
a
n
g
u
a
g
e
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
C
N
N
a
n
d
LST
M
,
”
A
d
v
a
n
c
e
s
i
n
S
c
i
e
n
c
e
,
T
e
c
h
n
o
l
o
g
y
a
n
d
E
n
g
i
n
e
e
r
i
n
g
S
y
s
t
e
m
s
J
o
u
r
n
a
l
,
v
o
l
.
5
,
n
o
.
5
,
p
p
.
2
8
2
–
2
8
7
,
2
0
2
0
,
d
o
i
:
1
0
.
2
5
0
4
6
/
a
j
0
5
0
5
3
5
.
[
1
5
]
T.
H
a
n
d
h
i
k
a
,
R
.
I
.
M
.
Z
e
n
,
M
u
r
n
i
,
D
.
P
.
L
e
st
a
r
i
,
a
n
d
I
.
S
a
r
i
,
“
G
e
st
u
r
e
r
e
c
o
g
n
i
t
i
o
n
f
o
r
I
n
d
o
n
e
si
a
n
S
i
g
n
L
a
n
g
u
a
g
e
(
B
I
S
I
N
D
O
)
,
”
i
n
J
o
u
rn
a
l
o
f
P
h
y
s
i
c
s:
C
o
n
f
e
r
e
n
c
e
S
e
r
i
e
s
2
n
d
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
t
a
t
i
st
i
c
s,
M
a
t
h
e
m
a
t
i
c
s
,
T
e
a
c
h
i
n
g
,
v
o
l
.
1
0
2
8
,
n
o
.
1
,
2
0
1
8
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
4
2
-
6
5
9
6
/
1
0
2
8
/
1
/
0
1
2
1
7
3
.
[
1
6
]
M
.
C
.
A
r
i
e
st
a
,
F
.
W
i
r
y
a
n
a
,
S
u
h
a
r
j
i
t
o
,
a
n
d
A
.
Z
a
h
r
a
,
“
S
e
n
t
e
n
c
e
l
e
v
e
l
I
n
d
o
n
e
si
a
n
si
g
n
l
a
n
g
u
a
g
e
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
3
D
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
a
n
d
b
i
d
i
r
e
c
t
i
o
n
a
l
r
e
c
u
r
r
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
1
s
t
I
n
d
o
n
e
s
i
a
n
Ass
o
c
i
a
t
i
o
n
f
o
r
P
a
t
t
e
r
n
Re
c
o
g
n
i
t
i
o
n
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
(
I
N
APR)
,
Ju
l
.
2
0
1
8
,
p
p
.
1
6
–
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
I
N
A
P
R
.
2
0
1
8
.
8
6
2
7
0
1
6
.
[
1
7
]
A
.
K
a
s
a
p
b
a
ş
i
,
A
.
E.
A
.
E
l
b
u
s
h
r
a
,
O
.
A
l
-
H
a
r
d
a
n
e
e
,
a
n
d
A
.
Y
i
l
maz
,
“
D
e
e
p
A
S
LR
:
A
C
N
N
-
b
a
se
d
h
u
m
a
n
–
c
o
m
p
u
t
e
r
i
n
t
e
r
f
a
c
e
f
o
r
A
meric
a
n
si
g
n
l
a
n
g
u
a
g
e
r
e
c
o
g
n
i
t
i
o
n
f
o
r
h
e
a
r
i
n
g
-
i
mp
a
i
r
e
d
i
n
d
i
v
i
d
u
a
l
s
,
”
C
o
m
p
u
t
e
r
M
e
t
h
o
d
s
a
n
d
Pro
g
r
a
m
s i
n
B
i
o
m
e
d
i
c
i
n
e
U
p
d
a
t
e
,
v
o
l
.
2
,
Ja
n
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
mp
b
u
p
.
2
0
2
1
.
1
0
0
0
4
8
.
[
1
8
]
M
.
S
.
M
a
r
c
o
l
i
n
o
e
t
a
l
.
,
“
S
i
g
n
l
a
n
g
u
a
g
e
r
e
c
o
g
n
i
t
i
o
n
s
y
st
e
m
f
o
r
d
e
a
f
p
a
t
i
e
n
t
s:
p
r
o
t
o
c
o
l
f
o
r
a
sy
s
t
e
m
a
t
i
c
r
e
v
i
e
w
,
”
J
MIR
Re
s
e
a
r
c
h
Pro
t
o
c
o
l
s
,
v
o
l
.
1
4
,
2
0
2
5
,
d
o
i
:
1
0
.
2
1
9
6
/
5
5
4
2
7
.
[
1
9
]
M
.
A
l
su
l
a
i
ma
n
e
t
a
l
.
,
“
F
a
c
i
l
i
t
a
t
i
n
g
t
h
e
c
o
mm
u
n
i
c
a
t
i
o
n
w
i
t
h
d
e
a
f
p
e
o
p
l
e
:
b
u
i
l
d
i
n
g
a
l
a
r
g
e
st
S
a
u
d
i
si
g
n
l
a
n
g
u
a
g
e
d
a
t
a
se
t
,
”
J
o
u
rn
a
l
o
f
K
i
n
g
S
a
u
d
U
n
i
v
e
rs
i
t
y
–
C
o
m
p
u
t
e
r
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
3
5
,
n
o
.
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
k
s
u
c
i
.
2
0
2
3
.
1
0
1
6
4
2
.
[
2
0
]
Z.
K
a
t
ı
l
mı
ş
a
n
d
C
.
K
a
r
a
k
u
z
u
,
“
EL
M
-
b
a
s
e
d
t
w
o
-
h
a
n
d
e
d
d
y
n
a
mi
c
Tu
r
k
i
s
h
s
i
g
n
l
a
n
g
u
a
g
e
(
TSL)
w
o
r
d
r
e
c
o
g
n
i
t
i
o
n
,
”
E
x
p
e
rt
S
y
st
e
m
s
w
i
t
h
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
8
2
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
2
1
.
1
1
5
2
1
3
.
[
2
1
]
Z.
K
a
t
ı
l
m
ı
ş
a
n
d
C
.
K
a
r
a
k
u
z
u
,
“
D
o
u
b
l
e
-
h
a
n
d
e
d
d
y
n
a
m
i
c
T
u
r
k
i
sh
s
i
g
n
l
a
n
g
u
a
g
e
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
Le
a
p
M
o
t
i
o
n
w
i
t
h
met
a
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
Ex
p
e
rt
S
y
s
t
e
m
s
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
2
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
2
3
.
1
2
0
4
5
3
.
[
2
2
]
S
.
S
i
d
d
i
q
u
e
,
S
.
I
sl
a
m,
E.
N
e
o
n
,
T.
S
a
b
b
i
r
,
I
.
N
a
h
e
e
n
,
a
n
d
R
.
K
h
a
n
,
“
D
e
e
p
l
e
a
r
n
i
n
g
-
b
a
s
e
d
B
a
n
g
l
a
s
i
g
n
l
a
n
g
u
a
g
e
d
e
t
e
c
t
i
o
n
w
i
t
h
a
n
e
d
g
e
d
e
v
i
c
e
,
”
I
n
t
e
l
l
i
g
e
n
t
S
y
st
e
m
s
w
i
t
h
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
8
,
M
a
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
sw
a
.
2
0
2
3
.
2
0
0
2
2
4
.
[
2
3
]
E.
S
.
El
f
a
r
,
D
.
M
.
A
.
K
i
s
h
k
,
A
.
M
.
I
b
r
a
h
i
m,
a
n
d
S
.
E.
A
b
d
e
l
r
a
o
u
f
,
“
S
i
l
e
n
t
l
i
f
e
sav
e
r
s:
B
r
e
a
k
i
n
g
b
a
r
r
i
e
r
s wi
t
h
a
si
g
n
l
a
n
g
u
a
g
e
h
e
a
l
t
h
e
d
u
c
a
t
i
o
n
v
i
d
e
o
f
o
r
s
t
u
d
e
n
t
s
w
i
t
h
d
e
a
f
n
e
ss
o
n
sc
h
o
o
l
f
i
r
st
a
i
d
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
A
f
ri
c
a
N
u
rs
i
n
g
S
c
i
e
n
c
e
s
,
v
o
l
.
2
0
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
j
a
n
s
.
2
0
2
4
.
1
0
0
7
2
5
.
[
2
4
]
M
.
S
a
n
a
u
l
l
a
h
e
t
a
l
.
,
“
S
i
g
n
l
a
n
g
u
a
g
e
t
o
se
n
t
e
n
c
e
f
o
r
m
a
t
i
o
n
:
a
r
e
a
l
t
i
me
so
l
u
t
i
o
n
f
o
r
d
e
a
f
p
e
o
p
l
e
,
”
C
o
m
p
u
t
e
r
s,
Ma
t
e
r
i
a
l
s
a
n
d
C
o
n
t
i
n
u
a
,
v
o
l
.
7
2
,
n
o
.
2
,
p
p
.
2
5
0
1
–
2
5
1
9
,
2
0
2
2
,
d
o
i
:
1
0
.
3
2
6
0
4
/
c
mc
.
2
0
2
2
.
0
2
1
9
9
0
.
[
2
5
]
B
.
G
a
r
g
e
t
a
l
.
,
“
S
i
g
n
l
a
n
g
u
a
g
e
d
e
t
e
c
t
i
o
n
d
a
t
a
s
e
t
:
a
r
e
so
u
r
c
e
f
o
r
A
I
-
b
a
sed
r
e
c
o
g
n
i
t
i
o
n
s
y
st
e
ms,
”
D
a
t
a
i
n
Br
i
e
f
,
v
o
l
.
6
1
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
d
i
b
.
2
0
2
5
.
1
1
1
7
0
3
.
[
2
6
]
E.
P
.
-
M
o
n
t
i
e
l
e
t
a
l
.
,
“
A
u
t
o
m
a
t
i
c
si
g
n
l
a
n
g
u
a
g
e
r
e
c
o
g
n
i
t
i
o
n
b
a
s
e
d
o
n
a
c
c
e
l
e
r
o
met
r
y
a
n
d
s
u
r
f
a
c
e
e
l
e
c
t
r
o
m
y
o
g
r
a
p
h
y
s
i
g
n
a
l
s
:
a
s
t
u
d
y
f
o
r
C
o
l
o
m
b
i
a
n
si
g
n
l
a
n
g
u
a
g
e
,
”
Bi
o
m
e
d
i
c
a
l
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
a
n
d
C
o
n
t
ro
l
,
v
o
l
.
7
1
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
b
sp
c
.
2
0
2
1
.
1
0
3
2
0
1
.
[
2
7
]
I
.
G
.
B
.
H
.
W
i
d
h
i
n
u
g
r
a
h
a
a
n
d
E.
R
a
k
u
n
,
“
I
n
d
o
n
e
s
i
a
n
l
a
n
g
u
a
g
e
si
g
n
s
y
st
e
m
(
S
I
B
I
)
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
t
h
r
e
s
h
o
l
d
c
o
n
d
i
t
i
o
n
a
l
r
a
n
d
o
m
f
i
e
l
d
s
,
”
I
C
C
PR
'
1
9
:
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
1
9
8
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
p
u
t
i
n
g
a
n
d
P
a
t
t
e
r
n
Re
c
o
g
n
i
t
i
o
n
,
2
0
1
9
,
p
p
.
3
8
0
-
3
8
4
,
d
o
i
:
1
0
.
1
1
4
5
/
3
3
7
3
5
0
9
.
3
3
7
3
5
9
1
.
[
2
8
]
A
.
K
.
S
.
a
n
d
M
.
G
o
k
i
l
a
v
a
n
i
,
“
A
st
u
d
y
o
f
me
d
i
c
a
l
i
ma
g
e
p
r
o
c
e
ss
i
n
g
a
n
d
s
e
g
me
n
t
a
t
i
o
n
me
t
h
o
d
s,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
n
o
v
a
t
i
v
e
R
e
se
a
rc
h
i
n
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
0
,
n
o
.
1
0
,
p
p
.
6
0
9
–
6
1
5
,
2
0
1
9
,
d
o
i
:
1
0
.
2
6
5
6
2
/
I
JI
R
A
E.
2
0
1
9
.
O
C
A
E1
0
0
8
2
.
[
2
9
]
A
.
D
e
r
a
t
,
“
A
p
p
l
i
e
d
d
e
e
p
l
e
a
r
n
i
n
g
–
p
a
r
t
4
:
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s,”
T
o
w
a
r
d
s
D
a
t
a
S
c
i
e
n
c
e
,
2
0
1
7
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
t
o
w
a
r
d
sd
a
t
a
sc
i
e
n
c
e
.
c
o
m/
a
p
p
l
i
e
d
-
d
e
e
p
-
l
e
a
r
n
i
n
g
-
p
a
r
t
-
4
-
c
o
n
v
o
l
u
t
i
o
n
a
l
-
n
e
u
r
a
l
-
n
e
t
w
o
r
k
s
-
5
8
4
b
c
1
3
4
c
1
e
2
.
[
3
0
]
S
.
S
a
h
a
,
“
A
c
o
mp
r
e
h
e
n
s
i
v
e
g
u
i
d
e
t
o
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
–
t
h
e
E
LI
5
w
a
y
,
”
M
e
d
i
u
m
,
2
0
1
8
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
t
o
w
a
r
d
sd
a
t
a
sc
i
e
n
c
e
.
c
o
m/
a
-
c
o
m
p
r
e
h
e
n
si
v
e
-
g
u
i
d
e
-
to
-
c
o
n
v
o
l
u
t
i
o
n
a
l
-
n
e
u
r
a
l
-
n
e
t
w
o
r
k
s
-
t
h
e
-
e
l
i
5
-
w
a
y
-
3
b
d
2
b
1
1
6
4
a
5
3
.
[
3
1
]
A
.
G
h
o
sh
,
A
.
S
u
f
i
a
n
,
F
.
S
u
l
t
a
n
a
,
A
.
C
h
a
k
r
a
b
a
r
t
i
,
a
n
d
D
.
D
e
,
“
F
u
n
d
a
m
e
n
t
a
l
c
o
n
c
e
p
t
s
o
f
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
i
n
Re
c
e
n
t
T
ren
d
s
i
n
S
i
g
n
a
l
a
n
d
I
m
a
g
e
Pr
o
c
e
ss
i
n
g
,
v
o
l
.
1
7
2
,
2
0
1
9
,
p
p
.
5
1
9
–
5
2
8
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
0
-
3
2
6
4
4
-
9
_
3
6
.
[
3
2
]
V
.
N
a
i
r
a
n
d
G
.
E.
H
i
n
t
o
n
,
“
R
e
c
t
i
f
i
e
d
l
i
n
e
a
r
u
n
i
t
s
i
mp
r
o
v
e
r
e
st
r
i
c
t
e
d
B
o
l
t
z
man
n
ma
c
h
i
n
e
s,
”
i
n
I
C
ML'
1
0
:
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
7
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
M
a
c
h
i
n
e
L
e
a
rn
i
n
g
,
2
0
1
0
,
p
p
.
8
0
7
–
8
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
5
0
0
8
-
5
0
1
6
5016
[
3
3
]
N
.
S
r
i
v
a
st
a
v
a
,
G
.
H
i
n
t
o
n
,
A
.
K
r
i
z
h
e
v
sk
y
,
I
.
S
u
t
s
k
e
v
e
r
,
a
n
d
R
.
S
a
l
a
k
h
u
t
d
i
n
o
v
,
“
D
r
o
p
o
u
t
:
a
si
mp
l
e
w
a
y
t
o
p
r
e
v
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
r
o
m o
v
e
r
f
i
t
t
i
n
g
,
”
J
o
u
r
n
a
l
o
f
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
R
e
se
a
rc
h
,
v
o
l
.
1
5
,
p
p
.
1
9
2
9
–
1
9
5
8
,
2
0
1
4
.
[
3
4
]
M
.
I
b
r
a
h
i
m,
A
.
S
h
a
a
w
a
t
,
a
n
d
M
.
To
r
k
i
,
“
C
o
v
a
r
i
a
n
c
e
p
o
o
l
i
n
g
l
a
y
e
r
f
o
r
t
e
x
t
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
Pr
o
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
8
9
,
p
p
.
6
1
–
6
6
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.
2
0
2
1
.
0
5
.
0
7
0
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
O
li
v
ia
K
e
m
b
u
a
n
re
c
e
iv
e
d
th
e
B.
En
g
.
d
e
g
re
e
in
I
n
fo
rm
a
ti
c
s
En
g
in
e
e
rin
g
fr
o
m
th
e
S
a
ty
a
Wac
a
n
a
Ch
r
isti
n
Un
i
v
e
rsit
y
,
Ce
n
tral
Ja
v
a
,
In
d
o
n
e
sia
,
in
2
0
1
0
,
a
n
d
t
h
e
m
a
ste
r’s
d
e
g
re
e
in
I
n
fo
rm
a
ti
c
s
E
n
g
i
n
e
e
rin
g
fro
m
th
e
G
a
d
jah
M
a
d
a
Un
i
v
e
rsity
,
Yo
g
y
a
k
a
rta,
I
n
d
o
n
e
sia
,
i
n
2
0
1
2
.
S
h
e
is
c
u
rre
n
tl
y
a
le
c
tu
re
r
o
f
th
e
De
p
a
rtme
n
t
o
f
In
f
o
rm
a
ti
c
s
En
g
i
n
e
e
rin
g
,
Un
i
v
e
rsity
o
f
M
a
n
a
d
o
,
In
d
o
n
e
sia
.
He
r
c
u
rre
n
t
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
m
a
c
h
in
e
lea
rn
i
n
g
,
c
o
m
p
u
ter
n
e
two
rk
i
n
g
,
a
n
d
a
u
g
m
e
n
ted
re
a
li
ty
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
o
li
v
ia
k
e
m
b
u
a
n
@
u
n
ima
.
a
c
.
id
.
H
a
r
y
a
n
t
o
is
a
lec
tu
re
r
i
n
th
e
G
ra
d
u
a
te
S
c
h
o
o
l
a
t
Y
o
g
y
a
k
a
rta
S
tate
Un
iv
e
rsit
y
(UN
Y),
In
d
o
n
e
sia
.
His
re
se
a
rc
h
i
n
tere
sts
c
e
n
tere
d
o
n
a
rti
ficia
l
in
te
ll
ig
e
n
t
c
o
n
tro
l,
e
d
u
c
a
ti
o
n
a
l
re
se
a
rc
h
a
n
d
e
v
a
lu
a
ti
o
n
,
a
n
d
te
c
h
n
ica
l
a
n
d
v
o
c
a
ti
o
n
a
l
e
d
u
c
a
ti
o
n
a
n
d
train
in
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
h
a
r
y
a
n
to
@
u
n
y
.
a
c
.
id
.
Mo
c
h
a
m
a
d
Br
u
r
i
Tr
iy
o
n
o
i
s
a
P
ro
fe
ss
o
r
i
n
t
h
e
G
ra
d
u
a
te
S
c
h
o
o
l,
Y
o
g
y
a
k
a
rt
a
S
tate
Un
iv
e
rsit
y
(UN
Y),
In
d
o
n
e
sia
.
His
re
se
a
rc
h
e
x
p
e
rti
se
e
n
c
o
m
p
a
ss
e
s
v
o
c
a
ti
o
n
a
l
e
d
u
c
a
ti
o
n
a
n
d
train
i
n
g
(VE
T),
c
u
rricu
lu
m
in
n
o
v
a
ti
o
n
,
lea
rn
i
n
g
i
n
n
o
v
a
ti
o
n
,
v
o
c
a
ti
o
n
a
l
tea
c
h
e
r
p
ro
fe
ss
io
n
a
l
d
e
v
e
lo
p
m
e
n
t,
a
n
d
p
a
rtn
e
rsh
ip
s
b
e
twe
e
n
v
o
c
a
ti
o
n
a
l
s
c
h
o
o
ls
a
n
d
i
n
d
u
stry
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
ru
ri
tri
y
o
n
o
@u
n
y
.
a
c
.
id
.
Evaluation Warning : The document was created with Spire.PDF for Python.