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.
15
,
No
.
4
,
A
u
g
u
s
t
20
26
,
p
p
.
3
7
8
2
~
3
7
9
1
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
4
.
p
p
3
7
8
2
-
3
7
9
1
3782
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Perf
o
rma
nce
of
ma
jority v
o
ting tr
a
nsfer learning
d
eep learning
mo
nkey
po
x
disea
se dete
ction
Nur
Na
f
iiy
a
h,
M
uh
a
m
m
a
d
Nurul H
ud
a
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
m
a
t
i
c
s
En
g
i
n
e
e
r
i
n
g
,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
a
n
d
Te
c
h
n
o
l
o
g
y
,
U
n
i
v
e
r
si
t
a
s I
sl
a
m
La
m
o
n
g
a
n
,
L
a
m
o
n
g
a
n
,
I
n
d
o
n
e
si
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
5
,
2
0
2
5
R
ev
is
ed
J
u
l 9
,
2
0
2
6
Acc
ep
ted
J
u
l 2
1
,
2
0
2
6
G
lo
b
a
l
h
e
a
lt
h
c
o
n
c
e
rn
s
h
a
v
e
b
e
e
n
ra
ise
d
b
y
th
e
a
d
v
e
n
t
o
f
m
o
n
k
e
y
p
o
x
fo
ll
o
wi
n
g
th
e
COV
ID
-
1
9
e
p
i
d
e
m
ic,
h
ig
h
li
g
h
ti
n
g
t
h
e
n
e
e
d
fo
r
re
li
a
b
le
a
u
to
m
a
ted
s
y
ste
m
s
to
su
p
p
o
rt
e
a
rly
sk
i
n
d
ise
a
se
sc
re
e
n
in
g
.
T
h
is
st
u
d
y
p
ro
p
o
se
s
a
m
o
n
k
e
y
p
o
x
sk
i
n
d
ise
a
se
c
l
a
ss
ifi
c
a
ti
o
n
fra
m
e
wo
rk
u
sin
g
a
m
a
jo
rit
y
v
o
ti
n
g
e
n
se
m
b
le
b
a
s
e
d
o
n
tran
sfe
r
lea
rn
i
n
g
.
Th
e
e
n
se
m
b
le
c
o
m
b
in
e
s
p
re
d
icti
o
n
s
fr
o
m
m
u
lt
ip
le
p
re
train
e
d
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
two
rk
s
(CNN
s)
to
imp
r
o
v
e
c
las
sifica
ti
o
n
ro
b
u
st
n
e
ss
.
A
p
u
b
li
c
ly
a
c
c
e
s
sib
le
d
a
tas
e
t
c
o
m
p
risin
g
f
o
u
r
c
las
se
s
:
m
o
n
k
e
y
p
o
x
,
c
h
ick
e
n
p
o
x
,
m
e
a
sle
s,
a
n
d
n
o
rm
a
l
sk
i
n
wa
s
u
se
d
fo
r
th
e
e
x
p
e
rime
n
ts.
Th
e
re
su
lt
s
in
d
ica
te
th
a
t
wh
il
e
a
sin
g
le
Re
sN
e
t5
0
m
o
d
e
l
a
c
h
iev
e
d
th
e
h
ig
h
e
st
o
v
e
ra
ll
a
c
c
u
ra
c
y
(9
9
.
1
5
%
),
t
h
e
m
a
jo
rit
y
v
o
ti
n
g
a
p
p
r
o
a
c
h
y
iel
d
e
d
h
i
g
h
e
r
p
re
c
isio
n
t
h
a
n
se
v
e
ra
l
in
d
i
v
id
u
a
l
m
o
d
e
ls,
d
e
m
o
n
stra
ti
n
g
imp
r
o
v
e
d
re
li
a
b
il
it
y
in
d
ist
in
g
u
ish
in
g
m
o
n
k
e
y
p
o
x
c
a
se
s.
Th
e
se
fin
d
i
n
g
s
s
u
g
g
e
st
t
h
a
t
e
n
se
m
b
le
-
b
a
se
d
m
a
jo
ri
ty
v
o
ti
n
g
c
a
n
e
n
h
a
n
c
e
th
e
ro
b
u
stn
e
ss
o
f
m
o
n
k
e
y
p
o
x
sk
i
n
d
ise
a
se
c
las
sifica
ti
o
n
a
n
d
m
a
y
su
p
p
o
rt
c
o
m
p
u
ter
-
a
id
e
d
sc
re
e
n
in
g
sy
st
e
m
s.
K
ey
w
o
r
d
s
:
Dee
p
lear
n
in
g
Ma
jo
r
ity
v
o
tin
g
Mo
n
k
ey
p
o
x
R
esNet5
0
T
r
an
s
f
er
lear
n
i
n
g
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
:
Nu
r
Naf
iiy
ah
Dep
ar
tm
en
t o
f
I
n
f
o
r
m
atics E
n
g
in
ee
r
in
g
,
Facu
lty
o
f
Scien
ce
an
d
T
ec
h
n
o
lo
g
y
,
Un
iv
e
r
s
itas
I
s
lam
L
am
o
n
g
an
L
am
o
n
g
a
n
6
2
2
1
1
,
I
n
d
o
n
esia
E
m
ail: m
y
n
af
f
@
u
n
is
la.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
ap
p
ar
en
t
s
im
ilar
ity
an
d
v
ar
iety
in
lesi
o
n
ap
p
ea
r
an
ce
o
f
s
k
in
ca
n
ce
r
an
d
o
t
h
er
d
er
m
ato
lo
g
ical
illn
ess
es
m
ak
e
d
iag
n
o
s
is
ex
t
r
em
ely
d
if
f
icu
lt.
T
h
e
ap
p
licatio
n
o
f
d
ee
p
lear
n
in
g
an
d
c
o
n
v
o
lu
ti
o
n
al
n
e
u
r
al
n
etwo
r
k
s
(
C
NN)
f
o
r
s
k
in
lesi
o
n
ca
teg
o
r
izatio
n
h
as
b
ee
n
t
h
o
r
o
u
g
h
ly
s
tu
d
ied
in
t
h
e
p
ast
to
h
elp
d
er
m
ato
lo
g
is
ts
in
cr
ea
s
e
th
e
ac
cu
r
ac
y
o
f
th
eir
d
iag
n
o
s
es.
Fo
r
in
s
tan
ce
,
s
tu
d
ie
s
u
s
in
g
d
en
s
ely
co
n
n
ec
ted
c
o
n
v
o
lu
tio
n
al
n
etwo
r
k
(
Den
s
eNe
t)
-
2
0
1
a
n
d
s
k
in
lesi
o
n
au
g
m
en
tatio
n
s
ty
le
-
b
ased
g
en
er
ativ
e
ad
v
er
s
ar
ial
n
etwo
r
k
(
SLA
-
Sty
leGAN
)
o
b
tain
ed
a
b
alan
ce
d
m
u
lticlas
s
ac
cu
r
ac
y
(
B
MA
)
o
f
9
3
.
6
4
%,
s
h
o
win
g
th
at
g
e
n
er
ativ
e
m
o
d
els
ca
n
im
p
r
o
v
e
d
ataset
d
iv
er
s
ity
an
d
m
o
d
el
r
o
b
u
s
tn
ess
[
1
]
.
Oth
er
r
esear
ch
h
as
d
ev
elo
p
ed
au
to
m
ated
im
ag
e
-
b
ased
s
y
s
tem
s
u
s
in
g
C
NNs
an
d
s
o
f
t
m
ax
cla
s
s
if
ier
s
to
p
r
o
d
u
ce
f
aster
an
d
m
o
r
e
r
eliab
le
d
iag
n
o
s
tic
r
esu
lts
,
wh
ich
ca
n
als
o
s
er
v
e
as
r
ea
l
-
tim
e
ed
u
ca
tio
n
al
to
o
ls
f
o
r
m
ed
ical
s
tu
d
en
ts
[
2
]
.
Similar
ly
,
d
ee
p
n
eu
r
al
n
et
wo
r
k
(
DNN)
-
b
ased
r
esear
ch
co
m
b
in
in
g
im
ag
e
p
r
ep
r
o
ce
s
s
in
g
,
s
eg
m
en
tatio
n
,
f
e
atu
r
e
ex
tr
ac
tio
n
,
an
d
class
if
ic
atio
n
h
as
ac
h
iev
ed
ac
cu
r
ac
y
u
p
to
8
4
.
4
5
%
o
n
I
SI
C
2
0
1
7
d
ataset
[
3
]
.
Ad
d
itio
n
al
ly
,
o
n
th
e
HAM
1
0
0
0
0
d
ataset,
d
ee
p
C
NN
tr
ain
ed
u
s
in
g
p
r
e
p
r
o
ce
s
s
in
g
,
n
o
r
m
aliza
tio
n
,
an
d
au
g
m
en
tatio
n
s
tr
ateg
ies
ac
h
iev
ed
9
1
.
9
3
%
test
in
g
ac
cu
r
ac
y
[
4
]
.
Usi
n
g
E
f
f
icien
tNet
with
tr
an
s
f
er
lear
n
in
g
an
d
th
e
r
a
n
g
er
o
p
tim
izer
,
m
o
r
e
r
ec
en
t
s
tu
d
y
a
ch
iev
ed
a
h
ig
h
ar
ea
u
n
d
er
th
e
cu
r
v
e
(
AUC)
-
r
ec
eiv
er
o
p
er
atin
g
ch
ar
a
cter
is
tic
(
R
OC
)
s
co
r
e
o
f
0
.
9
6
8
1
o
n
I
SIC
-
2
0
1
9
an
d
I
SIC
-
2
0
2
0
d
atasets
[
5
]
.
T
h
ese
r
esear
ch
f
i
n
d
in
g
s
d
em
o
n
s
tr
ate
th
at
d
ee
p
lear
n
in
g
b
ased
im
ag
e
an
al
y
s
is
p
r
o
v
i
d
es
r
eliab
le
p
er
f
o
r
m
an
ce
in
s
k
i
n
lesi
o
n
cla
s
s
if
icatio
n
task
s
.
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
P
erfo
r
ma
n
ce
o
f m
a
jo
r
ity
vo
tin
g
tr
a
n
s
fer lea
r
n
in
g
d
ee
p
le
a
r
n
in
g
mo
n
ke
y
p
o
x
d
is
ea
s
e
d
etec
ti
o
n
(
N
u
r
N
a
fiiya
h
)
3783
C
las
s
if
y
in
g
m
elan
o
m
a
an
d
o
th
er
s
k
in
m
alig
n
an
cies
h
as
ad
v
an
ce
d
s
ig
n
if
ican
tly
,
b
u
t
litt
le
s
tu
d
y
h
as
b
ee
n
d
o
n
e
o
n
in
f
ec
tio
u
s
s
k
in
c
o
n
d
itio
n
s
lik
e
m
o
n
k
e
y
p
o
x
,
wh
ich
h
as r
esu
r
f
ac
ed
as a
wo
r
ld
wid
e
h
ea
lth
co
n
ce
r
n
s
in
ce
th
e
C
OVI
D
-
1
9
p
a
n
d
em
ic.
Mo
s
t
co
m
m
o
n
in
C
en
tr
al
an
d
W
est
Af
r
ica,
m
o
n
k
ey
p
o
x
is
a
zo
o
n
o
tic
v
ir
al
illn
ess
th
at
ca
n
tr
an
s
f
er
f
r
o
m
wild
an
im
als
to
h
u
m
an
s
.
H
o
wev
er
,
it
ca
n
also
tr
av
el
to
tr
o
p
ical
ar
ea
s
lik
e
I
n
d
o
n
esia.
Dela
y
ed
id
en
tific
atio
n
o
f
m
o
n
k
ey
p
o
x
ca
n
a
cc
eler
ate
tr
an
s
m
is
s
io
n
an
d
in
cr
ea
s
e
m
o
r
tality
.
T
h
er
ef
o
r
e,
ea
r
ly
an
d
ac
cu
r
ate
d
etec
tio
n
u
s
in
g
ar
tific
ial
in
te
llig
en
ce
(
AI
)
a
n
d
c
o
m
p
u
ter
v
is
io
n
is
cr
u
cial
f
o
r
p
r
ev
en
tin
g
f
u
r
th
er
s
p
r
ea
d
[
6
]
.
Mo
n
k
ey
p
o
x
m
ay
b
e
s
u
cc
e
s
s
f
u
lly
d
is
tin
g
u
is
h
ed
f
r
o
m
s
u
p
er
f
icially
s
im
ilar
illn
ess
es l
ik
e
ch
ick
en
p
o
x
an
d
m
ea
s
les with
th
e
u
s
e
o
f
AI
-
b
a
s
ed
m
ed
ical
im
ag
e
an
al
y
s
is
[
7
]
.
Hig
h
ac
cu
r
ac
y
r
ates
h
av
e
b
ee
n
attain
ed
in
ea
r
lier
s
tu
d
ies
o
n
d
ee
p
lear
n
in
g
-
b
ased
m
o
n
k
e
y
p
o
x
ca
teg
o
r
izatio
n
.
Fo
r
e
x
am
p
le
,
Mo
b
ileNetV2
attain
ed
9
8
.
1
6
%
ac
cu
r
ac
y
an
d
m
o
d
if
ied
Den
s
eNe
t
-
2
0
1
9
8
.
9
1
%
[
7
]
–
[
9
]
.
H
o
wev
er
,
th
e
s
e
ap
p
r
o
ac
h
es
r
ely
h
ea
v
ily
o
n
lar
g
e
an
d
b
alan
ce
d
d
atasets
to
av
o
id
o
v
er
f
itti
n
g
,
wh
ich
r
em
ain
s
a
s
ig
n
if
ican
t
lim
itatio
n
.
R
esear
ch
er
s
h
av
e
u
s
ed
p
ictu
r
e
au
g
m
en
tatio
n
ap
p
r
o
ac
h
es
in
v
o
lv
in
g
s
ev
er
al
s
k
in
d
is
ea
s
e
cla
s
s
e
s
(
c
h
ick
en
p
o
x
,
m
ea
s
les,
m
o
n
k
ey
p
o
x
,
an
d
n
o
r
m
al)
to
r
eso
lv
e
d
at
a
im
b
alan
ce
[
7
]
.
I
n
o
r
d
er
t
o
g
et
9
3
.
3
9
%
ac
cu
r
ac
y
,
s
o
m
e
r
esear
ch
h
as
also
lo
o
k
ed
in
to
e
n
s
em
b
le
ap
p
r
o
ac
h
es
th
at
in
clu
d
e
m
an
y
d
ee
p
lear
n
in
g
m
o
d
els,
in
clu
d
in
g
e
x
tr
em
e
in
ce
p
tio
n
(
Xc
ep
tio
n
)
,
I
n
ce
p
tio
n
V3
,
an
d
Den
s
eNe
t1
6
9
[
1
0
]
.
Alth
o
u
g
h
en
s
em
b
le
lear
n
in
g
i
m
p
r
o
v
es
class
if
icatio
n
r
eliab
ilit
y
,
th
er
e
is
s
till
r
o
o
m
to
en
h
an
ce
p
r
ec
is
io
n
an
d
s
tab
ilit
y
,
p
ar
ticu
lar
ly
wh
en
d
e
alin
g
with
co
m
p
le
x
s
k
in
lesi
o
n
v
ar
iatio
n
s
.
I
n
lig
h
t
o
f
t
h
ese
co
n
s
tr
ai
n
ts
,
th
is
s
tu
d
y
s
u
g
g
ests
a
m
ajo
r
ity
v
o
tin
g
-
b
ased
cla
s
s
if
icatio
n
f
r
am
ewo
r
k
th
at
in
co
r
p
o
r
ates
s
ev
er
al
tr
an
s
f
er
lear
n
in
g
m
o
d
els,
in
clu
d
in
g
R
esNe
t
5
0
,
VGG1
6
,
Mo
b
ileNet,
I
n
ce
p
tio
n
V3
,
an
d
Xce
p
tio
n
,
to
in
cr
ea
s
e
th
e
p
r
ec
is
io
n
an
d
co
n
s
is
ten
cy
o
f
th
e
ca
teg
o
r
izat
io
n
o
f
m
o
n
k
ey
p
o
x
s
k
in
d
is
ea
s
es.
Ad
d
itio
n
ally
,
an
im
a
g
e
p
r
ep
ar
atio
n
s
te
p
ca
lled
c
o
n
tr
ast
lim
ited
ad
ap
tiv
e
h
is
to
g
r
am
eq
u
aliza
tio
n
(
C
L
AHE
)
is
in
clu
d
ed
to
im
p
r
o
v
e
im
ag
e
q
u
alit
y
b
y
b
alan
cin
g
co
n
tr
ast
an
d
b
r
ig
h
tn
ess
b
ef
o
r
e
to
class
if
icatio
n
.
T
h
is
s
tu
d
y
attem
p
ts
to
ad
d
r
ess
th
e
s
h
o
r
tco
m
in
g
s
o
f
ea
r
lier
ap
p
r
o
ac
h
es
an
d
ac
h
iev
e
m
o
r
e
p
r
ec
is
io
n
an
d
r
esil
ien
ce
i
n
th
e
d
etec
tio
n
o
f
m
o
n
k
e
y
p
o
x
s
k
in
illn
ess
b
y
co
m
b
in
in
g
C
L
AHE
p
r
ep
r
o
ce
s
s
in
g
with
m
ajo
r
ity
v
o
tin
g
.
2.
M
E
T
H
O
D
T
h
is
s
ec
tio
n
o
u
tlin
es
th
e
s
tu
d
y
’
s
g
en
er
al
m
eth
o
d
o
lo
g
y
,
wh
ich
in
clu
d
es
th
e
co
llectio
n
o
f
d
atasets
,
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es,
tr
an
s
f
er
lear
n
in
g
-
b
ased
m
o
d
el
tr
ain
in
g
,
an
d
m
ajo
r
ity
v
o
ti
n
g
-
b
ased
e
n
s
em
b
le
class
if
icatio
n
.
B
y
m
er
g
in
g
p
r
ed
ictio
n
s
f
r
o
m
s
ev
er
al
C
NN
m
o
d
els,
th
e
s
u
g
g
ested
ap
p
r
o
a
ch
s
ee
k
s
to
in
cr
ea
s
e
th
e
r
o
b
u
s
tn
ess
o
f
m
o
n
k
e
y
p
o
x
s
k
in
d
is
ea
s
e
class
if
icatio
n
.
An
o
v
er
v
iew
o
f
th
e
m
eth
o
d
o
lo
g
ical
p
ip
elin
e
is
p
r
esen
ted
b
ef
o
r
e
d
etailin
g
ea
c
h
co
m
p
o
n
en
t.
2
.
1
.
Da
t
a
s
et
s
T
h
is
s
tu
d
y
’
s
s
k
in
p
ictu
r
e
co
llectio
n
ca
m
e
f
r
o
m
a
p
u
b
licly
ac
ce
s
s
ib
le
d
ataset
th
at
h
ad
b
ee
n
u
s
ed
in
ea
r
lier
s
tu
d
ies
[
7
]
.
T
ab
le
1
s
u
m
m
ar
izes
th
e
7
7
0
o
r
ig
i
n
al
p
h
o
to
s
,
wh
ich
ar
e
d
iv
id
e
d
in
to
f
o
u
r
class
es:
ch
ick
en
p
o
x
(
1
0
7
im
ag
es),
m
e
asles
(
9
1
im
ag
es),
m
o
n
k
ey
p
o
x
(
2
7
9
im
a
g
es),
an
d
n
o
r
m
al
s
k
in
(
2
9
3
im
ag
es).
Sam
p
le
p
h
o
to
g
r
ap
h
s
f
r
o
m
ea
c
h
class
ar
e
s
h
o
wn
in
Fig
u
r
e
1
.
T
ab
le
1
.
Or
ig
i
n
al
d
ataset
d
is
tr
ib
u
tio
n
No
Ty
p
e
To
t
a
l
1
C
h
i
c
k
e
n
p
o
x
1
0
7
2
M
e
a
sl
e
s
91
3
M
o
n
k
e
y
p
o
x
2
7
9
4
N
o
r
mal
2
9
3
To
t
a
l
7
7
0
T
h
e
tr
ain
in
g
d
ata
was
s
u
b
ject
ed
to
d
ata
au
g
m
en
tatio
n
-
b
ase
d
o
v
e
r
s
am
p
lin
g
in
o
r
d
er
to
c
o
r
r
ec
t
th
e
class
im
b
alan
ce
.
Ho
r
izo
n
tal
a
n
d
v
er
tical
f
lip
p
in
g
,
r
o
tatio
n
,
c
en
ter
cr
o
p
p
in
g
,
b
r
ig
h
tn
ess
an
d
co
n
tr
ast
alter
atio
n
,
s
atu
r
atio
n
an
d
h
u
e
jitt
er
in
g
,
wid
th
an
d
h
eig
h
t
s
h
if
tin
g
,
an
d
n
o
is
e
ad
d
itio
n
wer
e
a
m
o
n
g
th
e
au
g
m
en
tatio
n
tech
n
iq
u
es
u
s
ed
.
T
h
e
d
ataset
was
ex
p
an
d
e
d
to
ap
p
r
o
x
im
a
tely
1
,
2
0
0
im
ag
es
p
e
r
class
to
en
s
u
r
e
s
u
f
f
icien
t
r
ep
r
esen
tatio
n
wh
ile
m
ain
tai
n
in
g
class
b
ala
n
ce
.
As
a
r
e
s
u
lt,
th
e
f
i
n
al
au
g
m
en
te
d
d
ataset
co
n
s
is
ted
o
f
4
,
6
9
2
im
ag
es,
as
s
h
o
wn
in
T
ab
le
2
.
Af
ter
th
at,
th
e
d
ataset
was
s
p
lit
8
0
:2
0
i
n
to
t
r
ain
in
g
an
d
test
in
g
s
ets,
y
ield
in
g
4
,
2
2
2
tr
ain
i
n
g
im
a
g
e
s
an
d
4
7
0
test
in
g
im
ag
es.
T
o
av
o
id
d
ata
leak
in
g
,
d
ata
au
g
m
en
tatio
n
was
o
n
ly
u
s
ed
o
n
th
e
tr
ain
in
g
s
et.
E
x
am
p
les o
f
au
g
m
en
ted
p
h
o
to
s
ar
e
s
h
o
wn
in
Fig
u
r
e
2
.
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.
15
,
No
.
4
,
Au
g
u
s
t
20
26
:
3
7
8
2
-
3
7
9
1
3784
C
l
a
s
s:
C
h
i
c
k
e
n
p
o
x
C
l
a
s
s:
M
e
a
sl
e
s
C
l
a
s
s:
M
o
n
k
e
y
p
o
x
C
l
a
s
s:
N
o
r
ma
l
Fig
u
r
e
1
.
Data
s
et
r
esear
ch
T
ab
le
2
.
Au
g
m
en
ted
d
ataset
d
is
tr
ib
u
tio
n
No
Ty
p
e
Tr
a
i
n
Te
st
To
t
a
l
1
C
h
i
c
k
e
n
p
o
x
1
,
0
8
0
1
2
0
1
,
2
0
0
2
M
e
a
sl
e
s
9
8
2
1
1
0
1
,
0
9
2
3
M
o
n
k
e
y
p
o
x
1
,
0
8
0
1
2
0
1
,
2
0
0
4
N
o
r
mal
1
,
0
8
0
1
2
0
1
,
2
0
0
To
t
a
l
4
,
2
2
2
4
7
0
4
,
6
9
2
2
.
2
.
P
r
o
po
s
ed
m
o
del
T
h
i
s
s
t
u
d
y
u
s
e
s
a
n
e
n
s
e
m
b
le
a
r
c
h
i
t
e
c
t
u
r
e
b
a
s
e
d
o
n
m
a
j
o
r
i
t
y
v
o
t
i
n
g
f
r
o
m
s
e
v
e
r
a
l
t
r
a
n
s
f
e
r
l
e
a
r
n
i
n
g
m
o
d
e
l
s
,
i
n
c
l
u
d
i
n
g
R
es
N
et
5
0
,
V
G
G
1
6
,
M
o
b
i
l
e
Ne
t
,
I
n
c
e
p
t
i
o
n
V
3
,
a
n
d
X
c
e
p
t
i
o
n
,
t
o
c
l
a
s
s
i
f
y
f
o
u
r
s
k
i
n
c
o
n
d
i
t
i
o
n
s
:
c
h
i
c
k
e
n
p
o
x
,
m
e
a
s
l
es
,
m
o
n
k
e
y
p
o
x
,
a
n
d
n
o
r
m
a
l
s
k
i
n
.
T
h
e
o
v
e
r
a
l
l
r
es
e
a
r
c
h
f
r
a
m
ew
o
r
k
is
i
llu
s
t
r
a
t
e
d
i
n
Fi
g
u
r
e
3
.
Acc
o
r
d
in
g
to
ea
r
lier
r
esear
c
h
,
th
e
m
ajo
r
ity
v
o
tin
g
m
et
h
o
d
u
s
es
a
h
ar
d
v
o
tin
g
s
ch
em
e
in
wh
ich
th
e
f
in
al
class
lab
el
is
d
ec
id
ed
b
y
t
h
e
class
th
at
ea
ch
m
o
d
el
p
r
ed
icts
th
e
m
o
s
t
f
r
eq
u
e
n
tly
[
1
1
]
.
W
ith
o
u
t
weig
h
tin
g
,
ea
c
h
m
o
d
el
m
ak
es
an
eq
u
al
co
n
tr
i
b
u
tio
n
to
th
e
u
ltima
te
ch
o
ice.
T
h
e
ch
o
s
en
tr
an
s
f
er
lear
n
in
g
ar
ch
itectu
r
es
h
av
e
p
r
o
v
e
n
to
p
er
f
o
r
m
well
in
p
r
e
v
io
u
s
m
ed
ical
p
ictu
r
e
ca
te
g
o
r
i
za
tio
n
s
tu
d
ies
[
7
]
.
I
m
a
g
eNe
t
p
r
e
-
tr
ain
ed
weig
h
ts
wer
e
u
s
ed
to
in
itialize
ea
ch
m
o
d
el,
an
d
it
was
s
et
u
p
to
tak
e
in
p
u
t
p
h
o
to
s
with
2
2
4
×
2
2
4
p
ix
els.
A
Glo
b
alAv
er
ag
ePo
o
lin
g
2
D
lay
er
,
a
f
u
lly
co
n
n
ec
ted
Den
s
e
lay
er
with
2
5
6
n
eu
r
o
n
s
an
d
r
ec
tifie
d
lin
ea
r
u
n
it
(
R
eL
U
)
ac
tiv
atio
n
,
b
atch
n
o
r
m
aliza
tio
n
,
d
r
o
p
o
u
t
with
a
r
a
te
o
f
0
.
5
,
a
n
d
a
f
in
al
Den
s
e
lay
er
with
s
o
f
tm
ax
ac
tiv
atio
n
f
o
r
f
o
u
r
-
class
p
r
ed
i
ctio
n
wer
e
ad
d
e
d
to
th
e
b
asic
n
etwo
r
k
s
.
As il
lu
s
tr
ate
d
i
n
Fi
g
u
r
e
4
,
im
a
g
e
p
r
ep
r
o
c
ess
i
n
g
was
c
ar
r
i
e
d
o
u
t
u
s
in
g
C
L
AH
E
to
i
m
p
r
o
v
e
c
o
n
tr
ast
a
n
d
less
e
n
l
ig
h
t
v
ar
ia
b
il
it
y
b
e
f
o
r
e
tr
a
in
in
g
.
T
h
e
A
d
a
m
o
p
ti
m
iz
e
r
was
u
s
e
d
to
t
r
ai
n
t
h
e
m
o
d
e
l
at
a
le
ar
n
i
n
g
r
a
te
o
f
0
.
0
0
0
1
.
A
cc
u
r
ac
y
,
wh
ic
h
was
co
m
p
u
te
d
u
s
in
g
(
1
)
,
was
th
e
m
ai
n
m
et
r
i
c
u
s
ed
to
ass
ess
m
o
d
el
p
er
f
o
r
m
a
n
c
e
[
9
]
.
T
h
is
s
tu
d
y
co
n
tr
ib
u
tes
b
y
i
n
v
e
s
tig
atin
g
th
e
ef
f
ec
tiv
en
ess
o
f
in
d
iv
id
u
al
tr
an
s
f
er
lear
n
in
g
m
o
d
els
an
d
ass
ess
in
g
th
e
r
esil
ien
ce
o
f
a
m
ajo
r
ity
v
o
tin
g
en
s
em
b
le
in
co
n
ju
n
ctio
n
with
co
n
tr
ast
en
h
an
ce
m
en
t
p
r
ep
r
o
ce
s
s
in
g
f
o
r
th
e
ca
teg
o
r
izatio
n
o
f
m
o
n
k
ey
p
o
x
s
k
in
d
is
ea
s
es.
2
.
3
.
M
o
del e
v
a
lua
t
i
o
n
Acc
u
r
ac
y
(
1
)
was
u
s
ed
to
ass
es
s
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
R
esNet5
0
,
VGG1
6
,
Mo
b
ileNet,
I
n
ce
p
tio
n
V3
,
a
n
d
Xce
p
tio
n
m
o
d
els.
T
h
is
m
etr
ic
o
f
f
er
s
a
th
o
r
o
u
g
h
e
v
alu
atio
n
o
f
th
e
m
o
d
el’
s
ef
f
icac
y
i
n
p
ictu
r
e
d
etec
tio
n
a
n
d
class
if
icatio
n
.
T
r
u
e
p
o
s
itiv
es
(
T
P),
tr
u
e
n
eg
ativ
es
(
T
N)
,
f
alse
p
o
s
itiv
es
(
FP
)
,
an
d
f
alse
n
eg
ativ
es
(
FN)
ar
e
all
in
clu
d
e
d
in
th
is
eq
u
atio
n
.
W
h
er
ea
s
FP
an
d
FN
r
ep
r
esen
t
in
ac
cu
r
at
e
f
o
r
ec
asts
,
T
P
an
d
T
N
r
ep
r
esen
t a
cc
u
r
ate
p
r
ed
icti
o
n
s
.
=
+
+
+
+
(
1
)
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
P
erfo
r
ma
n
ce
o
f m
a
jo
r
ity
vo
tin
g
tr
a
n
s
fer lea
r
n
in
g
d
ee
p
le
a
r
n
in
g
mo
n
ke
y
p
o
x
d
is
ea
s
e
d
etec
ti
o
n
(
N
u
r
N
a
fiiya
h
)
3785
O
r
i
g
i
n
a
l
H
o
r
i
z
o
n
t
a
l
f
l
i
p
V
e
r
t
i
c
a
l
f
l
i
p
C
e
n
t
r
a
l
c
r
o
p
B
r
i
g
h
t
n
e
ss
a
d
j
u
s
t
me
n
t
S
a
t
u
r
a
t
i
o
n
a
d
j
u
st
m
e
n
t
H
u
e
j
i
t
t
e
r
R
o
t
a
t
i
o
n
C
o
n
t
r
a
st
a
d
j
u
st
m
e
n
t
H
e
i
g
h
t
s
h
i
f
t
W
i
d
t
h
sh
i
f
t
N
o
i
se
a
u
g
me
n
t
a
t
i
o
n
Fig
u
r
e
2
.
Data
s
et
au
g
m
e
n
tatio
n
Fig
u
r
e
3
.
R
esear
ch
p
r
o
p
o
s
ed
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.
15
,
No
.
4
,
Au
g
u
s
t
20
26
:
3
7
8
2
-
3
7
9
1
3786
Fig
u
r
e
4
.
R
esu
lts
C
L
AHE
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
T
ra
ini
ng
m
et
ho
do
l
o
g
y
I
n
o
r
d
er
t
o
ca
teg
o
r
ize
f
o
u
r
k
in
d
s
o
f
s
k
in
d
is
ea
s
es:
ch
ick
en
p
o
x
,
m
ea
s
les,
m
o
n
k
e
y
p
o
x
,
an
d
n
o
r
m
al
s
k
in
,
th
is
s
tu
d
y
ass
ess
es
th
e
ef
f
ec
tiv
en
ess
o
f
tr
an
s
f
e
r
le
ar
n
in
g
m
o
d
els:
R
esNet5
0
,
VGG1
6
,
Mo
b
ileNet,
I
n
ce
p
tio
n
V3
,
an
d
Xce
p
tio
n
in
ad
d
itio
n
to
an
en
s
em
b
le
m
eth
o
d
b
ased
o
n
m
ajo
r
ity
v
o
tin
g
.
T
h
e
m
ajo
r
it
y
v
o
tin
g
s
tr
ateg
y
co
m
b
in
es
th
e
p
r
ed
ict
ed
class
lab
el
s
f
r
o
m
all
in
d
iv
id
u
al
m
o
d
els,
f
o
llo
win
g
a
h
ar
d
v
o
tin
g
s
ch
em
e
as
d
escr
ib
ed
in
p
r
e
v
io
u
s
s
tu
d
ies
[
1
1
]
.
E
ac
h
m
o
d
el
’
s
class
if
icat
io
n
ac
cu
r
ac
y
is
s
h
o
wn
in
T
ab
l
e
3
.
R
esNet5
0
h
ad
s
ig
n
if
ican
t
d
is
cr
im
in
ativ
e
c
ap
ac
ity
,
ac
h
iev
in
g
th
e
b
est
ac
cu
r
ac
y
o
f
9
9
.
1
5
%
am
o
n
g
th
e
in
d
iv
id
u
al
ar
ch
itectu
r
es.
W
ith
an
ac
cu
r
a
cy
o
f
9
8
.
9
4
%,
th
e
s
u
g
g
ested
m
ajo
r
ity
v
o
tin
g
en
s
em
b
le
o
u
tp
er
f
o
r
m
ed
VGG1
6
,
Mo
b
ileNet,
an
d
I
n
ce
p
tio
n
V
3
a
n
d
was o
n
p
ar
with
th
e
to
p
-
p
e
r
f
o
r
m
in
g
s
in
g
le
m
o
d
el.
T
ab
le
3
.
E
v
alu
atio
n
r
esu
lts
No
M
o
d
e
l
A
c
c
u
r
a
c
y
(
%)
1
R
e
sN
e
t
5
0
9
9
.
1
5
2
V
G
G
1
6
9
6
.
5
9
3
M
o
b
i
l
e
N
e
t
9
7
.
8
7
4
I
n
c
e
p
t
i
o
n
V
3
9
8
.
2
9
5
X
c
e
p
t
i
o
n
9
8
.
9
3
6
M
a
j
o
r
i
t
y
v
o
t
i
n
g
9
8
.
9
4
Alth
o
u
g
h
th
e
e
n
s
em
b
le
d
id
n
o
t
s
u
r
p
ass
R
esNet5
0
in
ter
m
s
o
f
o
v
er
all
ac
cu
r
ac
y
,
m
aj
o
r
ity
v
o
tin
g
p
r
o
v
id
es
im
p
r
o
v
ed
r
o
b
u
s
tn
ess
b
y
ag
g
r
eg
atin
g
p
r
ed
ictio
n
s
f
r
o
m
m
u
ltip
le
m
o
d
els,
t
h
er
eb
y
r
e
d
u
cin
g
th
e
in
f
lu
en
ce
o
f
in
d
iv
id
u
al
m
o
d
el
er
r
o
r
s
.
T
h
is
ch
ar
ac
ter
is
t
ic
is
p
ar
ticu
lar
ly
im
p
o
r
tan
t
in
m
ed
ical
im
a
g
e
class
if
icatio
n
,
wh
er
e
p
r
e
d
ictio
n
r
eliab
ilit
y
is
cr
itical.
T
h
e
ac
cu
r
ac
y
a
n
d
lo
s
s
cu
r
v
es
f
o
r
e
a
ch
tr
an
s
f
er
lear
n
in
g
m
o
d
el
d
u
r
in
g
tr
ain
in
g
an
d
v
a
lid
atio
n
ar
e
s
h
o
wn
in
Fig
u
r
e
5
.
R
esNet5
0
d
em
o
n
s
tr
ates
f
ast
co
n
v
er
g
en
ce
an
d
s
tab
le
lo
s
s
r
ed
u
ctio
n
,
wh
ile
I
n
ce
p
tio
n
V3
a
n
d
Xce
p
tio
n
ex
h
ib
it
s
m
o
o
th
er
co
n
v
er
g
e
n
ce
b
e
h
a
v
io
r
with
r
elativ
ely
lo
w
v
alid
atio
n
lo
s
s
.
T
h
ese
r
es
u
lts
in
d
icate
th
at
th
e
s
elec
ted
ar
ch
itectu
r
es
ar
e
well
-
s
u
ited
f
o
r
th
e
s
k
in
d
is
ea
s
e
class
if
icatio
n
task
.
3
.
2
.
Dis
cus
s
io
n
T
h
e
ex
p
e
r
im
en
tal
r
esu
lts
ar
e
d
is
cu
s
s
ed
in
th
is
s
ec
tio
n
b
y
c
o
n
tr
asti
n
g
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
with
r
elev
an
t
r
esear
ch
o
n
th
e
clas
s
if
icatio
n
o
f
s
k
in
d
is
ea
s
es
an
d
m
o
n
k
ey
p
o
x
.
As
s
h
o
wn
in
T
ab
le
4
,
n
u
m
e
r
o
u
s
s
tu
d
ies
h
av
e
r
e
p
o
r
te
d
h
i
g
h
cl
ass
if
icatio
n
ac
cu
r
ac
y
u
s
in
g
d
if
f
er
en
t
d
ee
p
lear
n
in
g
ar
c
h
itectu
r
es
an
d
p
r
o
b
le
m
f
o
r
m
u
latio
n
s
.
Sev
er
al
s
tu
d
ies
h
av
e
ac
h
ie
v
ed
v
e
r
y
h
ig
h
a
cc
u
r
ac
y
,
in
cl
u
d
in
g
Po
x
Net2
2
[
1
2
]
with
1
0
0
%
ac
cu
r
ac
y
a
n
d
R
esNet
-
b
ased
a
p
p
r
o
ac
h
es
[
1
3
]
–
[
1
5
]
.
I
t
is
cr
u
c
ial
to
r
em
em
b
e
r
th
at
t
h
e
m
ajo
r
ity
o
f
t
h
is
r
esear
ch
co
n
ce
n
tr
ate
o
n
b
i
n
ar
y
class
if
icatio
n
task
s
th
at
s
ep
ar
ate
in
s
tan
ce
s
o
f
m
o
n
k
ey
p
o
x
f
r
o
m
th
o
s
e
th
at
d
o
n
o
t.
B
in
ar
y
class
if
icatio
n
p
r
o
b
lem
s
ar
e
g
en
er
ally
less
co
m
p
lex
th
an
m
u
lti
-
class
s
ce
n
ar
io
s
.
T
h
e
s
u
g
g
ested
ap
p
r
o
ac
h
,
o
n
th
e
o
th
er
h
an
d
,
tack
les
a
m
o
r
e
d
if
f
icu
lt
f
o
u
r
-
class
class
if
icatio
n
p
r
o
b
lem
in
v
o
lv
in
g
m
ea
s
les,
ch
ick
en
p
o
x
,
m
o
n
k
ey
p
o
x
,
a
n
d
n
o
r
m
al
s
k
in
b
ec
au
s
e
o
f
th
e
v
is
u
al
s
im
ilar
itie
s
b
etwe
en
s
k
in
d
is
o
r
d
er
s
.
T
h
e
s
u
g
g
ested
m
ajo
r
ity
v
o
tin
g
en
s
em
b
le
ac
h
iev
es
an
ac
cu
r
ac
y
o
f
9
8
.
9
4
%
d
esp
ite
th
is
ad
d
ed
co
m
p
lex
ity
,
wh
ich
is
o
n
p
ar
with
o
r
b
etter
th
an
a
n
u
m
b
er
o
f
cu
r
r
en
t
f
o
u
r
-
class
m
eth
o
d
s
lik
e
Mo
n
k
ey
Net
[
7
]
,
Mo
b
ileNetV2
[
9
]
,
a
n
d
o
th
er
C
NN
-
b
ased
m
o
d
els
[
1
6
]
,
[
1
7
]
.
C
o
m
p
ar
ed
to
p
r
ev
io
u
s
en
s
em
b
le
-
b
ased
s
tu
d
ies,
s
u
ch
as
th
e
wo
r
k
in
[
1
0
]
,
wh
ich
co
m
b
i
n
e
d
m
u
ltip
le
d
ee
p
lear
n
in
g
m
o
d
els
f
o
r
b
in
a
r
y
m
o
n
k
e
y
p
o
x
class
if
icatio
n
a
n
d
ac
h
ie
v
ed
a
n
ac
c
u
r
ac
y
o
f
9
3
.
3
9
%,
th
e
p
r
o
p
o
s
ed
m
eth
o
d
d
em
o
n
s
tr
ates
s
u
p
er
io
r
p
er
f
o
r
m
a
n
ce
wh
ile
h
an
d
lin
g
a
m
o
r
e
co
m
p
le
x
m
u
lti
-
class
class
if
icatio
n
task
.
T
h
e
m
ajo
r
ity
v
o
te
en
s
em
b
le
p
r
o
v
id
es
b
etter
r
o
b
u
s
tn
ess
b
y
c
o
m
b
in
in
g
p
r
ed
ictio
n
s
f
r
o
m
v
a
r
io
u
s
ar
ch
itectu
r
es,
ev
en
th
o
u
g
h
a
s
in
g
le
R
esNet5
0
m
o
d
el
h
as
th
e
m
ax
im
u
m
ac
cu
r
ac
y
.
T
h
is
en
s
em
b
le
ap
p
r
o
ac
h
im
p
r
o
v
es
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
P
erfo
r
ma
n
ce
o
f m
a
jo
r
ity
vo
tin
g
tr
a
n
s
fer lea
r
n
in
g
d
ee
p
le
a
r
n
in
g
mo
n
ke
y
p
o
x
d
is
ea
s
e
d
etec
ti
o
n
(
N
u
r
N
a
fiiya
h
)
3787
p
r
ed
ictio
n
s
tab
ilit
y
an
d
less
en
s
th
e
in
f
lu
en
ce
o
f
in
d
iv
id
u
al
m
o
d
el
er
r
o
r
s
,
wh
ich
is
cr
u
cial
in
m
ed
ical
im
ag
e
an
aly
s
is
wh
er
e
tr
u
s
two
r
th
y
d
e
cisi
o
n
ass
is
tan
ce
is
n
ee
d
ed
.
R
e
sN
e
t
5
0
V
G
G
1
6
M
o
b
i
l
e
N
e
t
I
n
c
e
p
t
i
o
n
V
3
X
c
e
p
t
i
o
n
Fig
u
r
e
5
.
His
to
r
y
tr
ai
n
in
g
a
n
d
v
alid
atio
n
tr
an
s
f
er
lear
n
in
g
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.
15
,
No
.
4
,
Au
g
u
s
t
20
26
:
3
7
8
2
-
3
7
9
1
3788
T
ab
le
4
.
E
v
alu
atio
n
c
o
m
p
ar
is
o
n
No
M
e
t
h
o
d
s
C
l
a
s
si
f
i
c
a
t
i
o
n
A
c
c
u
r
a
c
y
(
%)
1
M
o
n
k
e
y
N
e
t
i
s
a
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
s
e
d
C
N
N
mo
d
e
l
t
h
a
t
w
a
s m
o
d
i
f
i
e
d
f
r
o
m
D
e
n
seN
e
t
-
2
0
1
[
7
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
8
.
9
1
2
F
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
w
i
t
h
G
o
o
g
l
e
N
e
t
a
n
d
D
e
c
i
s
i
o
n
Tr
e
e
c
l
a
ssi
f
i
c
a
t
i
o
n
[
8
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
4
.
3
5
3
M
o
b
i
l
e
N
e
t
V
2
[
9
]
4
c
l
a
ss
(
mo
n
k
e
y
p
o
x
,
n
o
r
ma
l
,
s
c
a
r
l
e
t
f
e
v
e
r
,
r
o
s
e
o
l
a
)
9
8
.
1
6
4
En
se
mb
l
e
f
r
o
m X
c
e
p
t
i
o
n
,
I
n
c
e
p
t
i
o
n
V
3
,
D
e
n
s
e
N
e
t
1
6
9
[
1
0
]
2
c
l
a
ss
(
mo
n
k
e
y
p
o
x
,
n
o
n
m
o
n
k
e
y
p
o
x
)
9
3
.
3
9
5
R
e
sN
e
t
1
0
1
[
1
3
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
4
.
2
5
6
R
e
sN
e
t
1
0
1
[
1
4
]
2
c
l
a
ss
(
mo
n
k
e
y
p
o
x
,
n
o
r
ma
l
)
9
9
.
0
0
7
P
o
x
N
e
t
2
2
[
1
2
]
2
c
l
a
ss
(
mo
n
k
e
y
p
o
x
,
n
o
n
m
o
n
k
e
y
p
o
x
)
1
0
0
.
0
0
8
R
e
sN
e
t
1
8
[
1
5
]
2
c
l
a
ss
(
mo
n
k
e
y
p
o
x
,
n
o
n
m
o
n
k
e
y
p
o
x
)
9
9
.
4
9
9
C
N
N
-
LSTM
h
y
b
r
i
d
m
o
d
e
l
[
1
8
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
8
7
.
0
0
10
I
n
t
e
g
r
a
t
i
n
g
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s wi
t
h
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
(
F
L)
a
n
d
C
y
c
l
e
G
A
N
[
1
9
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
7
.
9
0
11
F
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
w
i
t
h
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
w
e
i
g
h
t
e
d
n
a
ï
v
e
B
a
y
e
s (W
N
B
)
,
w
e
i
g
h
t
e
d
k
-
n
e
a
r
e
s
t
n
e
i
g
h
b
o
r
s
(
W
K
N
N
)
c
l
a
ssi
f
i
c
a
t
i
o
n
[
1
6
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
8
.
4
8
12
M
o
b
i
l
e
N
e
t
V
3
-
s
[
2
0
]
2
c
l
a
ss
(
p
o
si
t
i
v
e
,
n
e
g
a
t
i
v
e
mo
n
k
e
y
p
o
x
)
9
9
.
0
0
13
D
e
n
seN
e
t
2
0
1
[
1
7
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
5
.
1
8
14
M
o
b
i
l
e
N
e
t
V
2
[
2
1
]
2
c
l
a
ss
(
mo
n
k
e
y
p
o
x
,
n
o
r
ma
l
)
9
9
.
0
0
15
D
e
n
seN
e
t
2
0
1
[
2
2
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
9
.
9
4
16
A
l
e
x
N
e
t
+
V
G
G
1
6
[
2
3
]
2
c
l
a
ss
(
mo
n
k
e
y
p
o
x
,
n
o
n
m
o
n
k
e
y
p
o
x
)
9
5
.
5
5
17
Ef
f
i
c
i
e
n
t
N
e
t
B
3
[
2
4
]
6
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
c
o
w
p
o
x
,
mea
s
l
e
s
,
mo
n
k
e
y
p
o
x
,
h
e
a
l
t
h
y
,
sm
a
l
l
p
o
x
)
8
7
.
0
0
18
M
o
b
i
l
e
N
e
t
V
2
[
2
5
]
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
1
.
3
7
19
P
r
o
p
o
se
d
m
e
t
h
o
d
4
c
l
a
ss
(
c
h
i
c
k
e
n
p
o
x
,
me
a
sl
e
s,
mo
n
k
e
y
p
o
x
,
n
o
r
m
a
l
)
9
8
.
9
4
T
h
e
co
n
f
u
s
io
n
m
atr
ices
f
o
r
th
e
f
iv
e
tr
an
s
f
er
lear
n
in
g
m
o
d
els:
R
esNe
t5
0
,
VGG1
6
,
Mo
b
ileNet,
I
n
ce
p
tio
n
V3
,
an
d
Xce
p
tio
n
a
r
e
s
h
o
wn
in
Fig
u
r
e
6
.
T
h
e
h
ig
h
v
al
u
es
alo
n
g
th
e
m
ajo
r
d
iag
o
n
al
o
f
ea
ch
co
n
f
u
s
io
n
m
atr
ix
s
h
o
w
th
at,
o
v
er
all,
all
m
o
d
els
p
er
f
o
r
m
well
in
class
if
icatio
n
ac
r
o
s
s
t
h
e
f
o
u
r
s
k
i
n
d
is
ea
s
e
class
es
(
ch
ick
en
p
o
x
,
m
ea
s
les,
m
o
n
k
e
y
p
o
x
,
an
d
n
o
r
m
al)
.
W
ith
1
1
8
im
a
g
es
o
f
ch
ic
k
en
p
o
x
,
1
0
9
p
h
o
to
g
r
ap
h
s
o
f
m
ea
s
les,
1
1
9
im
ag
es
o
f
m
o
n
k
ey
p
o
x
,
an
d
1
2
0
im
a
g
es
o
f
n
o
r
m
al,
th
e
R
esNet5
0
m
o
d
el
h
as
ex
ce
p
tio
n
al
class
if
icatio
n
s
k
ill.
Stro
n
g
d
is
cr
im
in
ativ
e
p
o
wer
an
d
lo
w
m
is
clas
s
if
icatio
n
r
ates
ac
r
o
s
s
all
clas
s
es
ar
e
in
d
icate
d
b
y
th
e
lar
g
e
co
n
ce
n
tr
atio
n
o
f
ac
cu
r
ate
p
r
ed
ictio
n
s
o
n
th
e
d
iag
o
n
al.
Similar
ly
,
VGG1
6
ac
h
iev
es
r
o
b
u
s
t
p
er
f
o
r
m
an
ce
,
with
1
1
7
co
r
r
ec
tly
class
if
ied
ch
ick
en
p
o
x
c
ases
,
1
0
7
m
ea
s
les
ca
s
es,
1
1
9
m
o
n
k
ey
p
o
x
ca
s
es,
an
d
1
1
7
n
o
r
m
al
ca
s
e
s
.
Alth
o
u
g
h
a
s
m
all
n
u
m
b
er
o
f
m
is
class
if
icatio
n
s
ar
e
o
b
s
er
v
e
d
,
th
e
o
v
e
r
all
p
at
ter
n
s
u
g
g
ests
co
n
s
is
ten
t
an
d
r
eliab
le
p
r
ed
ictio
n
s
.
T
h
e
M
o
b
ileNet
m
o
d
el
also
p
er
f
o
r
m
s
v
er
y
well,
co
r
r
ec
tly
class
if
y
in
g
1
1
8
ch
ick
en
p
o
x
,
1
0
8
m
ea
s
les,
1
1
9
m
o
n
k
ey
p
o
x
,
an
d
1
2
0
n
o
r
m
al
im
ag
es.
T
h
is
r
esu
lt
h
ig
h
lig
h
t
s
Mo
b
ileNet’
s
ef
f
ec
tiv
en
ess
d
esp
ite
its
lig
h
tweig
h
t
a
r
ch
it
ec
tu
r
e,
m
a
k
in
g
it
s
u
itab
le
f
o
r
ef
f
icien
t c
lass
if
icatio
n
task
s
.
Fo
r
I
n
ce
p
tio
n
V3
,
th
e
co
n
f
u
s
io
n
m
atr
i
x
in
d
icate
s
s
tr
o
n
g
p
e
r
f
o
r
m
an
ce
,
with
1
1
5
co
r
r
ec
t
ch
ick
en
p
o
x
p
r
ed
ictio
n
s
,
1
0
7
m
ea
s
les
p
r
ed
ictio
n
s
,
1
1
9
m
o
n
k
ey
p
o
x
p
r
e
d
ictio
n
s
,
an
d
1
2
0
n
o
r
m
al
p
r
e
d
ictio
n
s
.
T
h
e
r
elativ
ely
lo
w
n
u
m
b
er
o
f
m
is
class
if
ied
s
am
p
les
s
u
g
g
ests
s
tab
le
co
n
v
er
g
en
ce
a
n
d
ef
f
ec
tiv
e
f
ea
tu
r
e
ex
tr
ac
tio
n
.
1
1
3
p
h
o
to
g
r
a
p
h
s
o
f
ch
ick
en
p
o
x
,
1
1
0
im
ag
es
o
f
m
ea
s
les,
1
1
9
im
ag
es
o
f
m
o
n
k
e
y
p
o
x
,
an
d
1
1
9
im
ag
es
o
f
n
o
r
m
al
ar
e
co
r
r
ec
tly
clas
s
if
ied
b
y
th
e
Xce
p
tio
n
m
o
d
e
l.
Miscla
s
s
if
icatio
n
er
r
o
r
s
ar
e
m
in
im
al,
with
o
n
ly
a
f
ew
ch
ick
en
p
o
x
s
am
p
les
in
co
r
r
ec
tly
p
r
ed
icted
as
m
ea
s
les,
m
o
n
k
ey
p
o
x
,
o
r
n
o
r
m
al.
T
h
is
f
u
r
th
e
r
d
em
o
n
s
tr
ates th
e
r
o
b
u
s
tn
ess
o
f
th
e
m
o
d
el.
Fro
m
a
clin
ical
p
er
s
p
ec
tiv
e,
th
e
co
n
f
u
s
io
n
m
atr
ices
s
h
o
w
th
a
t
m
is
clas
s
if
icatio
n
b
etwe
en
m
o
n
k
ey
p
o
x
an
d
o
th
er
v
is
u
ally
s
im
ilar
d
is
ea
s
es
is
lim
ited
,
wh
ich
is
p
ar
ticu
lar
ly
im
p
o
r
ta
n
t
f
o
r
r
ed
u
cin
g
f
alse
-
p
o
s
itiv
e
an
d
f
alse
-
n
eg
ativ
e
p
r
ed
ictio
n
s
.
W
h
en
th
ese
in
d
iv
id
u
al
m
o
d
els
a
r
e
co
m
b
in
ed
u
s
in
g
th
e
p
r
o
p
o
s
ed
m
ajo
r
ity
v
o
tin
g
en
s
em
b
le,
p
r
ed
ictio
n
r
eliab
ilit
y
is
en
h
an
ce
d
b
y
ag
g
r
e
g
atin
g
m
o
d
el
d
ec
is
io
n
s
,
th
er
eb
y
r
ed
u
cin
g
th
e
in
f
lu
e
n
ce
o
f
is
o
lated
class
if
icatio
n
er
r
o
r
s
.
E
v
en
wh
en
th
e
en
s
em
b
l
e
ac
cu
r
ac
y
is
m
ar
g
in
ally
less
th
an
th
at
o
f
th
e
to
p
-
p
er
f
o
r
m
i
n
g
s
in
g
le
m
o
d
el,
th
is
en
s
em
b
le
b
eh
av
io
r
ex
p
l
ain
s
th
e
o
b
s
er
v
ed
im
p
r
o
v
em
e
n
t
in
p
r
ec
is
io
n
a
n
d
o
v
er
all
r
o
b
u
s
tn
ess
.
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
P
erfo
r
ma
n
ce
o
f m
a
jo
r
ity
vo
tin
g
tr
a
n
s
fer lea
r
n
in
g
d
ee
p
le
a
r
n
in
g
mo
n
ke
y
p
o
x
d
is
ea
s
e
d
etec
ti
o
n
(
N
u
r
N
a
fiiya
h
)
3789
Re
sN
e
t5
0
VG
G
1
6
M
o
b
i
leN
e
t
In
c
e
p
ti
o
n
V3
Xc
e
p
ti
o
n
Fig
u
r
e
6
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
ev
alu
atio
n
m
o
d
el
4.
CO
NCLU
SI
O
N
I
n
o
r
d
er
to
class
if
y
m
o
n
k
ey
p
o
x
an
d
o
th
er
s
k
in
d
is
o
r
d
er
s
,
th
is
wo
r
k
s
h
o
ws
th
e
ef
f
icac
y
o
f
a
m
ajo
r
ity
v
o
tin
g
en
s
em
b
le
th
at
in
teg
r
at
es
f
iv
e
tr
an
s
f
er
lear
n
in
g
m
o
d
els:
R
esNet5
0
,
VGG1
6
,
Mo
b
ileNet,
I
n
ce
p
tio
n
V3
,
an
d
Xce
p
tio
n
.
Acc
o
r
d
in
g
to
e
x
p
er
im
en
tal
d
ata,
th
e
s
u
g
g
este
d
m
ajo
r
ity
v
o
tin
g
s
tr
ateg
y
o
b
t
ain
ed
a
co
m
p
ar
a
b
le
ac
cu
r
ac
y
o
f
9
8
.
9
4
%
an
d
p
r
o
v
id
ed
h
ig
h
er
p
r
ec
is
io
n
th
an
v
ar
i
o
u
s
in
d
iv
i
d
u
al
m
o
d
els,
wh
e
r
ea
s
R
esNet5
0
ea
r
n
ed
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.
15
,
No
.
4
,
Au
g
u
s
t
20
26
:
3
7
8
2
-
3
7
9
1
3790
th
e
b
est
in
d
iv
id
u
al
ac
cu
r
ac
y
(
9
9
.
1
5
%).
Ma
jo
r
ity
v
o
tin
g
im
p
r
o
v
es
p
r
e
d
ictio
n
r
eliab
ilit
y
b
y
co
m
b
in
in
g
o
u
tp
u
ts
f
r
o
m
s
ev
er
al
ar
ch
itectu
r
es,
s
o
less
en
in
g
th
e
im
p
ac
t
o
f
in
d
iv
id
u
al
m
o
d
el
f
au
lts
,
ev
en
wh
e
n
th
e
en
s
em
b
le
d
id
n
o
t
o
u
tp
er
f
o
r
m
th
e
to
p
-
p
er
f
o
r
m
in
g
s
in
g
le
m
o
d
el
in
ter
m
s
o
f
ac
cu
r
ac
y
.
W
h
en
co
m
b
i
n
ed
with
C
L
AHE
-
b
ased
im
ag
e
p
r
ep
r
o
ce
s
s
in
g
,
th
is
ap
p
r
o
ac
h
co
n
tr
i
b
u
tes
to
m
o
r
e
r
o
b
u
s
t
class
if
icatio
n
p
er
f
o
r
m
an
ce
,
wh
ich
i
s
p
ar
ticu
lar
ly
r
elev
a
n
t
f
o
r
m
ed
i
ca
l
im
ag
e
an
aly
s
is
task
s
in
v
o
lv
in
g
v
is
u
ally
s
im
ilar
s
k
in
co
n
d
itio
n
s
.
I
n
o
r
d
er
to
in
cr
ea
s
e
g
en
er
aliza
b
ilit
y
,
f
u
t
u
r
e
wo
r
k
will
c
o
n
ce
n
tr
ate
o
n
ad
d
in
g
m
o
r
e
v
ar
ie
d
r
ea
l
-
w
o
r
ld
im
a
g
es
to
t
h
e
d
ataset,
in
v
esti
g
atin
g
s
o
p
h
is
ticated
en
s
em
b
le
tec
h
n
iq
u
es
li
k
e
weig
h
ted
v
o
tin
g
a
n
d
s
tac
k
in
g
,
a
n
d
ass
ess
in
g
ex
tr
a
p
er
f
o
r
m
an
ce
m
etr
ics
in
clu
d
in
g
r
ec
all,
F1
-
s
co
r
e,
an
d
co
n
f
u
s
io
n
m
atr
ices.
Fu
r
th
er
m
o
r
e,
in
teg
r
atin
g
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
in
to
we
b
-
o
r
m
o
b
ile
-
b
ased
co
m
p
u
ter
-
aid
ed
s
cr
ee
n
i
n
g
s
y
s
tem
s
m
ay
s
u
p
p
o
r
t
ea
r
ly
a
n
d
r
eliab
le
m
o
n
k
e
y
p
o
x
d
etec
tio
n
in
p
r
ac
tical
h
ea
lth
ca
r
e
s
ettin
g
s
.
ACK
NO
WL
E
DG
M
E
N
T
S
W
e
wo
u
ld
lik
e
to
t
h
an
k
s
f
o
r
D
R
PM
Kem
d
ik
tis
ain
tek
an
d
th
e
Un
iv
er
s
itas
I
s
lam
L
am
o
n
g
an
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
was f
u
n
d
e
d
f
r
o
m
Un
iv
er
s
itas
I
s
lam
L
am
o
n
g
a
n
.
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
t
r
ib
u
to
r
R
o
les
T
a
x
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
i
d
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
Nu
r
Naf
iiy
ah
✓
✓
✓
✓
✓
✓
✓
✓
✓
M
u
h
a
m
m
a
d
N
u
r
u
l
H
u
d
a
✓
✓
✓
✓
✓
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
All r
esear
ch
er
s
d
o
n
o
t h
av
e
a
co
n
f
lict.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
av
ailab
ilit
y
f
r
o
m
a
n
o
p
e
n
d
ataset
th
at
h
as
b
ee
n
u
s
ed
p
r
ev
i
o
u
s
r
esear
ch
in
Kag
g
le
at
h
ttp
s
://www.
k
ag
g
le.
co
m
/d
atasets
/d
ip
u
iu
cse/m
o
n
k
ey
p
o
x
s
k
in
i
m
ag
ed
ataset
.
RE
F
E
R
E
NC
E
S
[
1
]
C
.
Zh
a
o
,
R
.
S
h
u
a
i
,
L.
M
a
,
W
.
Li
u
,
D
.
H
u
,
a
n
d
M
.
W
u
,
“
D
e
r
m
o
sc
o
p
y
i
ma
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
b
a
se
d
o
n
S
t
y
l
e
G
A
N
a
n
d
D
e
n
s
e
N
e
t
2
0
1
,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
9
,
p
p
.
8
6
5
9
–
8
6
7
9
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
1
.
3
0
4
9
6
0
0
.
[
2
]
S
.
I
n
t
h
i
y
a
z
e
t
a
l
.
,
“
S
k
i
n
d
i
s
e
a
se
d
e
t
e
c
t
i
o
n
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
A
d
v
a
n
c
e
s
i
n
E
n
g
i
n
e
e
ri
n
g
S
o
f
t
w
a
r
e
,
v
o
l
.
1
7
5
,
Ja
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
d
v
e
n
g
so
f
t
.
2
0
2
2
.
1
0
3
3
6
1
.
[
3
]
P
.
C
h
o
u
d
h
a
r
y
,
J.
S
i
n
g
h
a
i
,
a
n
d
J
.
S
.
Y
a
d
a
v
,
“
S
k
i
n
l
e
s
i
o
n
d
e
t
e
c
t
i
o
n
b
a
s
e
d
o
n
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s,
”
C
h
e
m
o
m
e
t
ri
c
s
a
n
d
I
n
t
e
l
l
i
g
e
n
t
L
a
b
o
r
a
t
o
ry
S
y
st
e
m
s
,
v
o
l
.
2
3
0
,
N
o
v
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
h
e
m
o
l
a
b
.
2
0
2
2
.
1
0
4
6
5
9
.
[
4
]
M
.
S
.
A
l
i
,
M
.
S
.
M
i
a
h
,
J
.
H
a
q
u
e
,
M
.
M
.
R
a
h
ma
n
,
a
n
d
M
.
K
.
I
sl
a
m,
“
A
n
e
n
h
a
n
c
e
d
t
e
c
h
n
i
q
u
e
o
f
s
k
i
n
c
a
n
c
e
r
c
l
a
ssi
f
i
c
a
t
i
o
n
u
s
i
n
g
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
w
i
t
h
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
m
o
d
e
l
s
,
”
Ma
c
h
i
n
e
L
e
a
r
n
i
n
g
w
i
t
h
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
5
,
S
e
p
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
m
l
w
a
.
2
0
2
1
.
1
0
0
0
3
6
.
[
5
]
S
.
M
.
Ja
i
sa
k
t
h
i
,
P
.
M
i
r
u
n
a
l
i
n
i
,
C
.
A
r
a
v
i
n
d
a
n
,
a
n
d
R
.
A
p
p
a
v
u
,
“
C
l
a
ss
i
f
i
c
a
t
i
o
n
o
f
sk
i
n
c
a
n
c
e
r
f
r
o
m
d
e
r
mo
s
c
o
p
i
c
i
mag
e
s
u
si
n
g
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
a
r
c
h
i
t
e
c
t
u
r
e
s,
”
M
u
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
8
2
,
n
o
.
1
0
,
p
p
.
1
5
7
6
3
–
1
5
7
7
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
0
4
2
-
022
-
1
3
8
4
7
-
3.
[
6
]
F
.
Y
a
s
m
i
n
,
M
.
M
.
H
a
s
s
a
n
,
S
.
Z
a
m
a
n
,
S
.
T
.
A
u
n
g
,
A
.
K
a
r
i
m
,
a
n
d
S
.
A
z
a
m
,
“
A
f
o
r
e
c
a
s
t
i
n
g
p
r
o
g
n
o
s
i
s
o
f
t
h
e
m
o
n
k
e
y
p
o
x
o
u
t
b
r
e
a
k
b
a
s
e
d
o
n
a
c
o
m
p
r
e
h
e
n
s
i
v
e
s
t
a
t
i
s
t
i
c
a
l
a
n
d
r
e
g
r
e
s
s
i
o
n
a
n
a
l
y
s
i
s
,
”
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
1
0
,
n
o
.
1
0
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
c
o
m
p
u
t
a
t
i
o
n
1
0
1
0
0
1
7
7
.
[
7
]
D
.
B
a
l
a
e
t
a
l
.
,
“
M
o
n
k
e
y
N
e
t
:
a
r
o
b
u
s
t
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
f
o
r
mo
n
k
e
y
p
o
x
d
i
sea
se
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
N
e
u
ra
l
N
e
t
w
o
rks
,
v
o
l
.
1
6
1
,
p
p
.
7
5
7
–
7
7
5
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
n
e
t
.
2
0
2
3
.
0
2
.
0
2
2
.
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
P
erfo
r
ma
n
ce
o
f m
a
jo
r
ity
vo
tin
g
tr
a
n
s
fer lea
r
n
in
g
d
ee
p
le
a
r
n
in
g
mo
n
ke
y
p
o
x
d
is
ea
s
e
d
etec
ti
o
n
(
N
u
r
N
a
fiiya
h
)
3791
[
8
]
A
.
H
.
A
l
h
a
r
b
i
e
t
a
l
.
,
“
D
i
a
g
n
o
s
i
s
o
f
m
o
n
k
e
y
p
o
x
d
i
se
a
se
u
s
i
n
g
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
a
n
d
b
i
n
a
r
y
a
d
v
a
n
c
e
d
d
i
p
p
e
r
t
h
r
o
a
t
e
d
o
p
t
i
m
i
z
a
t
i
o
n
a
l
g
o
r
i
t
h
m,
”
B
i
o
m
i
m
e
t
i
c
s
,
v
o
l
.
8
,
n
o
.
3
,
Ju
l
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
b
i
o
mi
m
e
t
i
c
s
8
0
3
0
3
1
3
.
[
9
]
A
.
S
.
Jara
d
a
t
e
t
a
l
.
,
“
A
u
t
o
ma
t
e
d
mo
n
k
e
y
p
o
x
s
k
i
n
l
e
s
i
o
n
d
e
t
e
c
t
i
o
n
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
E
n
v
i
ro
n
m
e
n
t
a
l
Re
s
e
a
r
c
h
a
n
d
P
u
b
l
i
c
H
e
a
l
t
h
,
v
o
l
.
2
0
,
n
o
.
5
,
M
a
r
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
i
j
e
r
p
h
2
0
0
5
4
4
2
2
.
[
1
0
]
R
.
P
r
a
ma
n
i
k
,
B
.
B
a
n
e
r
j
e
e
,
G
.
Ef
i
m
e
n
k
o
,
D
.
K
a
p
l
u
n
,
a
n
d
R
.
S
a
r
k
a
r
,
“
M
o
n
k
e
y
p
o
x
d
e
t
e
c
t
i
o
n
f
r
o
m
sk
i
n
l
e
si
o
n
i
ma
g
e
s
u
si
n
g
a
n
a
mal
g
a
ma
t
i
o
n
o
f
C
N
N
mo
d
e
l
s
a
i
d
e
d
w
i
t
h
B
e
t
a
f
u
n
c
t
i
o
n
-
b
a
se
d
n
o
r
m
a
l
i
z
a
t
i
o
n
sc
h
e
me,
”
PLO
S
O
N
E
,
v
o
l
.
1
8
,
n
o
.
4
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
3
7
1
/
j
o
u
r
n
a
l
.
p
o
n
e
.
0
2
8
1
8
1
5
.
[
1
1
]
N
.
N
a
f
i
i
y
a
h
,
N
.
S
.
F
a
t
o
n
a
h
,
R
.
W
a
r
d
h
a
n
i
,
B
.
Jo
k
o
n
o
w
o
,
a
n
d
T
.
A
.
Y
.
S
i
s
w
a
,
“
M
a
j
o
r
i
t
y
v
o
t
i
n
g
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
C
N
N
f
o
r
p
e
a
n
u
t
l
e
a
f
t
y
p
e
s
i
d
e
n
t
i
f
i
c
a
t
i
o
n
,
”
i
n
2
0
2
2
3
r
d
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
E
l
e
c
t
ri
c
a
l
E
n
g
i
n
e
e
r
i
n
g
a
n
d
I
n
f
o
rm
a
t
i
c
s
(
I
C
o
n
EEI)
,
O
c
t
.
2
0
2
2
,
p
p
.
7
0
–
7
4
.
d
o
i
:
1
0
.
1
1
0
9
/
I
C
o
n
E
EI
5
5
7
0
9
.
2
0
2
2
.
9
9
7
2
2
7
1
.
[
1
2
]
F
.
Y
a
smi
n
e
t
a
l
.
,
“
P
o
x
N
e
t
2
2
:
a
f
i
n
e
-
t
u
n
e
d
m
o
d
e
l
f
o
r
t
h
e
c
l
a
ssi
f
i
c
a
t
i
o
n
o
f
mo
n
k
e
y
p
o
x
d
i
s
e
a
s
e
u
si
n
g
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
,
”
I
E
E
E
Ac
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
2
4
0
5
3
–
2
4
0
7
6
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
2
5
3
8
6
8
.
[
1
3
]
M
.
L
a
k
s
h
m
i
a
n
d
R
.
D
a
s
,
“
C
l
a
ssi
f
i
c
a
t
i
o
n
o
f
m
o
n
k
e
y
p
o
x
i
ma
g
e
s
u
s
i
n
g
LI
M
E
-
e
n
a
b
l
e
d
i
n
v
e
st
i
g
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
,
”
D
i
a
g
n
o
st
i
c
s
,
v
o
l
.
1
3
,
n
o
.
9
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
s
t
i
c
s
1
3
0
9
1
6
3
9
.
[
1
4
]
M
.
M
.
A
h
s
a
n
e
t
a
l
.
,
“
D
e
e
p
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s
f
o
r
m
o
n
k
e
y
p
o
x
d
i
s
e
a
se
d
i
a
g
n
o
s
i
s,
”
E
x
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
1
6
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
s
w
a
.
2
0
2
2
.
1
1
9
4
8
3
.
[
1
5
]
T.
N
a
y
a
k
e
t
a
l
.
,
“
D
e
e
p
l
e
a
r
n
i
n
g
b
a
s
e
d
d
e
t
e
c
t
i
o
n
o
f
mo
n
k
e
y
p
o
x
v
i
r
u
s
u
s
i
n
g
sk
i
n
l
e
s
i
o
n
i
ma
g
e
s
,
”
M
e
d
i
c
i
n
e
i
n
N
o
v
e
l
T
e
c
h
n
o
l
o
g
y
a
n
d
D
e
v
i
c
e
s
,
v
o
l
.
1
8
,
Ju
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
me
d
n
t
d
.
2
0
2
3
.
1
0
0
2
4
3
.
[
1
6
]
A
.
I
.
S
a
l
e
h
a
n
d
A
.
H
.
R
a
b
i
e
,
“
H
u
m
a
n
mo
n
k
e
y
p
o
x
d
i
a
g
n
o
se
(
H
M
D
)
s
t
r
a
t
e
g
y
b
a
se
d
o
n
d
a
t
a
mi
n
i
n
g
a
n
d
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
t
e
c
h
n
i
q
u
e
s,
”
C
o
m
p
u
t
e
rs
i
n
B
i
o
l
o
g
y
a
n
d
M
e
d
i
c
i
n
e
,
v
o
l
.
1
5
2
,
Ja
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
b
i
o
m
e
d
.
2
0
2
2
.
1
0
6
3
8
3
.
[
1
7
]
A
.
S
.
A
z
a
r
,
A
.
N
a
e
mi
,
S
.
B
.
R
i
k
a
n
,
J.
B
.
M
o
h
a
s
e
f
i
,
H
.
P
i
r
n
e
j
a
d
,
a
n
d
U
.
K
.
W
i
i
l
,
“
M
o
n
k
e
y
p
o
x
d
e
t
e
c
t
i
o
n
u
s
i
n
g
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s,”
BM
C
I
n
f
e
c
t
i
o
u
s
D
i
se
a
s
e
s
,
v
o
l
.
2
3
,
n
o
.
1
,
Ju
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
8
6
/
s1
2
8
7
9
-
023
-
0
8
4
0
8
-
4.
[
1
8
]
F
.
U
y
sa
l
,
“
D
e
t
e
c
t
i
o
n
o
f
m
o
n
k
e
y
p
o
x
d
i
s
e
a
se
f
r
o
m
h
u
m
a
n
s
k
i
n
i
ma
g
e
s
w
i
t
h
a
h
y
b
r
i
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
,
”
D
i
a
g
n
o
st
i
c
s
,
v
o
l
.
1
3
,
n
o
.
1
0
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
st
i
c
s
1
3
1
0
1
7
7
2
.
[
1
9
]
D
.
K
u
n
d
u
e
t
a
l
.
,
“
F
e
d
e
r
a
t
e
d
d
e
e
p
l
e
a
r
n
i
n
g
f
o
r
m
o
n
k
e
y
p
o
x
d
i
sea
se
d
e
t
e
c
t
i
o
n
o
n
G
A
N
-
a
u
g
me
n
t
e
d
d
a
t
a
s
e
t
,
”
I
E
EE
Ac
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
3
2
8
1
9
–
3
2
8
2
9
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
4
.
3
3
7
0
8
3
8
.
[
2
0
]
M
.
A
l
t
u
n
,
H
.
G
ü
r
ü
l
e
r
,
O
.
Ö
z
k
a
r
a
c
a
,
F
.
K
h
a
n
,
J
.
K
h
a
n
,
a
n
d
Y
.
L
e
e
,
“
M
o
n
k
e
y
p
o
x
d
e
t
e
c
t
i
o
n
u
s
i
n
g
C
N
N
w
i
t
h
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
,
”
S
e
n
so
rs
,
v
o
l
.
2
3
,
n
o
.
4
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
s2
3
0
4
1
7
8
3
.
[
2
1
]
M
.
M
.
A
h
sa
n
e
t
a
l
.
,
“
M
o
n
k
e
y
p
o
x
d
i
a
g
n
o
s
i
s
w
i
t
h
i
n
t
e
r
p
r
e
t
a
b
l
e
d
e
e
p
l
e
a
r
n
i
n
g
,
”
I
EE
E
A
c
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
8
1
9
6
5
–
8
1
9
8
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
3
0
0
7
9
3
.
[
2
2
]
G
.
R
e
n
,
“
M
o
n
k
e
y
p
o
x
d
i
s
e
a
s
e
d
e
t
e
c
t
i
o
n
w
i
t
h
p
r
e
t
r
a
i
n
e
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
s,”
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
C
o
n
t
r
o
l
,
v
o
l
.
5
2
,
n
o
.
2
,
p
p
.
2
8
8
–
2
9
6
,
Ju
l
.
2
0
2
3
,
d
o
i
:
1
0
.
5
7
5
5
/
j
0
1
.
i
t
c
.
5
2
.
2
.
3
2
8
0
3
.
[
2
3
]
A
.
K
.
G
a
i
r
o
l
a
a
n
d
V
.
K
u
mar,
“
M
o
n
k
e
y
p
o
x
d
i
se
a
se
d
i
a
g
n
o
si
s
u
s
i
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
i
n
2
0
2
2
8
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
i
g
n
a
l
Pr
o
c
e
s
si
n
g
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
(
I
C
S
C
)
,
D
e
c
.
2
0
2
2
,
p
p
.
4
2
3
–
42
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
C
5
6
5
2
4
.
2
0
2
2
.
1
0
0
0
9
1
3
5
.
[
2
4
]
M
.
D
w
i
v
e
d
i
,
R
.
G
.
T
i
w
a
r
i
,
a
n
d
N
.
U
j
j
w
a
l
,
“
D
e
e
p
l
e
a
r
n
i
n
g
m
e
t
h
o
d
s
f
o
r
e
a
r
l
y
d
e
t
e
c
t
i
o
n
o
f
m
o
n
k
e
y
p
o
x
s
k
i
n
l
e
s
i
o
n
,
”
i
n
2
0
2
2
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
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
(
I
C
S
C
)
,
D
e
c
.
2
0
2
2
,
p
p
.
3
4
3
–
3
4
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
C
5
6
5
2
4
.
2
0
2
2
.
1
0
0
0
9
5
7
1
.
[
2
5
]
M
.
C
.
I
r
mak
,
T.
A
y
d
i
n
,
a
n
d
M
.
Y
a
g
a
n
o
g
l
u
,
“
M
o
n
k
e
y
p
o
x
sk
i
n
l
e
si
o
n
d
e
t
e
c
t
i
o
n
w
i
t
h
M
o
b
i
l
e
N
e
t
V
2
a
n
d
V
G
G
N
e
t
mo
d
e
l
s
,
”
i
n
2
0
2
2
Me
d
i
c
a
l
T
e
c
h
n
o
l
o
g
i
e
s
C
o
n
g
r
e
ss (T
I
P
T
EK
N
O
)
,
O
c
t
.
2
0
2
2
,
p
p
.
1
–
4
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
P
TE
K
N
O
5
6
5
6
8
.
2
0
2
2
.
9
9
6
0
1
9
4
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Nur
Na
fii
y
a
h
re
c
e
iv
e
d
h
e
r
B
a
c
h
e
lo
r
o
f
I
n
fo
rm
a
ti
c
s
En
g
in
e
e
rin
g
fro
m
Un
iv
e
rsitas
Isla
m
Lam
o
n
g
a
n
,
I
n
d
o
n
e
sia
(
2
0
0
5
-
2
0
0
9
)
a
n
d
M
a
ste
r
o
f
In
f
o
rm
a
ti
o
n
Tec
h
n
o
lo
g
y
fr
o
m
t
h
e
S
e
k
o
lah
Ti
n
g
g
i
Tek
n
ik
S
u
ra
b
a
y
a
,
In
d
o
n
e
sia
in
(2
0
1
1
-
2
0
1
3
)
.
S
h
e
is
th
e
P
h
.
D.
d
e
g
re
e
in
Co
m
p
u
ter
S
c
ien
c
e
with
De
p
a
rtme
n
t
o
f
I
n
fo
rm
a
ti
c
s
E
n
g
i
n
e
e
rin
g
,
F
a
c
u
l
ty
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
l
o
g
y
,
I
n
stit
u
t
Tek
n
o
l
o
g
i
S
e
p
u
l
u
h
N
o
p
e
m
b
e
r
2
0
2
3
.
S
h
e
is
c
u
rre
n
tl
y
in
tere
ste
d
i
n
AI,
d
e
e
p
lea
rn
in
g
,
a
n
d
c
o
m
p
u
ter
v
isi
o
n
.
S
h
e
is
c
u
rre
n
tl
y
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
a
t
t
h
e
De
p
a
rtme
n
t
o
f
In
f
o
rm
a
ti
c
s
a
t
Un
iv
e
rsitas
Isla
m
Lam
o
n
g
a
n
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
m
y
n
a
ff@u
n
isla.a
c
.
i
d
.
Mu
h
a
m
m
a
d
Nurul
H
u
d
a
re
c
e
iv
e
d
h
is
b
a
c
h
e
lo
r’s
d
e
g
re
e
fro
m
De
p
a
rtme
n
t
o
f
In
fo
rm
a
ti
c
s
E
n
g
i
n
e
e
rin
g
,
F
a
c
u
lt
y
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
lo
g
y
,
Un
iv
e
rsitas
Isla
m
Lam
o
n
g
a
n
,
In
d
o
n
e
sia
(2
0
2
1
-
2
0
2
5
).
His
c
u
rre
n
t
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
d
e
e
p
lea
rn
in
g
,
c
o
m
p
u
ter
v
isi
o
n
,
a
n
d
AI ap
p
li
c
a
ti
o
n
s.
He
c
a
n
b
e
c
o
n
tac
ted
a
t:
m
u
h
a
m
m
a
d
n
u
ru
lh
u
d
a
7
5
3
@
g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.