I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
ar
y
20
2
5
,
p
p
.
1
11
2
~
1
1
20
I
SS
N:
2
502
-
4
7
52
,
DOI
: 1
0
.
1
1
5
9
1
/ijee
cs
.v
3
7
.
i
2
.
pp
1
11
2
-
1
1
20
1112
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
Efficien
t
de
ep lea
rning
appro
a
ch f
o
r enhancing
pla
nt
lea
f
disea
se cla
ss
ificat
io
n
M
er
o
ua
B
elm
ir
1
,
Wa
f
a
Dif
a
l
la
h
2
,
Abdelk
a
der
G
ha
zli
1
1
La
b
o
r
a
t
o
r
y
o
f
C
o
d
i
n
g
a
n
d
I
n
f
o
r
ma
t
i
o
n
S
e
c
u
r
i
t
y
(
LA
C
O
S
I
)
,
F
a
c
u
l
t
y
o
f
E
x
a
c
t
S
c
i
e
n
c
e
s,
Ta
h
r
i
M
o
h
a
mm
e
d
U
n
i
v
e
r
si
t
y
o
f
B
e
c
h
a
r
,
B
e
c
h
a
r
,
A
l
g
e
r
i
a
2
La
b
o
r
a
t
o
r
y
o
f
E
n
e
r
g
e
t
i
c
i
n
A
r
i
d
Z
o
n
e
s (EN
ER
G
A
R
I
D
)
,
F
a
c
u
l
t
y
o
f
E
x
a
c
t
S
c
i
e
n
c
e
s,
Ta
h
r
i
M
o
h
a
mm
e
d
U
n
i
v
e
r
si
t
y
o
f
B
e
c
h
a
r
,
B
e
c
h
a
r
,
A
l
g
e
r
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
15
,
2
0
2
4
R
ev
is
ed
Sep
20
,
2
0
2
4
Acc
ep
ted
Sep
30
,
2
0
2
4
Th
e
wid
e
sp
re
a
d
o
c
c
u
rre
n
c
e
o
f
p
lan
t
d
ise
a
se
s
is
a
m
a
jo
r
fa
c
to
r
i
n
th
e
re
d
u
c
ti
o
n
o
f
a
g
ricu
lt
u
ra
l
o
u
tp
u
t,
a
ffe
c
ti
n
g
b
o
th
c
ro
p
q
u
a
li
t
y
a
n
d
q
u
a
n
ti
t
y
.
Th
e
se
d
ise
a
se
s
ty
p
ica
ll
y
b
e
g
i
n
o
n
th
e
lea
v
e
s,
in
fl
u
e
n
c
e
d
b
y
a
lt
e
ra
ti
o
n
s
i
n
p
lan
t
stru
c
t
u
re
a
n
d
g
r
o
win
g
tec
h
n
iq
u
e
s,
a
n
d
c
a
n
e
v
e
n
t
u
a
ll
y
sp
re
a
d
o
v
e
r
t
h
e
e
n
ti
re
p
lan
t.
Th
is
re
su
lt
s
in
a
n
o
tab
le
d
e
c
re
a
se
in
c
ro
p
v
a
riety
a
n
d
y
ield
.
S
u
c
c
e
ss
fu
ll
y
m
a
n
a
g
in
g
t
h
e
se
d
ise
a
se
s
d
e
p
e
n
d
s
o
n
a
c
c
u
ra
tely
c
las
si
fy
in
g
a
n
d
d
e
tec
ti
n
g
lea
f
i
n
fe
c
ti
o
n
s
e
a
rly
,
w
h
ich
is
e
ss
e
n
ti
a
l
f
o
r
c
o
n
t
ro
ll
i
n
g
t
h
e
ir
sp
re
a
d
a
n
d
e
n
s
u
rin
g
h
e
a
lt
h
y
p
lan
t
g
ro
wt
h
.
T
o
a
d
d
re
ss
th
e
se
c
h
a
ll
e
n
g
e
s,
t
h
is
p
a
p
e
r
in
tro
d
u
c
e
s
a
n
e
fficie
n
t
a
p
p
r
o
a
c
h
fo
r
d
e
tec
ti
n
g
p
la
n
t
lea
f
d
i
se
a
se
s.
A
c
o
n
c
a
ten
a
ti
o
n
o
f
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
r
k
s
(
CNN
)
fo
r
e
n
h
a
n
c
e
d
p
lan
t
lea
f
d
ise
a
se
u
sin
g
tran
sfe
r
lea
rn
in
g
tec
h
n
i
q
u
e
is
imp
lem
e
n
te
d
,
wit
h
a
sp
e
c
ifi
c
f
o
c
u
s
o
n
a
c
c
u
ra
te
e
a
rly
d
e
tec
ti
o
n
,
u
t
il
izin
g
t
h
e
c
o
m
p
re
h
e
n
siv
e
n
e
w
p
lan
t
d
ise
a
se
s
d
a
tas
e
t.
T
h
e
c
o
m
b
in
e
d
re
sid
u
a
l
n
e
two
r
k
-
5
0
(
R
es
N
et
-
50
)
wit
h
d
e
n
se
ly
c
o
n
n
e
c
ted
c
o
n
v
o
lu
ti
o
n
a
l
n
e
two
rk
-
1
2
1
(
D
e
n
se
N
et
-
121
)
a
rc
h
it
e
c
t
u
re
a
i
m
s
to
p
r
o
v
id
e
a
n
e
fficie
n
t
a
n
d
re
li
a
b
le
so
lu
ti
o
n
to
th
e
se
c
rit
ica
l
a
g
r
icu
lt
u
ra
l
c
o
n
c
e
rn
s.
Va
rio
u
s
e
v
a
lu
a
ti
o
n
m
e
tri
c
s
we
re
u
ti
li
z
e
d
t
o
e
v
a
l
u
a
te
th
e
r
o
b
u
stn
e
ss
o
f
th
e
p
ro
p
o
se
d
h
y
b
ri
d
m
o
d
e
l.
T
h
e
p
ro
p
o
se
d
R
es
N
et
-
5
0
wit
h
th
e
D
en
se
N
et
-
1
2
1
h
y
b
ri
d
m
o
d
e
l
a
c
h
iev
e
d
a
ra
te
o
f
a
c
c
u
ra
c
y
o
f
9
9
.
6
6
%
.
K
ey
w
o
r
d
s
:
Dee
p
lear
n
in
g
Den
s
eNe
t
-
121
Plan
t le
af
d
is
ea
s
e
R
esNet
-
50
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
:
Me
r
o
u
a
B
elm
ir
L
ab
o
r
ato
r
y
o
f
C
o
d
in
g
a
n
d
I
n
f
o
r
m
atio
n
Secu
r
ity
(
L
AC
OSI
)
,
Facu
lty
o
f
E
x
ac
t Scie
n
ce
s
T
ah
r
i M
o
h
am
m
ed
Un
iv
e
r
s
ity
o
f
B
ec
h
ar
K
en
ad
s
a
Stre
et,
B
ec
h
ar
0
8
0
0
0
,
Alg
er
ia
E
m
ail:
b
elm
ir
.
m
er
o
u
a@
u
n
iv
-
b
ec
h
ar
.
d
z
1.
I
NT
RO
D
UCT
I
O
N
Ma
s
s
iv
e
o
b
s
tacle
s
co
n
f
r
o
n
tin
g
th
e
wo
r
ld
to
d
a
y
in
clu
d
e
th
e
wo
r
ld
’
s
r
ap
id
ly
ex
p
an
d
in
g
p
o
p
u
latio
n
,
clim
ate
v
ar
iab
ilit
y
,
f
o
o
d
s
ec
u
r
ity
,
ec
o
lo
g
ical
d
ec
lin
e,
an
d
in
f
ec
tio
u
s
d
is
ea
s
e
p
an
d
em
ics.
T
en
b
illi
o
n
p
eo
p
le
ar
e
ex
p
ec
ted
to
liv
e
o
n
th
e
p
la
n
e
t
b
y
th
e
m
id
-
21
st
ce
n
tu
r
y
.
W
o
r
ld
wid
e
f
o
o
d
s
ec
u
r
ity
is
ex
p
ec
ted
to
b
ec
o
m
e
a
m
ajo
r
co
n
ce
r
n
s
in
ce
it is
p
r
ed
i
cted
th
at
f
o
o
d
d
em
a
n
d
will in
c
r
ea
s
e
b
y
7
0
% to
1
0
0
% b
y
2
0
5
0
[
1
]
.
E
n
h
an
cin
g
th
e
a
g
r
icu
ltu
r
al
s
ec
to
r
is
v
ital
f
o
r
o
p
tim
izin
g
p
r
o
d
u
ctio
n
an
d
elev
atin
g
q
u
alit
y
.
C
r
ea
tin
g
f
av
o
r
a
b
le
co
n
d
itio
n
s
f
o
r
th
e
h
ea
lth
y
d
e
v
elo
p
m
e
n
t
o
f
p
lan
ts
an
d
c
r
o
p
s
is
ess
en
tial.
Plan
t
d
i
s
ea
s
es
ar
e
o
f
ten
th
e
m
ain
f
ac
to
r
s
co
n
tr
ib
u
tin
g
to
t
h
eir
d
ec
lin
e.
B
ased
o
n
esti
m
ates
f
r
o
m
th
e
Fo
o
d
a
n
d
A
g
r
icu
ltu
r
e
Or
g
an
izatio
n
o
f
th
e
Un
ited
Natio
n
s
(
FAO)
,
th
ese
d
is
ea
s
es
r
esu
lt
in
an
a
n
n
u
al
ec
o
n
o
m
ic
im
p
ac
t
o
f
a
r
o
u
n
d
$
2
2
0
b
illi
o
n
g
lo
b
ally
,
f
r
e
q
u
en
tly
ca
u
s
in
g
d
am
ag
e
o
r
to
t
al
cr
o
p
lo
s
s
.
Pla
n
ts
ar
e
attac
k
ed
b
y
v
ar
io
u
s
p
ath
o
g
en
s
,
in
clu
d
in
g
b
ac
ter
ia,
f
u
n
g
i,
m
icr
o
s
co
p
ic
o
r
g
an
is
m
s
,
an
d
v
ir
u
s
es,
wh
ich
d
is
r
u
p
t
th
eir
v
ital
f
u
n
ctio
n
s
an
d
alter
th
eir
n
at
u
r
al
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
E
fficien
t d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
fo
r
en
h
a
n
cin
g
p
la
n
t le
a
f
d
is
ea
s
e
cla
s
s
if
ica
tio
n
(
Mer
o
u
a
B
elmir
)
1113
s
tr
u
ctu
r
e
[
2
]
.
E
ar
ly
id
e
n
tific
atio
n
o
f
p
lan
t
d
is
ea
s
e
in
th
e
f
ield
s
tan
d
s
as
a
p
iv
o
tal
in
itial
p
h
ase.
C
o
n
v
en
tio
n
a
l
d
is
ea
s
e
id
en
tific
atio
n
m
eth
o
d
s
h
ea
v
ily
lean
o
n
th
e
ass
is
tan
ce
o
f
ag
r
icu
ltu
r
al
ex
ten
s
io
n
o
r
g
an
izatio
n
s
.
Ho
wev
er
,
th
ese
m
eth
o
d
s
f
ac
e
co
n
s
tr
ain
ts
in
co
u
n
t
r
ies
with
in
s
u
f
f
icien
t
lo
g
is
tical
an
d
h
u
m
an
in
f
r
astru
ctu
r
e
ca
p
ac
ity
,
an
d
th
eir
s
ca
lab
ilit
y
co
m
es
at
a
h
ig
h
co
s
t.
I
n
s
u
ch
co
n
tex
ts
,
th
e
in
ter
n
et
o
f
th
i
n
g
s
,
with
s
m
ar
tp
h
o
n
e
an
d
u
n
m
an
n
ed
ae
r
ial
v
e
h
icle
t
ec
h
n
o
lo
g
ies,
p
r
esen
t
n
o
v
el
av
en
u
es
f
o
r
in
-
f
ield
p
lan
t
d
is
ea
s
e
d
etec
tio
n
th
r
o
u
g
h
au
to
m
ated
im
ag
e
r
ec
o
g
n
itio
n
.
T
h
is
ap
p
r
o
ac
h
h
o
l
d
s
p
r
o
m
is
e
f
o
r
en
a
b
lin
g
ea
r
l
y
d
etec
tio
n
o
n
a
b
r
o
ad
s
ca
le
[
3
]
.
C
r
o
p
y
ield
r
ed
u
ctio
n
is
a
cr
iti
ca
l
ar
ea
o
f
r
esear
ch
,
esp
ec
ially
in
ca
s
es
wh
er
e
d
is
ea
s
es
o
r
a
b
n
o
r
m
alities
d
is
r
u
p
t
ch
lo
r
o
p
h
y
ll
p
r
o
d
u
ctio
n
in
p
lan
t
leav
es,
r
esu
ltin
g
in
p
lan
t
m
o
r
tality
.
Ar
tific
ia
l
i
n
tellig
en
ce
(
AI
)
h
as
ev
o
lv
ed
as
a
m
ajo
r
av
en
u
e
f
o
r
ad
d
r
ess
in
g
th
is
is
s
u
e
[
4
]
.
R
esear
ch
er
s
h
av
e
in
tr
o
d
u
ce
d
a
v
ar
iety
o
f
d
ee
p
lear
n
i
n
g
alg
o
r
ith
m
s
aim
ed
at
th
e
id
e
n
tific
atio
n
an
d
class
if
icatio
n
o
f
m
u
ltip
le
d
is
ea
s
es
o
f
p
lan
ts
.
I
m
an
u
llo
h
et
a
l.
[
5
]
,
th
is
wo
r
k
p
r
o
p
o
s
ed
a
cu
s
to
m
-
d
esig
n
ed
co
n
v
o
l
u
tio
n
al
n
e
u
r
al
n
etwo
r
k
(
C
NN)
ap
p
r
o
ac
h
c
o
m
p
r
is
in
g
1
2
lay
er
s
,
with
eig
h
t
d
ed
icate
d
to
f
ea
tu
r
e
ex
tr
ac
tio
n
an
d
f
o
u
r
s
er
v
in
g
as
clas
s
if
ier
s
.
Uti
lizin
g
a
n
ew
p
lan
t
d
is
ea
s
es
d
ataset,
th
e
p
r
o
p
o
s
ed
m
o
d
el
ac
h
iev
ed
h
ig
h
p
er
f
o
r
m
a
n
ce
with
o
u
t
o
v
er
f
itti
n
g
,
attain
in
g
an
ac
cu
r
ac
y
r
ate
o
f
9
7
%.
B
elm
ir
et
a
l.
[
6
]
u
tili
ze
d
th
e
Plan
tVillag
e
d
ataset
wh
ich
co
n
tai
n
s
3
8
class
es
an
d
a
d
ee
p
C
NN
f
o
r
th
e
d
etec
tio
n
an
d
class
if
icatio
n
o
f
p
lan
t
leav
es
d
is
ea
s
es
ac
h
iev
in
g
a
test
ac
cu
r
ac
y
o
f
9
4
.
3
3
%.
Dif
f
er
e
n
t
ap
p
r
o
ac
h
es
h
av
e
b
ee
n
a
p
p
lie
d
to
h
a
n
d
le
th
e
class
if
icatio
n
o
f
s
in
g
le
cr
o
p
s
.
Fo
r
ex
am
p
l
e,
in
[
7
]
a
d
e
n
s
ely
co
n
n
ec
ted
co
n
v
o
l
u
tio
n
al
n
et
wo
r
k
-
1
2
1
(
Den
s
eNe
t
-
1
2
1
)
d
e
ep
lear
n
in
g
m
o
d
el
was
p
r
o
p
o
s
ed
to
id
en
tify
s
ix
ca
teg
o
r
ies
o
f
ap
p
le
leaf
d
is
ea
s
es.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
in
d
icate
d
an
ac
cu
r
ac
y
r
ate
o
f
9
3
.
7
1
%.
Ad
d
itio
n
ally
,
th
ese
s
o
lu
tio
n
s
[
8
]
,
[
9
]
im
p
le
m
en
ted
d
ee
p
lear
n
in
g
ap
p
r
o
a
ch
es
f
o
r
th
e
p
u
r
p
o
s
e
o
f
class
if
y
in
g
th
e
im
a
g
es
o
f
ap
p
le
leaf
d
is
ea
s
es,
also
f
o
r
t
h
e
class
if
y
in
g
o
f
t
o
m
ato
cr
o
p
[
1
0
]
-
[
1
2
]
.
I
n
a
n
o
th
er
s
tu
d
y
[
1
3
]
,
tr
an
s
f
e
r
lear
n
in
g
with
d
ee
p
C
NNs
wa
s
em
p
lo
y
ed
to
id
en
tif
y
p
lan
t
leaf
d
i
s
ea
s
es.
Pre
-
tr
ain
ed
m
o
d
els,
in
itially
tr
ain
ed
o
n
ex
ten
s
iv
e
d
atasets
,
wer
e
a
d
ap
t
ed
to
th
e
s
p
ec
if
ic
task
u
s
in
g
th
e
r
ice
a
n
d
m
aize
Plan
tVillag
e
d
ataset.
Mo
r
eo
v
e
r
,
Kh
an
et
a
l.
[
1
4
]
an
d
Gu
p
ta
et
a
l.
[
1
5
]
ap
p
lied
d
ee
p
lear
n
in
g
tech
n
iq
u
es
to
class
if
y
im
ag
es
o
f
m
aize
leaf
d
is
ea
s
es.
T
h
e
v
is
u
al
g
eo
m
etr
y
g
r
o
u
p
n
etwo
r
k
(
VGGNet
)
,
wh
ich
is
p
r
e
-
tr
ain
ed
o
n
th
e
I
m
a
g
eNe
t,
an
d
also
th
e
I
n
ce
p
tio
n
m
o
d
el
was
ch
o
s
en
f
o
r
th
is
p
u
r
p
o
s
e,
r
esu
ltin
g
i
n
a
n
ac
cu
r
ac
y
o
f
9
2
.
0
0
%.
C
h
en
e
t
a
l.
[
1
6
]
,
e
n
h
a
n
ce
d
an
ar
tific
ial
n
e
u
r
al
n
etwo
r
k
b
y
in
p
u
ttin
g
e
x
tr
ac
ted
p
ix
el
a
n
d
f
ea
tu
r
e
v
alu
es
f
o
r
im
ag
e
s
eg
m
en
tatio
n
.
Nex
t,
a
C
NN
-
b
ased
m
o
d
el
was
e
s
tab
lis
h
ed
,
an
d
th
e
s
eg
m
en
ted
im
ag
es
wer
e
class
if
ied
u
s
in
g
th
e
p
r
o
p
o
s
ed
C
NN
m
o
d
el.
E
x
p
er
im
en
tal
f
in
d
in
g
s
in
d
icate
d
an
a
v
er
ag
e
ac
c
u
r
ac
y
o
f
9
3
.
7
5
%.
W
ith
in
th
e
s
co
p
e
o
f
t
h
is
r
esear
ch
,
we
in
tr
o
d
u
ce
a
s
p
ec
if
ically
tailo
r
ed
h
y
b
r
i
d
m
o
d
el
d
esig
n
ed
f
o
r
th
e
id
en
tific
atio
n
a
n
d
class
if
icatio
n
o
f
p
lan
t
leaf
d
is
ea
s
es
wh
ich
co
m
b
in
ed
p
r
e
-
tr
ai
n
ed
r
esid
u
al
n
etwo
r
k
-
50
(
R
es
N
et
-
50
)
with
D
en
s
e
N
et
-
121
.
T
h
e
m
et
h
o
d
o
lo
g
ical
f
r
am
ewo
r
k
co
m
p
r
is
es
th
r
ee
p
iv
o
tal
s
tep
s
:
d
ata
co
llectio
n
,
p
r
e
-
p
r
o
ce
s
s
in
g
o
f
d
ata,
an
d
th
e
class
if
icatio
n
.
T
h
e
o
v
er
ar
ch
i
n
g
g
o
al
is
to
cr
a
f
t
a
m
o
d
el
p
r
o
f
icien
t
in
d
if
f
er
en
tiatin
g
h
ea
lth
y
f
r
o
m
in
f
ec
ted
p
lan
t
f
o
liag
e.
T
h
e
n
e
w
p
lan
t
d
ataset
,
en
co
m
p
ass
in
g
a
d
iv
er
s
e
ar
r
ay
o
f
p
lan
t
v
ar
ieties,
s
er
v
es
as
th
e
f
o
u
n
d
atio
n
f
o
r
th
is
s
tu
d
y
.
T
h
e
d
is
ce
r
n
ib
le
r
esu
lts
u
n
d
er
s
co
r
e
th
e
s
ig
n
if
ican
tly
h
eig
h
ten
ed
ac
cu
r
ac
y
e
x
h
ib
ite
d
b
y
th
e
p
r
o
p
o
s
ed
m
o
d
el
i
n
co
m
p
a
r
is
o
n
to
co
n
v
en
tio
n
al
m
ac
h
in
e
lear
n
in
g
class
if
ier
s
wh
en
d
etec
tin
g
p
lan
t
leaf
d
is
ea
s
es.
T
o
ass
es
s
h
o
w
well
o
u
r
s
u
g
g
ested
s
o
lu
tio
n
p
er
f
o
r
m
s
,
we
co
m
p
ar
ed
its
p
er
f
o
r
m
a
n
ce
ag
ain
s
t
a
d
ee
p
lear
n
in
g
m
o
d
els
tr
ain
ed
o
n
th
e
s
am
e
d
ataset.
T
h
is
s
tu
d
y
m
ak
es
a
co
n
tr
ib
u
tio
n
to
co
m
b
in
in
g
d
ee
p
tr
an
s
f
er
lear
n
i
n
g
m
o
d
els
to
i
n
cr
ea
s
e
p
r
ed
ictio
n
ac
cu
r
ac
y
.
T
h
e
p
ap
er
is
s
y
s
tem
atica
lly
s
tr
u
ctu
r
ed
in
t
o
f
iv
e
d
is
tin
ct
s
ec
tio
n
s
f
o
r
clar
ity
.
T
h
e
s
ec
o
n
d
s
ec
tio
n
d
etails
th
e
m
eth
o
d
s
em
p
lo
y
ed
.
T
h
e
th
ir
d
s
ec
tio
n
d
elv
es
in
to
th
e
p
r
o
p
o
s
ed
m
eth
o
d
o
lo
g
y
f
o
r
id
e
n
tify
in
g
p
la
n
t
le
af
d
is
ea
s
es.
Sectio
n
f
o
u
r
th
o
r
o
u
g
h
ly
e
x
am
in
es
th
e
ex
p
er
im
en
tal
r
esu
lts
an
d
i
n
clu
d
es
co
m
p
ar
is
o
n
s
o
f
th
e
f
in
d
in
g
s
.
T
h
e
f
i
n
al
s
ec
tio
n
co
n
clu
d
es th
e
r
esear
ch
p
ap
er
.
2.
M
E
T
H
O
D
2
.
1
.
Da
t
a
c
o
llect
io
n
R
ec
en
tly
,
s
ev
er
al
ef
f
o
r
ts
h
a
v
e
b
ee
n
in
itiated
in
t
h
e
r
ea
lm
o
f
d
ata
g
ath
er
in
g
.
On
e
s
u
ch
in
itiativ
e
in
v
o
lv
es
th
e
ac
q
u
is
itio
n
o
f
i
m
ag
es
d
ep
ictin
g
m
u
ltip
le
p
la
n
t
s
p
ec
ies
af
f
ec
ted
b
y
d
if
f
er
e
n
t
d
is
ea
s
es
f
r
o
m
th
e
Kag
g
le
p
latf
o
r
m
,
s
p
ec
if
ically
f
r
o
m
th
e
d
ataset
titl
ed
‘
New
p
lan
t
d
is
ea
s
es
d
ataset
’
.
T
h
is
d
ataset
h
as
b
ee
n
r
ec
r
ea
ted
th
r
o
u
g
h
au
g
m
en
tat
io
n
f
r
o
m
th
e
o
r
ig
in
al
d
atas
et
“
Plan
tVillag
e
d
ataset
”
.
T
h
e
o
r
ig
in
al
d
ataset
co
n
tain
s
a
wid
e
v
ar
iety
o
f
im
ag
es
o
f
p
lan
t
s
p
ec
ies
af
f
ec
ted
b
y
d
if
f
e
r
en
t
d
is
ea
s
es.
T
h
e
n
ew
p
lan
t
d
is
ea
s
es
d
ataset
in
clu
d
es
8
7
,
8
6
7
r
ed
,
g
r
ee
n
,
an
d
b
lu
e
(
R
GB
)
im
ag
es
o
f
1
4
d
if
f
er
en
t
ty
p
es
o
f
cr
o
p
l
ea
v
es,
b
o
th
h
ea
lth
y
an
d
th
o
s
e
af
f
ec
ted
b
y
th
e
d
is
ea
s
e,
class
if
ied
in
to
3
8
class
es
o
f
p
lan
t
d
is
ea
s
es
ca
teg
o
r
ies
an
d
ca
n
b
e
ac
ce
s
s
ed
at
“
h
ttp
s
://www.
k
ag
g
le.
co
m
/d
atasets
/v
ip
o
o
o
o
o
l/n
ew
-
p
la
n
t
-
d
is
ea
s
es
-
d
ataset
”
.
Fo
r
p
r
e
d
ictiv
e
m
o
d
elin
g
,
th
e
d
ataset
was
s
p
lited
in
t
o
tr
ain
in
g
(
5
6
,
2
3
6
im
ag
es),
v
alid
atio
n
(
1
4
,
0
5
9
im
ag
e
s
)
,
an
d
test
s
ets
(
1
7
,
5
7
2
im
ag
es).
An
ex
p
e
r
im
en
t
u
tili
zin
g
1
4
cr
o
p
s
was
p
er
f
o
r
m
ed
,
Fig
u
r
e
1
r
ep
r
esen
ts
class
es
o
f
th
e
ap
p
le
leaf
d
atas
et
in
clu
d
e:
Fig
u
r
e
1
(
a)
h
ea
lth
y
leav
es,
Fig
u
r
e
1
(
b
)
r
u
s
t
leav
es,
Fig
u
r
e
1
(
c)
b
lack
r
o
t
leav
es,
an
d
Fig
u
r
e
1
(
d
)
s
ca
b
leav
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
a
r
y
20
2
5
:
1
11
2
-
1
1
20
1114
(
a)
(
b
)
(
c)
(
d
)
Fig
u
r
e
1
.
S
am
p
le
im
a
g
es r
ep
r
e
s
en
tin
g
v
ar
io
u
s
class
es wi
th
in
th
e
ap
p
le
leaf
d
ataset
:
(
a)
ap
p
l
e
h
ea
lth
y
,
(
b
)
ap
p
le
r
u
s
t,
(
c
)
ap
p
le
b
lack
r
o
t,
an
d
(
d
)
ap
p
le
s
ca
b
2
.
2
.
Dee
p lea
rning
On
e
o
f
th
e
m
o
s
t
s
ig
n
if
ican
t
b
r
ea
k
th
r
o
u
g
h
s
in
co
m
p
u
ter
s
cie
n
ce
,
f
u
n
d
am
en
tally
tr
a
n
s
f
o
r
m
i
n
g
th
e
d
ata
m
in
in
g
in
d
u
s
tr
y
,
is
d
ee
p
lear
n
in
g
.
I
t
h
as
tak
en
n
ea
r
ly
two
d
ec
ad
es
to
r
ea
ch
its
cu
r
r
en
t
lev
el
o
f
s
o
p
h
is
ticatio
n
,
d
r
iv
en
b
y
th
e
in
cr
ea
s
ed
av
ail
ab
ilit
y
o
f
p
u
b
lic
d
ata,
th
e
p
o
wer
f
u
l
p
a
r
allel
p
r
o
ce
s
s
in
g
ca
p
ab
ilit
ies
o
f
g
r
ap
h
ics
p
r
o
ce
s
s
in
g
u
n
its
(
GPU)
,
an
d
th
e
cr
ea
tio
n
o
f
s
p
ec
ialized
d
ee
p
lea
r
n
in
g
h
ar
d
wa
r
e
[
1
7
]
.
Dee
p
lear
n
in
g
f
r
am
ewo
r
k
s
ar
e
ex
ten
s
iv
ely
u
tili
ze
d
in
n
u
m
er
o
u
s
ap
p
li
ca
tio
n
s
o
f
th
e
class
if
icat
io
n
,
in
clu
d
in
g
im
ag
e
r
ec
o
g
n
itio
n
[
1
8
]
,
r
ec
o
g
n
itio
n
o
f
m
u
s
ic
[
1
9
]
,
an
d
m
ed
ical
d
is
ea
s
e
r
ec
o
g
n
itio
n
[
2
0
]
.
C
NNs
ar
e
a
u
n
i
q
u
e
ty
p
e
o
f
n
eu
r
al
n
etwo
r
k
d
esig
n
e
d
f
o
r
im
ag
e
r
ec
o
g
n
itio
n
an
d
class
if
icatio
n
,
ac
h
iev
in
g
e
x
ce
p
tio
n
al
r
esu
lts
.
Un
lik
e
tr
a
d
itio
n
al
ap
p
r
o
ac
h
es,
C
NNs
ca
n
au
t
o
m
atica
lly
lear
n
co
m
p
lex
f
ea
tu
r
es
f
r
o
m
r
aw
im
a
g
es,
elim
in
atin
g
th
e
n
ec
ess
ity
f
o
r
m
an
u
al
f
ea
tu
r
e
ex
tr
ac
tio
n
.
I
n
task
s
s
u
ch
as
id
en
tify
in
g
p
lan
t
s
p
ec
ies
an
d
d
iag
n
o
s
in
g
d
is
ea
s
es,
C
NNs h
av
e
d
em
o
n
s
tr
ated
g
r
ea
ter
ef
f
ec
tiv
en
ess
co
m
p
a
r
ed
to
co
n
v
en
tio
n
al
m
eth
o
d
s
[
1
3
]
.
2
.
3
.
T
ra
ns
f
er
lea
rning
Alth
o
u
g
h
d
ee
p
lear
n
in
g
h
as
d
em
o
n
s
tr
ated
g
r
ea
t
ef
f
ec
tiv
e
n
e
s
s
in
n
u
m
er
o
u
s
ap
p
licatio
n
s
,
t
h
er
e
ar
e
a
n
u
m
b
er
o
f
co
n
s
tr
ain
ts
th
at
p
r
ev
en
t
d
ee
p
lear
n
in
g
f
r
o
m
b
ei
n
g
u
s
ed
in
ce
r
tain
co
n
tex
ts
.
T
o
p
r
o
p
e
r
l
y
tr
ain
th
e
m
o
d
el
p
a
r
am
eter
s
,
a
s
u
b
s
tan
tial
am
o
u
n
t
o
f
la
b
eled
d
ata
is
n
ee
d
ed
.
T
h
is
is
o
n
e
m
ajo
r
r
estr
ictio
n
.
Gen
er
atin
g
lar
g
e
-
s
ca
le
tag
g
ed
d
atasets
is
f
r
eq
u
e
n
tly
n
o
t
f
ea
s
ib
le.
Ov
er
f
itti
n
g
ca
n
o
cc
u
r
wh
en
a
d
ee
p
n
eu
r
al
n
etwo
r
k
is
tr
ain
ed
en
tire
ly
f
r
o
m
s
cr
atc
h
u
s
in
g
s
p
ar
s
e
d
ata.
T
h
is
p
r
o
b
lem
is
s
o
lv
ed
th
r
o
u
g
h
tr
an
s
f
er
lear
n
in
g
,
wh
ich
ap
p
lies
th
e
k
n
o
wled
g
e
g
ain
e
d
f
r
o
m
o
n
e
ac
tiv
ity
to
o
th
er
r
elate
d
task
s
.
Ma
n
g
o
es
an
d
av
o
ca
d
o
s
ca
n
b
e
class
if
ied
u
s
in
g
a
m
o
d
el
th
at
was
tr
ain
ed
to
id
en
tify
p
h
o
to
s
o
f
ap
p
les
an
d
m
an
g
o
es,
f
o
r
in
s
tan
ce
.
T
h
e
I
m
ag
eNe
t
d
ataset
co
n
tain
s
im
ag
es
o
f
v
ar
io
u
s
r
ea
l
-
life
s
u
b
jects,
h
as
b
ee
n
in
s
tr
u
m
en
ta
l
in
p
r
o
m
o
tin
g
th
e
u
s
e
o
f
tr
an
s
f
er
lear
n
in
g
[
1
7
]
.
T
r
an
s
f
er
lear
n
in
g
u
tili
ze
s
k
n
o
wled
g
e
f
r
o
m
m
o
d
els
in
itiall
y
tr
ain
ed
o
n
lar
g
er
b
en
ch
m
ar
k
d
atasets
,
s
u
ch
as
I
m
ag
eNe
t,
an
d
ap
p
lies
it
to
s
im
ilar
o
r
d
if
f
er
e
n
t
task
s
,
lik
e
class
if
y
in
g
d
is
ea
s
e
im
ag
es.
Ho
wev
er
,
b
ec
a
u
s
e
o
f
th
e
d
if
f
er
e
n
ce
s
b
etwe
en
th
e
s
o
u
r
ce
d
ataset
(
I
m
a
g
eNe
t)
an
d
tar
g
et
d
atasets
(
n
ew
p
lan
t
d
is
ea
s
es
)
,
o
u
r
wo
r
k
f
o
cu
s
es
o
n
e
x
p
er
im
e
n
tin
g
with
th
e
f
o
llo
win
g
p
r
e
-
tr
ain
e
d
m
o
d
els:
R
esNet
-
50
an
d
Den
s
eNe
t
-
121
.
Fig
u
r
e
2
s
h
o
ws th
e
m
ec
h
an
is
m
o
f
th
is
tr
an
s
f
er
lear
n
in
g
.
Fig
u
r
e
2
.
T
h
e
m
ec
h
a
n
is
m
o
f
k
n
o
wled
g
e
tr
a
n
s
f
er
with
in
tr
an
s
f
er
lear
n
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
E
fficien
t d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
fo
r
en
h
a
n
cin
g
p
la
n
t le
a
f
d
is
ea
s
e
cla
s
s
if
ica
tio
n
(
Mer
o
u
a
B
elmir
)
1115
2
.
4
.
P
er
f
o
r
m
a
nce
e
v
a
lua
t
io
n
T
o
co
n
v
in
cin
g
l
y
d
em
o
n
s
tr
ate
th
e
ef
f
ec
tiv
e
n
ess
o
f
th
e
s
u
g
g
ested
h
y
b
r
id
m
o
d
el,
it
is
es
s
en
tial
to
ass
es
s
th
e
ef
f
icien
cy
o
f
th
e
p
r
ed
ictio
n
m
o
d
el
t
h
o
r
o
u
g
h
l
y
.
T
h
er
e
ar
e
n
u
m
er
o
u
s
m
etr
ics
av
ai
lab
le
to
ass
ess
h
o
w
well
a
m
o
d
el
p
r
e
d
icts
o
u
tco
m
es.
I
n
th
is
s
tu
d
y
,
we
f
o
cu
s
o
n
s
ev
er
al
k
ey
p
e
r
f
o
r
m
an
ce
m
ea
s
u
r
es,
wh
ich
ar
e
d
etailed
b
elo
w,
to
h
i
g
h
lig
h
t
t
h
e
m
o
d
el
’
s
p
r
e
d
ictiv
e
ac
cu
r
ac
y
an
d
r
eliab
ilit
y
.
Acc
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e
th
ese
u
s
ed
m
etr
ics
ar
e
d
er
iv
e
d
f
r
o
m
v
alu
es
o
b
tain
e
d
f
r
o
m
t
h
e
co
n
f
u
s
io
n
m
atr
ix
.
Fo
r
b
o
t
h
b
i
n
ar
y
an
d
m
u
lticlas
s
class
if
ica
tio
n
task
s
,
th
e
c
o
n
f
u
s
io
n
m
atr
i
x
is
a
wid
ely
u
s
ed
tech
n
iq
u
e
f
o
r
e
v
alu
atin
g
th
e
ef
f
ec
tiv
en
ess
o
f
class
if
icatio
n
m
o
d
els.
T
h
e
co
u
n
ts
o
f
t
h
e
ex
p
ec
ted
a
n
d
ac
t
u
al
v
alu
e
s
ar
e
s
h
o
wn
.
“
TN
”
r
ep
r
esen
ts
tr
u
e
n
eg
ativ
e
,
d
en
o
t
in
g
th
e
q
u
an
tity
o
f
n
eg
ativ
e
s
am
p
les
th
at
h
av
e
b
ee
n
ac
cu
r
ate
ly
id
en
tifie
d
.
“
TP
”
r
ep
r
esen
ts
tr
u
e
p
o
s
itiv
e
,
w
h
ich
in
d
icate
s
th
e
q
u
an
tity
o
f
p
o
s
itiv
e
ca
s
es
th
at
h
av
e
b
ee
n
co
r
r
ec
tly
class
if
ied
.
Fals
e
p
o
s
itiv
e
(
ab
b
r
ev
iated
“
FP
”
)
is
th
e
n
u
m
b
e
r
o
f
n
eg
ativ
e
ca
s
es
th
at
ar
e
m
is
tak
en
ly
ca
te
g
o
r
ized
as
p
o
s
itiv
e.
Fals
e
n
eg
ativ
e,
o
r
“
FN
”
f
o
r
s
h
o
r
t,
is
th
e
q
u
an
tity
o
f
p
o
s
itiv
e
ex
am
p
les
th
at
a
r
e
m
is
tak
e
n
ly
ca
teg
o
r
ized
as
n
eg
ativ
e
[
2
1
]
.
T
h
e
p
er
ce
n
tag
e
o
f
th
e
m
o
d
el
’
s
p
r
ed
ictio
n
s
th
at
co
m
e
tr
u
e
is
it
s
ac
cu
r
ac
y
.
A
m
o
d
el
’
s
r
ec
all,
o
r
s
en
s
itiv
ity
,
m
ea
s
u
r
es
h
o
w
wel
l
it
id
en
tifie
s
tr
u
e
p
o
s
itiv
e
ev
en
ts
o
u
t
o
f
all
th
e
ac
t
u
al
p
o
s
i
tiv
e
in
s
tan
ce
s
.
T
h
e
tr
u
e
p
o
s
itiv
e
to
to
tal
tr
u
e
p
o
s
itiv
e
an
d
f
alse
p
o
s
itiv
e
r
atio
is
k
n
o
wn
as
p
r
ec
is
io
n
.
T
h
e
p
r
ec
is
io
n
an
d
r
ec
al
l
b
alan
ce
ar
e
in
d
icate
d
b
y
th
e
F
1
v
alu
e
[
2
2
]
.
=
+
+
+
+
(
1
)
=
+
(
2
)
=
+
(
3
)
1
−
=
2
.
(
.
)
(
+
)
(
4
)
3.
T
H
E
P
RO
P
O
SE
D
SYS
T
E
M
I
n
th
is
p
a
p
er
,
th
e
s
u
g
g
ested
p
r
o
ce
s
s
co
n
s
is
ts
o
f
f
iv
e
s
tep
s
,
as
illu
s
tr
ated
in
Fig
u
r
e
3
.
T
h
e
m
eth
o
d
o
l
o
g
y
o
f
th
is
wo
r
k
s
tar
ts
with
th
e
s
tep
o
f
d
ata
co
lle
ctio
n
,
wh
ich
i
n
v
o
lv
es
g
at
h
er
i
n
g
a
v
a
r
iety
o
f
p
lan
t
leaf
im
ag
es f
r
o
m
th
e
n
ew
p
lan
t d
is
ea
s
e
s
d
ataset
,
o
f
f
er
in
g
a
r
ich
ar
r
ay
o
f
d
ata
f
o
r
an
aly
s
is
.
Fo
llo
win
g
th
is
,
d
ata
p
r
e
-
p
r
o
ce
s
s
in
g
b
ec
o
m
es
cr
u
ci
al,
as
it
d
ea
ls
with
th
e
d
iv
er
s
e
s
h
ap
es
an
d
r
eso
lu
tio
n
s
o
f
th
e
co
lle
cted
im
ag
es.
T
o
en
s
u
r
e
co
n
s
is
ten
cy
ac
r
o
s
s
all
im
ag
es,
we
r
esize
th
em
to
a
s
tan
d
ar
d
ized
d
im
e
n
s
io
n
o
f
2
5
6
×
2
5
6
×
3
an
d
ap
p
ly
tech
n
iq
u
es
lik
e
im
ag
e
au
g
m
en
tatio
n
.
W
ith
th
e
p
r
e
-
p
r
o
ce
s
s
ed
d
ataset
in
h
an
d
,
th
e
s
u
b
s
eq
u
en
t
s
tep
is
m
o
d
el
b
u
ild
i
n
g
,
wh
er
e
a
h
y
b
r
id
m
o
d
el
is
co
n
s
tr
u
cted
.
T
h
i
s
m
ix
ed
m
o
d
el
co
m
b
in
ed
p
r
e
tr
ain
ed
R
esNet
-
50
-
Den
s
eNe
t
-
121
is
s
p
ec
if
ically
d
esig
n
ed
to
class
if
y
p
lan
t
d
is
ea
s
es.
T
h
e
s
tep
af
ter
b
u
ild
in
g
th
e
m
o
d
el
is
to
tr
ain
an
d
test
it.
Mo
d
el
e
v
alu
atio
n
b
ec
o
m
es
p
iv
o
tal.
T
h
is
s
tag
e
ass
ess
es
th
e
ef
f
icac
y
o
f
th
e
h
y
b
r
id
m
o
d
el
,
d
eter
m
in
in
g
its
ac
cu
r
ac
y
an
d
ef
f
ec
tiv
en
ess
in
class
if
y
in
g
p
lan
t
d
is
ea
s
es
b
ased
o
n
th
e
p
r
o
v
id
e
d
test
d
ata.
T
h
r
o
u
g
h
th
is
co
m
p
r
eh
e
n
s
iv
e
e
v
alu
atio
n
,
in
s
ig
h
ts
ar
e
g
ain
ed
i
n
to
th
e
m
o
d
el
’
s
ca
p
ab
ilit
ies.
Fig
u
r
e
3
.
Diag
r
a
m
f
o
r
th
e
s
u
g
g
ested
m
o
d
el
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
a
r
y
20
2
5
:
1
11
2
-
1
1
20
1116
3
.
1
.
P
r
o
po
s
ed
t
ra
ns
f
er
lea
rning
f
ra
m
ewo
r
k
T
h
e
R
esNet
-
50
an
d
Den
s
e
Net
-
121
m
o
d
els
p
r
e
-
tr
ai
n
ed
o
n
th
e
I
m
ag
eNe
t
d
ataset
ar
e
u
s
ed
.
I
n
th
is
s
cien
tific
ex
p
er
im
en
t
,
we
h
av
e
em
p
lo
y
e
d
R
esNet
-
50
an
d
Den
s
eNe
t
-
121
s
tem
m
in
g
f
o
r
th
e
f
o
llo
win
g
r
ea
s
o
n
s
:
−
T
h
e
tr
ain
in
g
p
r
o
ce
s
s
o
f
th
e
R
esNet
-
50
an
d
Den
s
eNe
t
-
1
2
1
m
o
d
els
is
ef
f
icien
t,
with
R
esNet
-
5
0
f
a
m
o
u
s
f
o
r
its
p
r
o
f
icien
cy
in
d
ee
p
n
etwo
r
k
tr
ain
in
g
ef
f
ec
tiv
ely
th
r
o
u
g
h
r
esid
u
al
co
n
n
ec
tio
n
s
an
d
Den
s
eNe
t
-
121
f
o
r
its
f
ea
tu
r
e
r
eu
s
e,
t
h
er
eb
y
r
ed
u
cin
g
p
ar
am
eter
co
u
n
t a
n
d
e
n
h
an
ci
n
g
co
m
p
u
tatio
n
al
ef
f
icien
cy
.
−
R
esNet
-
50
an
d
Den
s
eNe
t
-
121
ar
e
b
o
th
h
i
g
h
ly
r
ec
o
m
m
en
d
ed
d
ee
p
lear
n
in
g
m
o
d
els
f
o
r
v
ar
io
u
s
ap
p
licatio
n
s
,
p
r
o
v
id
in
g
s
tr
o
n
g
p
er
f
o
r
m
an
ce
an
d
r
o
b
u
s
tn
ess
.
3
.
1
.
1
.
ResNet
-
50
a
rc
hite
ct
ure
R
esNet
-
50
i
n
tr
o
d
u
ce
d
b
y
He
et
a
l
.
[
2
3
]
,
it
em
er
g
es
as
a
s
tr
e
am
lin
ed
ar
c
h
itectu
r
e
with
in
th
e
r
ea
lm
o
f
d
ee
p
C
NNs
.
C
o
m
p
r
is
in
g
5
0
lay
er
s
,
it
d
iv
er
g
es
f
r
o
m
co
n
v
en
tio
n
al
ap
p
r
o
ac
h
es
b
y
r
e
d
ef
in
in
g
its
lear
n
in
g
p
r
o
ce
s
s
: in
s
tead
o
f
g
r
a
p
p
lin
g
with
u
n
r
estra
in
ed
f
u
n
ctio
n
s
,
it
p
iv
o
ts
to
war
d
s
m
aster
in
g
r
esid
u
al
f
u
n
ctio
n
s
wh
ile
s
tay
in
g
g
r
o
u
n
d
ed
i
n
r
ef
er
e
n
ce
to
th
e
lay
er
in
p
u
ts
.
T
h
is
R
esNet
m
o
d
el
is
s
tr
u
ctu
r
ed
ar
o
u
n
d
a
s
er
ies
o
f
r
esid
u
al
b
lo
ck
s
,
ea
ch
f
u
n
ctio
n
in
g
ak
in
to
a
s
tack
o
f
co
n
v
o
lu
tio
n
al
lay
er
s
.
No
tab
ly
,
th
e
o
u
tp
u
t o
f
ea
c
h
b
lo
ck
in
ter
twin
es
with
its
o
wn
in
p
u
t
v
ia
an
id
e
n
tity
m
ap
p
in
g
p
ath
way
.
T
h
is
d
esig
n
n
o
t
o
n
ly
u
p
h
o
ld
s
th
e
t
im
e
co
m
p
lex
ity
p
er
lay
er
b
u
t
also
p
er
p
etu
ates
th
e
d
o
wn
s
am
p
lin
g
o
f
f
ea
tu
r
e
m
a
p
p
in
g
th
r
o
u
g
h
s
tr
id
ed
co
n
v
o
l
u
tio
n
,
co
n
c
u
r
r
en
tly
am
p
lify
in
g
c
h
an
n
el
d
ep
th
[
2
3
]
,
[
2
4
]
.
3
.
1
.
2
.
DenseNet
-
1
2
1
a
rc
hite
ct
ure
Den
s
eNe
t
-
121
h
as
g
ain
e
d
r
ec
o
g
n
itio
n
in
th
e
r
ea
lm
o
f
C
NNs
d
u
e
to
its
d
is
tin
ctiv
e
d
en
s
e
co
n
n
ec
tiv
ity
p
atter
n
,
wh
ich
en
c
o
u
r
ag
es
e
f
f
ec
tiv
e
f
ea
tu
r
e
r
eu
s
e
a
n
d
p
r
o
p
ag
atio
n
ac
r
o
s
s
th
e
n
etwo
r
k
.
T
h
e
ar
ch
itectu
r
e
co
n
s
is
ts
o
f
1
2
1
la
y
er
s
o
r
g
an
i
ze
d
in
to
d
en
s
e
b
lo
c
k
s
an
d
tr
a
n
s
itio
n
lay
er
s
,
co
n
cl
u
d
in
g
with
a
g
lo
b
al
av
er
a
g
e
p
o
o
lin
g
lay
er
f
o
llo
wed
b
y
a
So
f
tMa
x
class
if
ier
.
W
ith
in
d
en
s
e
b
lo
c
k
s
,
co
n
v
o
l
u
tio
n
al
lay
er
s
ar
e
d
en
s
ely
in
ter
co
n
n
ec
ted
,
wh
ile
tr
an
s
itio
n
lay
er
s
r
eg
u
late
s
p
atial
d
im
e
n
s
io
n
s
an
d
f
ea
tu
r
e
m
ap
q
u
an
ti
ties
b
etwe
en
d
en
s
e
b
lo
ck
s
.
T
h
ese
tr
an
s
itio
n
la
y
er
s
f
r
eq
u
e
n
tly
em
p
lo
y
tech
n
iq
u
e
s
s
u
ch
as
b
atch
n
o
r
m
aliza
tio
n
,
1
×
1
co
n
v
o
lu
tio
n
s
,
an
d
av
er
a
g
e
p
o
o
lin
g
t
o
ac
h
iev
e
d
im
en
s
io
n
ality
r
e
d
u
ctio
n
a
n
d
d
o
wn
s
am
p
lin
g
[
2
5
]
.
3
.
2
.
T
he
pro
po
s
ed
hy
brid
m
o
del
T
r
ain
in
g
a
d
ee
p
n
e
u
r
al
n
etwo
r
k
f
r
o
m
s
cr
atch
r
eq
u
ir
es
a
s
ig
n
if
ican
t
am
o
u
n
t
o
f
d
ata.
T
h
er
ef
o
r
e,
w
e
u
tili
ze
two
p
r
e
-
tr
ain
ed
d
ee
p
le
ar
n
in
g
m
o
d
els,
R
esNet
-
50
an
d
Den
s
eNe
t
-
121
,
b
o
th
o
f
wh
ich
h
av
e
b
ee
n
tr
ain
ed
o
n
th
e
I
m
ag
eNe
t
d
ataset.
E
ac
h
m
o
d
el
p
r
o
ce
s
s
es
th
e
d
ata
in
d
ep
en
d
e
n
tly
,
a
n
d
th
eir
o
u
t
p
u
t
s
ar
e
m
er
g
e
d
b
ef
o
r
e
p
ass
in
g
th
r
o
u
g
h
f
u
lly
c
o
n
n
ec
ted
lay
er
s
f
o
r
im
ag
e
class
if
icatio
n
in
to
3
8
ca
te
g
o
r
ies.
T
h
e
f
in
al
class
if
icatio
n
lay
er
ad
d
r
ess
es
3
8
class
es
o
f
p
lan
t
leaf
d
is
ea
s
es,
as
d
ep
icte
d
in
Fig
u
r
e
4
.
Su
b
s
eq
u
e
n
tly
,
th
e
m
o
d
el
is
f
i
ne
-
tu
n
ed
o
n
th
e
Plan
tVillag
e
d
ataset.
T
o
ad
ju
s
t
th
e
weig
h
t
s
o
f
th
e
n
eu
r
al
n
etwo
r
k
,
we
u
tili
ze
th
e
Ad
am
o
p
tim
izer
.
W
e
im
p
lem
en
t
ea
r
ly
s
to
p
p
in
g
b
ased
o
n
th
e
lo
west
v
alid
atio
n
lo
s
s
,
with
a
p
atien
ce
p
ar
am
eter
o
f
5
.
T
r
ea
tin
g
th
e
task
as
a
m
u
ltip
l
e
class
es
’
ca
te
g
o
r
izatio
n
is
s
u
e,
ev
er
y
i
n
s
tan
ce
is
class
if
ied
in
to
o
n
e
o
f
th
e
3
8
class
es.
T
h
e
h
y
p
er
p
ar
am
eter
s
f
o
r
tr
ain
in
g
ar
e:
lear
n
in
g
r
ate=
1
e
-
3
,
b
atch
s
ize=
3
2
,
e
p
o
ch
s
=2
0
,
an
d
p
atien
ce
=5
.
T
h
ese
v
alu
es,
r
ec
o
m
m
en
d
e
d
b
y
p
r
ev
io
u
s
r
esear
ch
,
en
s
u
r
e
o
p
tim
al
p
er
f
o
r
m
an
ce
.
T
h
e
m
o
d
el
’
s
ar
ch
itectu
r
e
is
d
ev
elo
p
e
d
an
d
ev
alu
ated
u
s
in
g
Ker
as lib
r
ar
y
t
h
at
o
p
er
ates
with
T
en
s
o
r
Flo
w
as its
b
ac
k
en
d
.
Fig
u
r
e
4
.
Ar
c
h
itectu
r
e
o
f
th
e
s
u
g
g
ested
h
y
b
r
id
m
o
d
el
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
E
fficien
t d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
fo
r
en
h
a
n
cin
g
p
la
n
t le
a
f
d
is
ea
s
e
cla
s
s
if
ica
tio
n
(
Mer
o
u
a
B
elmir
)
1117
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
p
r
im
ar
y
aim
o
f
th
e
s
u
g
g
ested
p
ap
er
is
to
ass
e
s
s
th
e
ef
f
icac
y
o
f
th
e
p
r
etr
ai
n
ed
R
esNet
-
50
with
Den
s
eNe
t
-
121
h
y
b
r
id
a
r
ch
itec
tu
r
e
in
class
if
y
in
g
p
lan
t
lea
f
d
is
ea
s
es
an
d
to
co
m
p
ar
e
its
p
er
f
o
r
m
an
ce
with
t
h
at
o
f
s
tate
-
of
-
th
e
-
ar
t
d
ee
p
lear
n
i
n
g
m
o
d
els
in
th
e
l
iter
atu
r
e.
T
o
ass
ess
th
e
p
r
o
p
o
s
ed
s
y
s
tem
,
we
u
tili
ze
d
a
r
an
g
e
o
f
class
if
icatio
n
p
er
f
o
r
m
a
n
ce
in
d
icato
r
s
in
clu
d
in
g
th
e
co
n
f
u
s
io
n
m
atr
ix
,
ac
cu
r
ac
y
,
p
r
ec
is
i
o
n
,
r
ec
all,
an
d
th
e
F1
-
s
co
r
e.
T
h
is
s
elec
tio
n
was
m
ad
e
d
u
e
to
t
h
e
p
r
e
v
alen
ce
o
f
th
ese
cr
iter
ia
in
p
a
p
e
r
s
ev
alu
atin
g
th
eir
ap
p
r
o
ac
h
es
o
n
th
e
Plan
tVillag
e
d
ataset.
Ad
o
p
tin
g
th
ese
m
etr
ics
f
ac
ilit
ates
co
m
p
ar
is
o
n
an
d
e
n
ab
les
u
s
to
ef
f
ec
tiv
ely
p
o
s
itio
n
o
u
r
wo
r
k
in
th
e
c
o
n
tex
t
o
f
o
th
er
s
t
ate
-
of
-
th
e
-
a
r
t
m
eth
o
d
s
.
T
h
e
m
ix
ed
R
esNet
-
50
-
Den
s
eNe
t
-
121
m
o
d
el
was
tr
ai
n
ed
u
s
ed
GPU.
W
e
ac
h
iev
ed
9
9
.
8
7
%
o
n
t
h
e
tr
ain
in
g
ac
cu
r
ac
y
.
T
h
e
ac
h
iev
ed
o
u
tco
m
es
ar
e
d
ep
icted
in
Fig
u
r
e
s
5
a
n
d
6
.
T
h
e
r
esu
lts
o
f
t
h
e
s
u
g
g
ested
R
esNet
-
50
-
Den
s
eNe
t
-
121
a
p
p
r
o
ac
h
in
d
icate
an
ac
c
u
r
ac
y
o
f
9
9
.
6
6
%,
p
r
ec
is
io
n
o
f
9
9
.
6
7
%,
r
ec
al
l
o
f
9
9
.
6
5
%,
an
d
F1
-
s
co
r
e
o
f
9
9
.
6
6
%.
Fig
u
r
e
5
illu
s
tr
ates
th
e
ev
o
lu
tio
n
o
f
th
e
tr
ain
in
g
/
v
alid
atio
n
lo
s
s
f
u
n
ctio
n
an
d
also
th
e
f
lu
ctu
atio
n
s
o
f
th
e
ac
c
u
r
ac
y
ac
r
o
s
s
th
e
tr
ain
in
g
ep
o
c
h
s
.
Ob
s
er
v
in
g
Fig
u
r
e
5
(
a)
w
h
ich
r
e
p
r
esen
ts
th
e
g
r
ap
h
o
f
tr
ain
a
n
d
v
alid
atio
n
ac
cu
r
ac
y
,
it
’
s
ev
id
en
t
th
at
th
e
m
o
d
el
’
s
a
cc
u
r
ac
y
in
itially
in
cr
ea
s
es
r
ap
id
ly
,
s
tab
ilizes
af
ter
a
ce
r
tain
n
u
m
b
er
o
f
ep
o
ch
s
,
an
d
ev
en
t
u
ally
r
ea
ch
es
its
p
ea
k
p
er
f
o
r
m
an
ce
,
ac
h
iev
in
g
9
9
.
8
7
%
ac
cu
r
ac
y
af
ter
2
0
ep
o
ch
s
.
I
n
Fig
u
r
e
5
(
b
)
wh
ich
r
ep
r
esen
ts
th
e
g
r
a
p
h
o
f
tr
ain
an
d
v
alid
atio
n
lo
s
s
,
th
e
lo
s
s
f
u
n
ctio
n
ex
h
ib
its
an
in
v
er
s
e
tr
en
d
,
s
tead
ily
d
ec
r
ea
s
in
g
in
itially
,
th
en
s
tab
ilizin
g
,
an
d
u
ltima
tely
co
n
v
er
g
in
g
to
its
m
in
im
u
m
v
alu
e.
T
h
e
co
n
f
u
s
io
n
m
atr
ix
p
r
o
d
u
ce
d
b
y
th
e
s
u
g
g
ested
m
o
d
el
is
s
h
o
wn
in
Fig
u
r
e
6
.
T
h
e
p
r
i
m
ar
y
d
iag
o
n
al
c
o
m
p
o
n
en
ts
o
f
t
h
e
m
at
r
ix
,
wh
ich
s
tan
d
f
o
r
tr
u
e
p
o
s
itiv
es
th
at
ar
e
h
ig
h
ly
v
alu
ed
in
all
ca
teg
o
r
ies,
in
d
icatin
g
th
at
th
e
d
ataset
’
s
ex
am
p
les
with
in
ea
ch
class
wer
e
ac
c
u
r
ately
ca
teg
o
r
ized
.
T
h
is
co
n
f
ir
m
s
th
e
p
r
ec
is
io
n
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el
in
ca
teg
o
r
izin
g
e
f
f
ec
tiv
ely
.
(
a)
(
b
)
Fig
u
r
e
5
.
O
b
tain
e
d
r
esu
lts
u
s
in
g
h
y
b
r
id
m
o
d
el
o
f
(
a
)
tr
ain
a
n
d
v
alid
atio
n
ac
cu
r
ac
y
an
d
(
b
)
tr
ain
an
d
v
alid
atio
n
lo
s
s
g
r
ap
h
T
h
e
o
u
tco
m
es
f
r
o
m
th
is
r
esear
ch
p
ap
er
h
ig
h
lig
h
t
th
e
r
o
b
u
s
tn
ess
o
f
th
e
m
ix
ed
m
o
d
el
in
ta
ck
lin
g
th
e
ch
allen
g
es
o
f
p
lan
t
leaf
d
is
ea
s
e
class
if
icat
io
n
.
Ou
r
m
o
d
el
ac
h
iev
ed
im
p
r
ess
iv
e
f
in
d
in
g
s
,
with
an
ac
cu
r
ac
y
r
ate
o
f
9
9
.
6
6
%.
T
h
ese
o
u
tco
m
es
em
p
h
asize
th
e
m
o
d
el
’
s
r
o
b
u
s
tn
ess
in
im
ag
e
-
b
ased
p
lan
t
d
i
s
ea
s
e
id
en
tific
atio
n
task
s
.
E
v
alu
atin
g
th
e
h
y
b
r
i
d
m
o
d
el
’
s
p
er
f
o
r
m
an
ce
f
u
r
th
er
r
ev
ea
ls
its
ex
ce
p
tio
n
al
ca
p
ab
ilit
ies
in
p
lan
t
d
is
ea
s
e
class
if
icatio
n
.
T
h
e
m
o
d
el
ex
h
ib
ited
r
em
ar
k
ab
le
p
r
ec
is
io
n
,
r
e
ca
ll,
an
d
F1
-
s
co
r
e
v
alu
es
ac
r
o
s
s
m
u
ltip
le
d
is
ea
s
e
ty
p
es,
in
d
icatin
g
r
em
ar
k
a
b
le
a
cc
u
r
ac
y
an
d
r
eliab
ilit
y
.
Pre
cisi
o
n
v
alu
es
c
o
n
s
is
ten
tly
r
ef
lect
t
h
e
m
o
d
el
’
s
s
k
ill
in
p
r
ec
is
ely
id
en
tify
i
n
g
s
u
cc
ess
f
u
l
in
s
tan
ce
s
wh
ile
m
in
im
izin
g
f
alse
p
o
s
itiv
es.
Similar
ly
,
r
ec
all
v
alu
es
with
in
th
e
s
am
e
r
an
g
e
d
em
o
n
s
tr
ate
th
e
m
o
d
el
’
s
ef
f
ec
tiv
en
e
s
s
in
r
ec
o
g
n
izin
g
m
o
s
t
tr
u
e
p
o
s
itiv
e
ca
s
es.
T
h
e
co
n
s
is
ten
tly
h
ig
h
F1
-
s
co
r
es
u
n
d
er
s
co
r
e
t
h
e
m
o
d
el
’
s
ac
cu
r
ac
y
an
d
r
o
b
u
s
tn
ess
in
d
is
tin
g
u
is
h
in
g
b
etwe
en
h
ea
lth
y
an
d
d
is
ea
s
ed
p
lan
ts
.
I
n
s
u
m
m
ar
y
,
th
ese
r
esu
lts
af
f
ir
m
th
e
m
o
d
el
’
s
s
u
itab
ilit
y
f
o
r
th
e
cr
itical
task
o
f
p
lan
t
d
is
ea
s
e
d
etec
tio
n
,
p
r
esen
tin
g
a
p
r
o
m
is
in
g
s
o
lu
tio
n
to
im
p
r
o
v
e
ag
r
icu
ltu
r
al
p
r
ac
tices a
n
d
cr
o
p
m
a
n
a
g
em
en
t.
T
o
g
ain
a
c
o
m
p
r
e
h
en
s
iv
e
p
er
s
p
ec
tiv
e,
we
co
m
p
a
r
ed
o
u
r
r
es
u
lts
with
th
o
s
e
f
r
o
m
p
r
e
v
io
u
s
s
tu
d
ies,
as
s
h
o
wn
in
T
a
b
le
1
,
f
o
cu
s
in
g
o
n
cr
o
p
ty
p
e,
d
ataset
u
s
ed
,
n
u
m
b
er
o
f
class
es,
m
o
d
els
u
s
ed
,
an
d
o
b
tain
ed
r
esu
lts
.
W
h
ile
s
o
m
e
s
tu
d
ies
f
o
cu
s
ed
o
n
a
s
in
g
le
cr
o
p
lik
e
ap
p
le,
m
ai
ze
,
o
r
r
ice,
o
u
r
e
x
p
er
im
en
t
en
c
o
m
p
ass
ed
m
u
ltip
le
cr
o
p
s
.
Ad
d
itio
n
ally
,
we
u
s
ed
a
d
ataset
with
3
8
class
es,
wh
ich
is
lar
g
e
r
th
a
n
t
h
o
s
e
in
s
o
m
e
o
th
er
s
tu
d
ies
th
at
u
s
ed
s
m
aller
d
atasets
.
T
h
e
p
lan
t
d
is
ea
s
e
id
en
tific
atio
n
m
eth
o
d
u
s
in
g
R
esNet
-
50
-
Den
s
eNe
t
-
121
h
y
b
r
id
m
o
d
el
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
a
r
y
20
2
5
:
1
11
2
-
1
1
20
1118
p
r
esen
ted
in
th
is
wo
r
k
is
co
m
p
ar
ed
with
s
ev
er
al
o
t
h
er
cl
ass
if
icatio
n
ap
p
r
o
ac
h
es,
in
clu
d
in
g
C
NN
[
5
]
,
[
6
]
,
Den
s
eNe
t
-
121
[
7
]
,
p
r
e
-
tr
ain
e
d
VGGN
et
with
I
n
ce
p
tio
n
[
1
3
]
,
a
n
d
s
eg
m
e
n
tatio
n
with
C
NN
m
o
d
el
-
b
ased
class
if
icatio
n
[
1
6
]
.
T
ab
le
1
il
lu
s
tr
ates
th
e
ac
cu
r
ac
y
co
m
p
a
r
is
o
n
f
o
r
th
e
ca
teg
o
r
izatio
n
o
f
th
e
Plan
tVillag
e
d
ataset.
T
h
e
o
b
tain
ed
o
u
tco
m
es
d
em
o
n
s
tr
ate
th
at
th
e
p
r
o
p
o
s
ed
h
y
b
r
id
m
o
d
el
ac
h
iev
es
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
.
I
n
o
u
r
an
al
y
s
is
,
we
o
b
s
er
v
ed
th
e
lim
ited
u
s
e
o
f
h
y
b
r
id
m
o
d
e
ls
ac
r
o
s
s
m
o
s
t
s
tu
d
ies,
with
th
e
ex
ce
p
tio
n
in
[
1
3
]
an
d
it
ac
h
iev
ed
lo
wer
ac
cu
r
ac
y
in
th
e
o
u
tco
m
es
co
m
p
a
r
ed
to
th
is
s
tu
d
y
.
I
n
c
o
n
tr
ast,
o
u
r
r
esear
ch
lev
er
a
g
es
th
e
s
tr
en
g
th
s
o
f
h
y
b
r
id
m
o
d
els,
wh
ich
co
m
b
in
e
v
a
r
io
u
s
m
eth
o
d
o
lo
g
ies to
e
n
h
an
ce
p
er
f
o
r
m
an
ce
.
As
a
r
esu
lt,
we
ac
h
iev
e
a
h
ig
h
er
lev
el
o
f
ac
cu
r
ac
y
an
d
r
o
b
u
s
tn
ess
in
o
u
r
f
in
d
in
g
s
,
s
h
o
w
ca
s
in
g
th
e
ad
v
an
tag
es
o
f
in
co
r
p
o
r
atin
g
h
y
b
r
id
ap
p
r
o
ac
h
es
in
d
ata
an
aly
s
is
an
d
p
r
ed
ictiv
e
m
o
d
elin
g
.
Ou
r
r
esu
lts
u
n
d
er
s
co
r
e
th
e
p
o
ten
tial
o
f
h
y
b
r
id
m
o
d
els
to
o
u
tp
e
r
f
o
r
m
tr
a
d
itio
n
al
m
eth
o
d
s
,
h
ig
h
lig
h
tin
g
th
e
im
p
o
r
ta
n
ce
o
f
in
n
o
v
atio
n
an
d
in
teg
r
atio
n
in
a
d
v
an
cin
g
th
e
f
ield
.
Fig
u
r
e
6
.
C
o
n
f
u
s
io
n
m
atr
i
x
T
ab
le
1
.
C
o
m
p
a
r
ativ
e
an
aly
s
e
s
o
f
s
im
ilar
wo
r
k
s
an
d
t
h
e
p
r
o
p
o
s
ed
h
y
b
r
id
a
r
ch
itectu
r
e
R
e
f
e
r
e
n
c
e
s
C
r
o
p
f
o
c
u
s
D
a
t
a
s
e
t
N
u
mb
e
r
o
f
c
l
a
sse
s
M
o
d
e
l
R
e
s
u
l
t
s
[
5
]
S
e
v
e
r
a
l
N
e
w
p
l
a
n
t
d
i
sea
ses
38
C
N
N
A
c
c
u
r
a
c
y
:
9
7
%
[
6
]
S
e
v
e
r
a
l
P
l
a
n
t
V
i
l
l
a
g
e
38
C
N
N
A
c
c
u
r
a
c
y
:
9
4
.
3
3
%
[
7
]
A
p
p
l
e
AI
-
c
h
a
l
l
e
n
g
e
r
p
l
a
n
t
d
i
s
e
a
se
r
e
c
o
g
n
i
t
i
o
n
6
D
e
n
seN
e
t
-
1
2
1
A
c
c
u
r
a
c
y
:
9
3
.
7
1
%
[
1
3
]
R
i
c
e
a
n
d
m
a
i
z
e
P
l
a
n
t
V
i
l
l
a
g
e
4
V
G
G
N
e
t
w
i
t
h
i
n
c
e
p
t
i
o
n
A
c
c
u
r
a
c
y
:
9
2
%
[
1
6
]
G
r
a
p
e
P
l
a
n
t
V
i
l
l
a
g
e
4
S
e
g
C
N
N
A
c
c
u
r
a
c
y
:
9
3
.
7
5
%
P
r
o
p
o
se
d
w
o
r
k
S
e
v
e
r
a
l
N
e
w
p
l
a
n
t
d
i
sea
ses
38
R
e
sN
e
t
-
50
w
i
t
h
D
e
n
seN
e
t
-
1
2
1
A
c
c
u
r
a
c
y
9
9
.
6
6
%
5.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
in
v
esti
g
ates
th
e
ap
p
licatio
n
o
f
tr
an
s
f
er
lear
n
i
n
g
to
im
p
r
o
v
e
leaf
d
is
ea
s
e
d
etec
tio
n
.
I
t
ex
am
in
es
h
o
w
lev
er
ag
in
g
p
r
e
-
tr
ain
ed
m
o
d
els
ca
n
en
h
an
c
e
th
e
ac
cu
r
ac
y
an
d
ef
f
icien
c
y
o
f
d
is
ea
s
e
d
etec
tio
n
in
p
lan
t
leav
es.
W
e
p
r
o
p
o
s
ed
a
d
ee
p
lear
n
in
g
a
r
ch
itectu
r
e
ca
lled
R
esNet
-
50
an
d
Den
s
eNe
t
-
121
f
o
r
id
e
n
tify
in
g
p
lan
t
d
is
ea
s
e
im
ag
es.
Ou
r
ex
p
er
im
en
t
u
til
ized
th
e
n
ew
p
lan
t
d
is
ea
s
es
d
ataset
f
o
r
class
if
y
in
g
3
8
ty
p
es
o
f
p
la
n
t
leaf
d
is
ea
s
es.
Ou
r
h
y
b
r
id
m
o
d
el
ac
h
iev
ed
an
ac
cu
r
ac
y
r
ate,
s
u
r
p
ass
in
g
r
esu
lts
r
ep
o
r
ted
in
o
th
er
s
tu
d
ies.
Ou
r
ex
p
er
im
en
t
was
u
n
iq
u
e
i
n
th
at
we
co
n
s
id
er
ed
m
u
ltip
le
cr
o
p
s
an
d
u
s
ed
a
m
o
d
if
ied
d
ataset
wi
th
m
o
r
e
class
es.
Ov
er
all,
o
u
r
ex
p
e
r
im
en
t
o
f
f
er
s
an
ef
f
icien
t
ap
p
r
o
a
ch
f
o
r
m
u
lti
-
cr
o
p
d
is
ea
s
e
d
etec
tio
n
.
T
h
e
h
y
b
r
i
d
p
r
e
-
tr
ain
ed
m
o
d
el
d
em
o
n
s
tr
ated
an
im
p
r
ess
iv
e
ac
cu
r
ac
y
r
ate
o
f
9
9
.
6
6
%,
p
r
ec
is
io
n
o
f
9
9
.
6
7
%,
r
ec
all
o
f
9
9
.
6
5
%,
a
n
d
F
1
-
s
co
r
e
o
f
9
9
.
6
6
%
d
em
o
n
s
tr
atin
g
th
e
ef
f
ica
cy
o
f
o
u
r
m
eth
o
d
in
e
n
h
an
ci
n
g
th
e
ac
cu
r
ac
y
o
f
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
E
fficien
t d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
fo
r
en
h
a
n
cin
g
p
la
n
t le
a
f
d
is
ea
s
e
cla
s
s
if
ica
tio
n
(
Mer
o
u
a
B
elmir
)
1119
p
lan
t
d
is
ea
s
e
class
if
icatio
n
en
d
ea
v
o
r
s
.
T
h
e
r
esear
ch
c
o
m
m
u
n
ity
will
b
en
ef
it
g
r
ea
tly
f
r
o
m
th
ese
f
in
d
in
g
s
as
th
ey
en
h
an
ce
th
e
ac
c
u
r
ac
y
an
d
d
ep
en
d
ab
ilit
y
o
f
leaf
d
is
ea
s
e
d
etec
tio
n
.
Ad
d
itio
n
ally
,
th
e
y
d
em
o
n
s
tr
ate
th
e
p
r
o
m
is
e
o
f
AI
-
d
r
iv
e
n
,
s
ca
lab
le
s
o
lu
tio
n
s
,
o
p
en
in
g
th
e
d
o
o
r
f
o
r
m
o
r
e
s
tu
d
y
in
to
t
h
e
d
ev
el
o
p
m
en
t
o
f
ef
f
ec
tiv
e
m
o
n
ito
r
in
g
s
y
s
tem
s
th
at
m
ay
b
e
u
s
ed
in
lar
g
e
-
s
ca
le
f
ar
m
in
g
o
p
er
atio
n
s
.
Ou
r
u
p
co
m
in
g
p
r
o
jects
in
clu
d
e
en
g
ag
in
g
with
p
lan
tatio
n
s
to
in
cr
ea
s
e
th
e
d
iv
er
s
ity
o
f
d
atasets
b
y
u
tili
zin
g
d
r
o
n
es
o
r
R
GB
ca
m
er
a
to
c
o
llect
d
ata.
RE
F
E
R
E
NC
E
S
[
1
]
F
.
Z
h
u
e
t
a
l
.
,
“
J
o
i
n
t
h
e
g
r
e
e
n
t
e
a
m:
I
n
d
u
c
e
r
s
o
f
p
l
a
n
t
i
mm
u
n
i
t
y
i
n
t
h
e
p
l
a
n
t
d
i
se
a
se
su
s
t
a
i
n
a
b
l
e
c
o
n
t
r
o
l
t
o
o
l
b
o
x
,
”
J
o
u
rn
a
l
o
f
Ad
v
a
n
c
e
d
R
e
se
a
rc
h
,
v
o
l
.
5
7
,
p
p
.
1
5
–
4
2
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
a
r
e
.
2
0
2
3
.
0
4
.
0
1
6
.
[
2
]
I
.
B
o
u
a
c
i
d
a
,
B
.
F
a
r
o
u
,
L.
D
j
a
k
h
d
j
a
k
h
a
,
H
.
S
e
r
i
d
i
,
a
n
d
M
.
K
u
r
u
l
a
y
,
“
I
n
n
o
v
a
t
i
v
e
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
f
o
r
c
r
o
ss
-
c
r
o
p
p
l
a
n
t
d
i
s
e
a
se
d
e
t
e
c
t
i
o
n
:
a
g
e
n
e
r
a
l
i
z
e
d
me
t
h
o
d
f
o
r
i
d
e
n
t
i
f
y
i
n
g
u
n
h
e
a
l
t
h
y
l
e
a
v
e
s
,
”
I
n
f
o
rm
a
t
i
o
n
Pro
c
e
ss
i
n
g
i
n
A
g
r
i
c
u
l
t
u
r
e
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
n
p
a
.
2
0
2
4
.
0
3
.
0
0
2
.
[
3
]
A
.
R
a
mc
h
a
r
a
n
,
K
.
B
a
r
a
n
o
w
s
k
i
,
P
.
M
c
C
l
o
s
k
e
y
,
B
.
A
h
m
e
d
,
J.
Le
g
g
,
a
n
d
D
.
P
.
H
u
g
h
e
s
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
i
ma
g
e
-
b
a
se
d
c
a
ss
a
v
a
d
i
s
e
a
se
d
e
t
e
c
t
i
o
n,
”
Fr
o
n
t
i
e
rs i
n
Pl
a
n
t
S
c
i
e
n
c
e
,
v
o
l
.
8
,
p
.
1
8
5
2
,
O
c
t
.
2
0
1
7
,
d
o
i
:
1
0
.
3
3
8
9
/
f
p
l
s.
2
0
1
7
.
0
1
8
5
2
.
[
4
]
E.
M
o
u
p
o
j
o
u
e
t
a
l
.
,
“
F
i
e
l
d
P
l
a
n
t
:
a
d
a
t
a
set
o
f
f
i
e
l
d
p
l
a
n
t
i
m
a
g
e
s
f
o
r
p
l
a
n
t
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
w
i
t
h
d
e
e
p
l
e
a
r
n
i
n
g,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
3
5
3
9
8
–
3
5
4
1
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
2
6
3
0
4
2
.
[
5
]
S
.
B
.
I
man
u
l
l
o
h
,
A
.
R
.
M
u
sl
i
k
h
,
a
n
d
D
.
R
.
I
.
M
.
S
e
t
i
a
d
i
,
“
P
l
a
n
t
d
i
se
a
ses
c
l
a
ss
i
f
i
c
a
t
i
o
n
b
a
se
d
l
e
a
v
e
s
i
m
a
g
e
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
,
”
J
o
u
rn
a
l
o
f
C
o
m
p
u
t
i
n
g
T
h
e
o
r
i
e
s
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
,
n
o
.
1
,
p
p
.
1
–
1
0
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
6
3
3
/
j
c
t
a
.
v
1
i
1
.
8
8
7
7
.
[
6
]
M
.
B
e
l
m
i
r
,
W
.
D
i
f
a
l
l
a
h
,
a
n
d
A
.
G
h
a
z
l
i
,
“
P
l
a
n
t
l
e
a
f
d
i
sea
s
e
p
r
e
d
i
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
i
n
2
0
2
3
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
D
e
c
i
s
i
o
n
A
i
d
S
c
i
e
n
c
e
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
(
D
AS
A)
,
S
e
p
.
2
0
2
3
,
p
p
.
5
3
6
–
5
4
0
,
d
o
i
:
1
0
.
1
1
0
9
/
D
A
S
A
5
9
6
2
4
.
2
0
2
3
.
1
0
2
8
6
6
7
2
.
[
7
]
Y
.
Zh
o
n
g
a
n
d
M
.
Z
h
a
o
,
“
R
e
sea
r
c
h
o
n
d
e
e
p
l
e
a
r
n
i
n
g
i
n
a
p
p
l
e
l
e
a
f
d
i
se
a
se
r
e
c
o
g
n
i
t
i
o
n
,
”
C
o
m
p
u
t
e
rs a
n
d
El
e
c
t
r
o
n
i
c
s
i
n
A
g
r
i
c
u
l
t
u
r
e
,
v
o
l
.
1
6
8
,
p
.
1
0
5
1
4
6
,
J
a
n
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
p
a
g
.
2
0
1
9
.
1
0
5
1
4
6
.
[
8
]
Z.
u
r
R
e
h
m
a
n
e
t
a
l
.
,
“
R
e
c
o
g
n
i
z
i
n
g
a
p
p
l
e
l
e
a
f
d
i
s
e
a
ses
u
si
n
g
a
n
o
v
e
l
p
a
r
a
l
l
e
l
r
e
a
l
‐
t
i
me
p
r
o
c
e
ssi
n
g
f
r
a
mew
o
r
k
b
a
s
e
d
o
n
M
A
S
K
R
C
N
N
a
n
d
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
:
a
n
a
p
p
l
i
c
a
t
i
o
n
f
o
r
smar
t
a
g
r
i
c
u
l
t
u
r
e
,
”
I
ET
I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
v
o
l
.
1
5
,
n
o
.
1
0
,
p
p
.
2
1
5
7
–
2
1
6
8
,
A
u
g
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
4
9
/
i
p
r
2
.
1
2
1
8
3
.
[
9
]
K
.
S
a
n
g
e
e
t
h
a
,
P
.
R
i
ma,
P
.
K
u
mar,
a
n
d
S
.
P
r
e
e
t
h
e
e
s
,
“
Ap
p
l
e
l
e
a
f
d
i
s
e
a
s
e
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
,
”
i
n
2
0
2
2
6
t
h
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
i
n
g
M
e
t
h
o
d
o
l
o
g
i
e
s
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
(
I
C
C
MC
)
,
M
a
r
.
2
0
2
2
,
p
p
.
1
0
6
3
–
1
0
6
7
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
M
C
5
3
4
7
0
.
2
0
2
2
.
9
7
5
3
9
8
5
.
[
1
0
]
T.
A
.
S
a
l
i
h
,
A
.
J.
A
l
i
,
a
n
d
M
.
N
.
A
h
me
d
,
“
D
e
e
p
l
e
a
r
n
i
n
g
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
t
o
d
e
t
e
c
t
a
n
d
c
l
a
ssi
f
y
t
o
m
a
t
o
p
l
a
n
t
l
e
a
f
d
i
s
e
a
se
s,
”
O
ALi
b
,
v
o
l
.
0
7
,
n
o
.
0
5
,
p
p
.
1
–
1
2
,
2
0
2
0
,
d
o
i
:
1
0
.
4
2
3
6
/
o
a
l
i
b
.
1
1
0
6
2
9
6
.
[
1
1
]
D
.
R
.
H
a
mm
o
u
a
n
d
M
.
B
o
u
b
a
k
e
r
,
“
To
m
a
t
o
p
l
a
n
t
d
i
se
a
se
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ssi
f
i
c
a
t
i
o
n
u
s
i
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
a
r
c
h
i
t
e
c
t
u
r
e
s
t
e
c
h
n
o
l
o
g
i
e
s
,
”
N
e
t
w
o
r
k
i
n
g
,
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
a
n
d
S
e
c
u
ri
t
y
:
Pr
o
c
e
e
d
i
n
g
s
o
f
N
I
S
S
2
0
2
1
.
p
p
.
3
3
–
4
4
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
16
-
3
6
3
7
-
0
_
3
.
[
1
2
]
G
.
S
a
k
k
a
r
v
a
r
t
h
i
,
G
.
W
.
S
a
t
h
i
a
n
e
sa
n
,
V
.
S
.
M
u
r
u
g
a
n
,
A
.
J
.
R
e
d
d
y
,
P
.
Ja
y
a
g
o
p
a
l
,
a
n
d
M
.
E
l
si
s
i
,
“
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
o
f
t
o
m
a
t
o
c
r
o
p
d
i
sea
s
e
u
s
i
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
,
”
E
l
e
c
t
r
o
n
i
c
s
,
v
o
l
.
1
1
,
n
o
.
2
1
,
p
.
3
6
1
8
,
N
o
v
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s1
1
2
1
3
6
1
8
.
[
1
3
]
J.
C
h
e
n
,
J
.
C
h
e
n
,
D
.
Z
h
a
n
g
,
Y
.
S
u
n
,
a
n
d
Y
.
A
.
N
a
n
e
h
k
a
r
a
n
,
“
U
s
i
n
g
d
e
e
p
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
f
o
r
i
ma
g
e
-
b
a
sed
p
l
a
n
t
d
i
s
e
a
s
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
,
”
C
o
m
p
u
t
e
rs
a
n
d
El
e
c
t
r
o
n
i
c
s
i
n
A
g
r
i
c
u
l
t
u
r
e
,
v
o
l
.
1
7
3
,
p
.
1
0
5
3
9
3
,
J
u
n
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
a
g
.
2
0
2
0
.
1
0
5
3
9
3
.
[
1
4
]
F
.
K
h
a
n
,
N
.
Z
a
f
a
r
,
M
.
N
.
Ta
h
i
r
,
M
.
A
q
i
b
,
H
.
W
a
h
e
e
d
,
a
n
d
Z.
H
a
r
o
o
n
,
“
A
mo
b
i
l
e
-
b
a
s
e
d
sy
s
t
e
m
f
o
r
m
a
i
z
e
p
l
a
n
t
l
e
a
f
d
i
sea
s
e
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ssi
f
i
c
a
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
F
ro
n
t
i
e
rs
i
n
P
l
a
n
t
S
c
i
e
n
c
e
,
v
o
l
.
1
4
,
p
.
1
0
7
9
3
6
6
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
3
3
8
9
/
f
p
l
s
.
2
0
2
3
.
1
0
7
9
3
6
6
.
[
1
5
]
S
.
G
u
p
t
a
,
J.
V
i
sh
n
o
i
,
a
n
d
A
.
S
.
R
a
o
,
“
D
i
se
a
se
d
e
t
e
c
t
i
o
n
i
n
mai
z
e
p
l
a
n
t
u
si
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
,
”
i
n
2
0
2
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
C
o
m
m
u
n
i
c
a
t
i
o
n
a
n
d
E
l
e
c
t
r
o
n
i
c
s
S
y
s
t
e
m
s
(
I
C
C
E
S
)
,
J
u
n
.
2
0
2
2
,
p
p
.
1
3
3
0
–
1
3
3
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
ES5
4
1
8
3
.
2
0
2
2
.
9
8
3
5
7
3
7
.
[
1
6
]
J.
C
h
e
n
,
J.
C
h
e
n
,
D
.
Z
h
a
n
g
,
Y
.
A
.
N
a
n
e
h
k
a
r
a
n
,
a
n
d
Y
.
S
u
n
,
“
A
c
o
g
n
i
t
i
v
e
v
i
si
o
n
me
t
h
o
d
f
o
r
t
h
e
d
e
t
e
c
t
i
o
n
o
f
p
l
a
n
t
d
i
s
e
a
se
i
m
a
g
e
s,
”
Ma
c
h
i
n
e
V
i
si
o
n
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
3
2
,
n
o
.
1
,
p
.
3
1
,
J
a
n
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s0
0
1
3
8
-
020
-
0
1
1
5
0
-
w.
[
1
7
]
V
.
K
.
S
i
n
g
h
a
n
d
M
.
H
.
K
o
l
e
k
a
r
,
“
D
e
e
p
l
e
a
r
n
i
n
g
e
m
p
o
w
e
r
e
d
C
O
V
I
D
-
1
9
d
i
a
g
n
o
si
s
u
s
i
n
g
c
h
e
s
t
C
T
sca
n
i
ma
g
e
s
f
o
r
c
o
l
l
a
b
o
r
a
t
i
v
e
e
d
g
e
-
c
l
o
u
d
c
o
m
p
u
t
i
n
g
p
l
a
t
f
o
r
m,
”
Mu
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
1
,
n
o
.
1
,
p
p
.
3
–
3
0
,
Ja
n
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
0
4
2
-
021
-
1
1
1
5
8
-
7.
[
1
8
]
L.
Li
,
“
A
p
p
l
i
c
a
t
i
o
n
o
f
d
e
e
p
l
e
a
r
n
i
n
g
i
n
i
ma
g
e
r
e
c
o
g
n
i
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
P
h
y
si
c
s:
C
o
n
f
e
re
n
c
e
S
e
ri
e
s
,
v
o
l
.
1
6
9
3
,
n
o
.
1
,
p
.
0
1
2
1
2
8
,
D
e
c
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
4
2
-
6
5
9
6
/
1
6
9
3
/
1
/
0
1
2
1
2
8
.
[
1
9
]
A
.
S
o
l
a
n
k
i
a
n
d
S
.
P
a
n
d
e
y
,
“
M
u
si
c
i
n
st
r
u
me
n
t
r
e
c
o
g
n
i
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
s,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
4
,
n
o
.
3
,
p
p
.
1
6
5
9
–
1
6
6
8
,
M
a
y
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
4
1
8
7
0
-
0
1
9
-
0
0
2
8
5
-
y.
[
2
0
]
D
.
K
.
S
h
a
r
ma,
M
.
C
h
a
t
t
e
r
j
e
e
,
G
.
K
a
u
r
,
a
n
d
S
.
V
a
v
i
l
a
l
a
,
“
D
e
e
p
l
e
a
r
n
i
n
g
a
p
p
l
i
c
a
t
i
o
n
s
f
o
r
d
i
sea
se
d
i
a
g
n
o
s
i
s,
”
i
n
D
e
e
p
L
e
a
r
n
i
n
g
f
o
r
Me
d
i
c
a
l
A
p
p
l
i
c
a
t
i
o
n
s
w
i
t
h
U
n
i
q
u
e
D
a
t
a
,
E
l
se
v
i
e
r
,
2
0
2
2
,
p
p
.
3
1
–
5
1
.
[
2
1
]
A
.
K
u
l
k
a
r
n
i
,
D
.
C
h
o
n
g
,
a
n
d
F
.
A
.
B
a
t
a
r
se
h
,
“
F
o
u
n
d
a
t
i
o
n
s
o
f
d
a
t
a
i
m
b
a
l
a
n
c
e
a
n
d
s
o
l
u
t
i
o
n
s
f
o
r
a
d
a
t
a
d
e
m
o
c
r
a
c
y
,
”
i
n
D
a
t
a
D
e
m
o
c
r
a
c
y
,
El
s
e
v
i
e
r
,
2
0
2
0
,
p
p
.
8
3
–
1
0
6
.
[
2
2
]
O
.
J.
A
w
u
j
o
o
l
a
,
F
.
N
.
O
g
w
u
e
l
e
k
a
,
P
.
O
.
O
d
i
o
n
,
A
.
E.
A
w
u
j
o
o
l
a
,
a
n
d
O
.
R
.
A
d
e
l
e
g
a
n
,
“
G
e
n
o
m
i
c
d
a
t
a
sc
i
e
n
c
e
s
y
s
t
e
ms
o
f
P
r
e
d
i
c
t
i
o
n
a
n
d
p
r
e
v
e
n
t
i
o
n
o
f
p
n
e
u
m
o
n
i
a
f
r
o
m
c
h
e
s
t
X
-
r
a
y
i
m
a
g
e
s
u
si
n
g
a
t
w
o
-
c
h
a
n
n
e
l
d
u
a
l
-
st
r
e
a
m 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
D
a
t
a
S
c
i
e
n
c
e
f
o
r
G
e
n
o
m
i
c
s
,
E
l
se
v
i
e
r
,
2
0
2
3
,
p
p
.
2
1
7
–
2
2
8
.
[
2
3
]
R
.
N
a
n
d
i
a
n
d
M
.
M
u
l
i
ma
n
i
,
“
D
e
t
e
c
t
i
o
n
o
f
C
O
V
I
D
-
1
9
f
r
o
m
X
-
r
a
y
s
u
si
n
g
h
y
b
r
i
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
s
,
”
Re
s
e
a
r
c
h
o
n
B
i
o
m
e
d
i
c
a
l
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
3
7
,
n
o
.
4
,
p
p
.
6
8
7
–
6
9
5
,
D
e
c
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s
4
2
6
0
0
-
0
2
1
-
0
0
1
8
1
-
0.
[
2
4
]
K
.
H
e
,
X
.
Z
h
a
n
g
,
S
.
R
e
n
,
a
n
d
J
.
S
u
n
,
“
D
e
e
p
r
e
si
d
u
a
l
l
e
a
r
n
i
n
g
f
o
r
i
ma
g
e
r
e
c
o
g
n
i
t
i
o
n,
”
i
n
2
0
1
6
I
EEE
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
si
o
n
a
n
d
Pa
t
t
e
rn
Re
c
o
g
n
i
t
i
o
n
(
C
V
P
R)
,
Ju
n
.
2
0
1
6
,
p
p
.
7
7
0
–
7
7
8
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
.
2
0
1
6
.
9
0
.
[
2
5
]
F
.
T
u
j
,
J.
F
a
r
i
a
,
M
.
B
i
n
M
o
i
n
,
P
.
D
e
b
n
a
t
h
,
A
.
I
.
F
a
h
i
m,
a
n
d
F
.
M
.
S
h
a
h
,
“
Ex
p
l
a
i
n
a
b
l
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
f
o
r
r
e
t
i
n
a
l
f
u
n
d
u
s
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
n
d
c
u
t
t
i
n
g
-
e
d
g
e
s
e
g
me
n
t
a
t
i
o
n
m
o
d
e
l
s
f
o
r
r
e
t
i
n
a
l
b
l
o
o
d
v
e
sse
l
s
f
r
o
m
f
u
n
d
u
s
i
m
a
g
e
s
,
”
a
rXi
v
p
re
p
r
i
n
t
a
rXi
v
:
2
4
0
5
.
0
7
3
3
8
,
2
0
2
4
,
d
o
i
:
1
0
.
4
8
5
5
0
/
a
r
X
i
v
.
2
4
0
5
.
0
7
3
3
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
a
r
y
20
2
5
:
1
11
2
-
1
1
20
1120
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Mer
o
u
a
Be
l
m
ir
is
a
P
h
.
D
.
stu
d
e
n
t
i
n
a
rti
ficia
l
in
telli
g
e
n
c
e
a
n
d
it
s
a
p
p
li
c
a
ti
o
n
s
a
t
Tah
ri
M
o
h
a
m
m
e
d
U
n
iv
e
rsit
y
(
UTM
B),
Al
g
e
ria.
S
h
e
re
c
e
iv
e
d
h
e
r
b
a
c
h
e
l
o
r
’
s
a
n
d
m
a
ste
r
’
s
d
e
g
re
e
s
in
c
o
m
p
u
ter
sc
ien
c
e
in
2
0
2
0
a
n
d
in
a
rti
ficia
l
in
telli
g
e
n
c
e
a
n
d
d
e
c
isio
n
m
a
k
in
g
in
2
0
2
2
re
sp
e
c
ti
v
e
ly
,
fr
o
m
Tah
ri
M
o
h
a
m
m
e
d
Un
iv
e
rsity
o
f
Be
c
h
a
r,
A
lg
e
ria.
He
r
c
u
rre
n
t
a
re
a
o
f
in
tere
st
is
sm
a
rt
fa
rm
in
g
a
p
p
li
c
a
ti
o
n
u
si
n
g
a
rti
ficia
l
in
tell
ig
e
n
c
e
tec
h
n
i
q
u
e
s
e
sp
e
c
ially
m
a
c
h
in
e
lea
rn
i
n
g
a
n
d
d
e
e
p
lea
rn
in
g
,
ima
g
e
p
ro
c
e
ss
in
g
,
a
n
d
ima
g
e
a
n
a
ly
s
is.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
e
lmir.me
ro
u
a
@u
n
i
v
-
b
e
c
h
a
r.
d
z
.
Dr
.
Wa
fa
Difa
l
la
h
is
a
lec
tu
re
r
in
c
o
m
p
u
ter
sc
ien
c
e
a
t
Tah
ri
M
o
h
a
m
e
d
Un
iv
e
rsity
o
f
Be
c
h
a
r,
Alg
e
ria.
S
h
e
re
c
e
iv
e
d
h
e
r
b
a
c
h
e
lo
r
’
s
a
n
d
m
a
ste
r
’
s
d
e
g
re
e
s
in
M
a
th
e
m
a
ti
c
s
a
n
d
C
o
m
p
u
ter
S
c
ien
c
e
in
2
0
0
9
a
n
d
i
n
C
o
m
p
u
ter
S
y
ste
m
a
n
d
Ne
two
rk
s
in
2
0
1
1
re
sp
e
c
ti
v
e
ly
,
fro
m
Tah
ri
M
o
h
a
m
m
e
d
Un
i
v
e
rsity
o
f
Be
c
h
a
r,
Alg
e
ri
a
.
S
h
e
re
c
e
iv
e
d
h
e
r
P
h
.
D
.
in
in
telli
g
e
n
t
c
o
m
p
u
ter
sy
ste
m
fro
m
Tah
ri
M
o
h
a
m
e
d
Un
iv
e
rsit
y
o
f
B
e
c
h
a
r
Alg
e
ria
in
2
0
1
8
.
He
r
c
u
rre
n
t
re
se
a
rc
h
in
tere
sts
a
re
wa
ter
o
p
ti
m
iza
ti
o
n
,
sm
a
rt
fa
rm
in
g
a
n
d
se
c
u
rit
y
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
d
i
fa
ll
a
h
.
wa
fa
@u
n
i
v
-
b
e
c
h
a
r.
d
z
.
Dr
.
Abd
e
lk
a
d
e
r
G
h
a
z
li
is
a
lec
tu
re
r
a
t
Tah
ri
M
o
h
a
m
m
e
d
Un
iv
e
rsit
y
in
Be
c
h
a
r,
Alg
e
ria.
He
o
b
tai
n
e
d
a
n
e
n
g
in
e
e
rin
g
d
ip
l
o
m
a
in
c
o
m
p
u
ter
sc
ien
c
e
fro
m
Un
iv
e
rsity
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
l
o
g
y
i
n
Ora
n
(UST
O),
Alg
e
ria
i
n
2
0
0
5
.
He
o
b
t
a
in
e
d
a
tea
c
h
in
g
d
ip
l
o
m
a
i
n
c
o
m
p
u
ter
sc
ien
c
e
fro
m
U
n
iv
e
rs
it
y
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
l
o
g
y
i
n
Ora
n
(UST
O),
Alg
e
ria
i
n
2
0
0
9
.
His
re
se
a
rc
h
in
tere
sts
a
re
a
rti
ficia
l
in
telli
g
e
n
c
e
,
m
o
b
il
e
n
e
t
wo
rk
s,
c
y
b
e
r
se
c
u
rit
y
,
b
ra
i
n
c
o
m
p
u
ter
in
terfa
c
e
,
a
n
d
e
m
b
e
d
d
e
d
sy
ste
m
s.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
g
h
a
z
e
k
@g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.