I
n
t
e
r
n
at
ion
al
Jou
r
n
a
l
of
I
n
f
o
r
m
at
ics
an
d
Com
m
u
n
icat
ion
T
e
c
h
n
ol
ogy
(
I
J
-
I
CT
)
Vo
l
.
1
5
,
N
o
.
3
,
S
e
p
t
e
m
b
e
r
20
2
6
,
pp.
1303
~
1312
I
S
S
N:
2252
-
8776
,
DO
I
:
10
.
11591/i
ji
c
t
.
v
1
5
i
3
.
pp
13
03
-
1312
1303
Jou
r
n
al
h
o
m
e
page
:
ht
tp:
//
ij
ict
.
iaes
c
or
e
.
c
om
A
t
t
e
n
t
io
n
-
e
n
h
an
c
e
d
V
GG
-
16
ar
c
h
ite
c
t
u
r
e
f
or
p
r
e
c
is
io
n
w
e
e
d
d
e
t
e
c
t
io
n
i
n
w
h
e
at
f
ie
ld
s
Ak
an
k
s
h
a
B
od
h
al
e
1
,
S
e
e
m
a
Ve
r
m
a
2
,
Ais
h
wa
r
y
B
od
h
al
e
3
1
D
e
pa
r
tm
e
nt
of
C
o
mpu
te
r
S
c
ie
n
c
e
E
ng
in
e
e
r
in
g,
B
a
na
s
th
a
li
V
id
y
a
pi
th
, R
a
ja
s
th
a
n, I
ndi
a
2
D
e
pa
r
tm
e
nt
of
E
l
e
c
tr
i
c
a
l
a
nd
E
l
e
c
tr
o
ni
c
s
E
ngi
n
e
e
r
in
g,
N
a
ti
o
na
l
I
ns
ti
tu
te
of
T
e
c
hn
ic
a
l
T
e
a
c
h
e
r
s
'
T
r
a
in
in
g a
nd R
e
s
e
a
r
c
h
(
N
I
T
T
T
R
)
B
h
o
pa
l,
I
ndi
a
3
D
e
pa
r
tm
e
nt
of
B
us
in
e
s
s
A
dmi
ni
s
tr
a
ti
o
n
,
U
ni
ve
r
s
it
y
of
t
h
e
C
u
mbe
r
la
nds
,
L
a
s
V
e
ga
s
, U
ni
te
d S
ta
te
s
Ar
t
ic
l
e
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
i
ve
d
J
u
l
24,
2025
R
e
vi
s
e
d
A
pr
24,
2026
A
c
c
e
pt
e
d
M
a
y
21,
2026
T
h
i
s
s
t
u
d
y
d
el
i
n
e
at
e
s
f
i
v
e
ad
v
an
ced
co
n
v
o
l
u
t
i
o
n
a
l
n
eu
ral
n
e
t
w
o
r
k
d
e
s
i
g
n
s
t
h
at
em
p
l
o
y
d
ee
p
l
e
arn
i
n
g
fo
r
t
h
e
c
l
as
s
i
fi
c
at
i
o
n
o
f
w
h
e
a
t
w
eed
s
.
T
h
e
d
at
as
e
t
c
o
n
s
i
s
t
s
o
f
g
r
ey
s
c
al
e
p
h
o
t
o
s
t
ak
e
n
i
n
ag
r
i
c
u
l
t
u
ral
fi
e
l
d
s
,
e
n
h
an
ced
w
i
t
h
RG
B
l
i
g
h
t
i
n
g
a
n
d
c
r
o
p
p
e
d
t
o
2
5
6
×
2
5
6
×
3
p
i
x
e
l
s
.
N
o
r
m
a
l
i
zat
i
o
n
,
b
at
ch
-
w
i
s
e
au
g
me
n
t
at
i
o
n
,
c
o
n
t
ras
t
s
t
re
t
c
h
i
n
g
,
o
r
h
i
s
t
o
g
ra
m
e
q
u
al
i
z
at
i
o
n
w
e
r
e
al
l
p
art
o
f
t
h
e
p
re
p
ro
ce
s
s
i
n
g
t
h
at
i
m
p
ro
v
ed
p
i
c
t
u
r
e
q
u
al
i
t
y
a
n
d
m
o
d
e
l
l
e
ar
n
i
n
g
e
ffi
ci
e
n
cy
.
T
h
e
s
e
e
n
h
an
ceme
n
t
s
d
ev
e
l
o
p
ed
f
e
at
u
r
e
e
x
t
rac
t
i
o
n
b
y
i
n
c
r
e
as
i
n
g
p
i
c
t
u
r
e
co
n
t
ras
t
an
d
h
o
mo
g
en
e
i
t
y
.
A
V
G
G
1
6
-
b
a
s
e
d
mo
d
e
l
w
i
t
h
f
u
rt
h
e
r
Co
n
v
2
D
l
a
y
ers
an
d
s
p
at
i
al
at
t
e
n
t
i
o
n
i
s
o
n
e
o
f
t
h
e
f
i
v
e
mo
d
e
l
s
r
eco
mme
n
d
e
d
.
A
n
ad
d
i
t
i
o
n
a
l
d
e
s
i
g
n
,
i
n
s
p
i
red
b
y
R
e
s
N
e
t
5
0
t
h
at
ap
p
l
i
e
s
re
s
i
d
u
al
b
l
o
ck
s
an
d
w
o
rl
d
w
i
d
e
a
v
e
rag
e
p
o
o
l
i
n
g
.
A
h
y
b
ri
d
RN
N
i
n
t
e
g
rat
e
d
ICN
A
-
C
N
N
o
r
L
ST
M
fo
r
s
p
at
i
al
-
t
em
p
o
ral
co
n
t
e
n
t
l
e
arn
i
n
g
i
s
an
o
t
h
e
r.
L
as
t
l
y
,
In
ce
p
t
i
o
n
R
e
s
N
e
t
V
2
i
s
en
h
a
n
ce
d
w
i
t
h
CN
N
l
a
y
e
rs
an
d
a
q
u
e
r
y
-
k
ey
at
t
en
t
i
o
n
mec
h
a
n
i
s
m
.
W
e
u
s
e
d
t
h
e
A
d
a
m
o
p
t
i
m
i
z
e
r
an
d
c
at
eg
o
r
i
c
al
c
r
o
s
s
-
e
n
t
ro
p
y
t
o
e
s
t
i
m
at
e
t
h
e
l
o
s
s
t
h
ro
u
g
h
o
u
t
t
h
e
t
rai
n
i
n
g
fo
r
t
h
e
ro
u
n
d
.
A
t
t
h
e
e
n
d
,
al
l
t
h
e
m
o
d
e
l
s
h
ad
R
eL
U
ac
t
i
v
at
i
o
n
,
b
at
ch
n
o
r
m
a
l
i
zat
i
o
n
,
Max
Po
o
l
i
n
g
2
D
,
an
d
a
d
ro
p
o
u
t
ch
arg
e
o
f
0
.
5
t
o
k
ee
p
t
h
em
fro
m
o
v
e
rfi
t
t
i
n
g
.
U
p
o
n
e
v
a
l
u
at
i
n
g
t
h
e
i
r
p
e
rfo
r
m
a
n
ce
b
as
ed
o
n
acc
u
ra
cy
,
r
ec
a
l
l
,
rel
i
ab
i
l
i
t
y
,
an
d
t
r
ai
n
i
n
g
l
o
s
s
,
V
G
G
1
6
e
me
r
g
e
d
as
t
h
e
s
t
an
d
o
u
t
mo
d
el
,
a
c
h
i
e
v
i
n
g
an
i
m
p
re
s
s
i
v
e
9
9
.
0
7
%
p
re
c
i
s
i
o
n
an
d
a
rem
ar
k
ab
l
y
l
o
w
t
rai
n
i
n
g
l
o
s
s
o
f
j
u
s
t
0
.
0
0
7
9
.
T
h
e
s
t
u
d
y
d
e
t
e
r
m
i
n
ed
t
h
at
V
G
G
1
6
i
s
t
h
e
o
p
t
i
m
al
ch
o
i
ce
fo
r
p
reci
s
e
w
h
e
at
–
w
ee
d
c
at
eg
o
r
i
zat
i
o
n
b
ec
au
s
e
t
o
i
t
s
s
u
p
e
ri
o
r
g
en
e
ral
i
zab
i
l
i
t
y
a
n
d
a
ccu
ra
cy
.
K
e
y
w
o
r
d
s
:
De
e
p
-
l
e
a
r
ni
ng
I
C
NA
-
C
NN
I
n
c
e
pt
i
o
n
R
e
s
Ne
t
V2
M
a
c
hi
ne
-
l
e
a
r
ni
n
g
VG
G16
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
cen
s
e.
C
or
r
e
s
pon
din
g
A
u
th
or
:
A
ka
n
k
s
h
a
B
o
dh
a
l
e
De
pa
r
t
m
e
n
t
o
f
C
o
m
put
e
r
S
c
i
e
n
c
e
E
n
g
i
ne
e
r
i
n
g,
B
a
n
a
s
t
h
a
li
V
i
d
y
a
p
i
t
h
R
a
j
a
s
t
h
a
n
,
I
n
d
i
a
E
m
a
i
l
:
a
ka
ns
h
a
2611
@
g
m
a
il
.
c
o
m
1.
I
NT
RODU
C
T
I
ON
P
a
r
t
i
c
u
l
a
r
l
y
i
n
pr
e
c
i
s
i
o
n
f
a
r
m
i
ng,
t
h
e
us
e
s
o
f
a
r
t
i
f
ic
i
a
l
i
n
t
e
ll
i
ge
n
c
e
(
AI
)
a
n
d
m
a
c
hi
ne
l
e
a
r
ni
ng
i
n
t
h
e
f
a
r
m
i
ng
s
e
c
t
o
r
h
a
ve
r
e
s
u
l
t
e
d
i
n
r
e
v
o
l
ut
i
o
na
r
y
s
hi
f
t
s
.
W
e
e
d
c
o
n
tr
o
l
,
whi
c
h
gr
e
a
t
l
y
r
e
duc
e
s
a
gr
i
c
u
l
t
u
r
a
l
y
i
e
l
d
a
n
d
qua
li
t
y
,
i
s
o
n
e
o
f
t
h
e
bi
gge
s
t
pr
o
bl
e
m
s
i
n
c
o
n
t
e
m
po
r
a
r
y
f
a
r
m
i
ng.
M
e
c
h
a
ni
c
a
l
r
e
m
o
v
a
l
a
n
d
h
u
m
a
n
o
b
s
e
r
v
a
t
i
o
n
,
t
h
e
t
w
o
m
a
i
ns
t
a
y
s
o
f
t
r
a
d
i
t
i
o
na
l
we
e
d
i
d
e
n
t
i
f
i
c
a
t
i
o
n
pr
o
c
e
dur
e
s
,
a
r
e
b
ot
h
t
e
di
o
us
a
n
d
pr
o
n
e
to
h
u
m
a
n
mi
s
t
a
ke
[
1
]
–
[
4
]
.
T
r
a
di
t
i
o
n
a
l
m
e
t
h
o
ds
do
n
'
t
wo
r
k
f
o
r
m
a
k
i
ng
de
c
i
s
i
o
ns
i
n
r
e
a
l
t
i
m
e
o
r
f
o
r
l
a
r
ge
a
r
e
a
s
o
f
f
a
r
m
i
ng
[
5
]
–
[
9
]
.
M
a
ny
p
e
o
p
l
e
a
r
e
e
n
t
h
us
i
a
s
t
i
c
a
b
o
ut
s
m
a
r
t
f
a
r
m
i
ng
a
n
d
c
o
m
put
e
r
vi
s
i
o
n
a
n
d
c
o
m
put
e
r
vi
s
i
o
n
t
e
c
h
ni
que
s
f
o
r
a
u
to
m
a
t
e
d
c
r
o
p
a
n
d
we
e
d
c
l
a
s
s
if
i
c
a
t
i
o
n
due
to
t
h
e
i
nc
r
e
a
s
i
ng
de
m
a
n
d
f
o
r
s
u
s
t
a
i
na
bl
e
a
gr
i
c
u
l
t
ur
e
a
n
d
t
h
e
g
l
o
b
a
l
t
r
e
n
d
to
wa
r
ds
a
u
to
m
a
t
i
o
n
[
10
]
–
[1
5
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
3
,
S
e
p
te
m
b
e
r
20
2
6
:
1303
-
1312
1304
W
e
e
d
i
nf
e
s
t
a
t
i
o
ns
c
a
n
s
e
v
e
r
e
ly
d
im
i
n
i
s
h
c
r
o
p
qu
a
l
i
t
y
a
n
d
qua
n
t
i
t
y
i
f
n
o
t
pr
o
m
pt
l
y
d
i
s
c
o
v
e
r
e
d
a
n
d
c
o
n
t
r
o
l
l
e
d
;
t
hi
s
i
s
e
s
pe
c
i
a
ll
y
t
r
ue
o
f
whe
a
t
,
a
s
t
a
p
l
e
c
r
o
p
a
r
o
un
d
t
h
e
wo
r
l
d.
I
n
o
r
de
r
to
m
i
n
i
mi
z
e
t
he
us
e
o
f
h
e
r
bi
c
i
de
s
,
l
e
s
s
e
n
t
h
e
e
nvi
r
o
nm
e
n
t
a
l
e
f
f
e
c
t
,
a
n
d
i
nc
r
e
a
s
e
c
r
o
p
h
e
a
l
t
h
,
i
t
i
s
c
r
uc
i
a
l
t
o
de
t
e
c
t
we
e
ds
a
m
o
n
g
wh
e
a
t
c
r
o
ps
e
a
r
l
y
a
n
d
a
c
c
ur
a
t
e
l
y
.
A
ut
o
m
a
t
e
d
vi
s
u
a
l
c
l
a
s
s
if
i
c
a
t
i
o
n
s
y
s
t
e
m
s
t
h
a
t
us
e
de
e
p
l
e
a
r
ni
ng
a
r
e
a
f
a
s
t
e
r
a
n
d
c
h
e
a
pe
r
wa
y
t
o
d
o
i
ns
pe
c
t
i
o
n
s
t
h
a
n
ha
vi
n
g
pe
o
p
l
e
do
t
h
e
m
.
T
h
e
s
e
s
y
s
t
e
m
s
s
t
ud
y
f
i
e
l
d
i
m
a
ge
s
to
r
e
c
o
gni
z
e
r
e
gi
o
n
s
w
h
e
r
e
whe
a
t
pl
a
n
t
s
r
e
m
a
i
n
h
e
a
l
t
hy
a
n
d
f
i
e
l
ds
wh
e
r
e
we
e
d
gr
o
w
t
h
i
s
o
n
t
h
e
r
i
s
e
.
C
o
n
v
o
l
ut
i
o
n
a
l
ne
ur
a
l
ne
t
wor
ks
(
C
NN
s
)
h
a
v
e
f
u
n
da
m
e
n
t
a
l
ly
r
e
v
o
l
u
t
i
o
ni
z
e
d
t
hi
s
do
m
a
i
n
by
e
na
bl
i
ng
m
o
de
l
s
to
i
de
n
t
i
f
y
s
p
a
t
i
a
l
pa
t
t
e
r
n
s
a
n
d
e
x
t
r
a
c
t
i
n
t
r
i
c
a
t
e
vi
s
ua
l
f
e
a
t
ur
e
s
f
r
o
m
a
gr
i
c
u
l
t
ur
a
l
p
h
o
to
g
r
a
ph
s
[
16
]
–
[
22
]
.
C
NN
s
a
n
d
ot
h
e
r
de
e
p
l
e
a
r
ni
ng
a
r
c
hi
t
e
c
t
ur
e
s
a
r
e
e
x
c
e
pt
i
o
n
a
ll
y
e
f
f
e
c
t
i
v
e
a
t
im
a
ge
c
l
a
s
s
i
f
i
c
a
t
i
o
n
b
e
c
a
u
s
e
t
h
e
y
a
ut
o
n
o
m
o
us
ly
lea
r
n
r
e
l
e
va
n
t
pr
o
pe
r
t
i
e
s
w
i
t
h
o
ut
r
e
qui
r
i
ng
m
a
n
u
a
l
h
u
m
a
n
i
n
t
e
r
v
e
n
t
i
o
n
.
T
h
e
pr
e
s
e
n
c
e
o
f
b
a
c
kgr
o
un
d
n
o
i
s
e
,
v
a
r
i
a
t
i
o
n
s
i
n
li
g
h
t
i
n
g,
o
r
i
m
pe
d
im
e
n
t
s
m
i
g
h
t
n
e
ve
r
t
h
e
l
e
s
s
c
a
us
e
a
gr
i
c
u
l
t
ur
a
l
p
i
c
t
ur
e
s
to
o
c
c
a
s
i
o
n
a
ll
y
d
i
s
p
l
a
y
i
r
r
e
gu
l
a
r
i
t
i
e
s
.
As
a
r
e
s
u
l
t
,
hi
g
h
-
qua
l
i
t
y
im
a
g
e
s
a
n
d
r
e
l
i
a
bl
e
m
o
de
l
t
r
a
i
ni
ng
de
pe
n
d
o
n
e
f
f
e
c
t
i
ve
pr
e
pr
o
c
e
s
s
in
g
t
e
c
hni
qu
e
s
.
No
r
m
a
l
i
z
a
t
i
o
n
s
,
c
o
n
t
r
a
s
t
s
tr
e
t
c
hi
n
g,
a
n
d
hi
s
t
o
gr
a
m
e
qua
l
i
z
a
t
i
o
n
a
r
e
c
o
m
m
o
n
wa
y
s
to
s
t
a
n
d
a
r
d
i
s
e
i
m
a
g
e
da
t
a
a
n
d
i
m
pr
o
v
e
c
r
uc
i
a
l
vi
s
ua
l
s
i
g
na
l
s
.
T
h
i
s
a
l
l
o
w
s
f
o
r
b
e
tt
e
r
l
e
a
r
ni
ng
a
n
d
ge
ne
r
a
l
i
s
a
t
i
o
n
[
23
]
–
[
27
]
.
T
h
e
s
e
m
o
de
l
s
a
ppe
a
r
c
a
pa
bl
e
o
f
d
i
s
t
i
n
gu
i
s
hi
ng
be
twe
e
n
v
a
r
i
o
us
we
e
d
s
pe
c
i
e
s
by
r
e
c
o
gni
z
i
ng
s
u
b
t
l
e
c
h
a
n
g
e
s
i
n
l
e
a
f
s
ha
pe
,
c
o
l
o
r
a
t
i
o
n
,
a
n
d
t
e
x
t
ur
e
a
c
r
o
s
s
v
a
r
i
o
us
e
nvi
r
o
nm
e
n
t
a
l
c
o
n
d
i
t
i
o
n
s
.
M
o
r
e
o
v
e
r
,
a
gr
i
c
u
l
t
ur
a
l
i
m
a
ge
r
y
c
a
n
o
f
t
e
n
be
i
n
c
o
n
s
i
s
t
e
n
t
du
e
to
b
a
c
kgr
o
un
d
c
l
ut
t
e
r
,
c
h
a
n
g
i
ng
s
unli
g
h
t
c
o
n
d
i
t
i
o
n
s
,
o
r
phy
s
i
c
a
l
o
b
s
t
r
uc
t
i
o
n
s
.
A
c
hi
e
vi
ng
hi
g
h
-
qua
l
i
t
y
im
a
ge
s
a
n
d
t
r
a
i
ni
ng
r
o
b
us
t
m
o
de
l
s
de
m
a
n
ds
e
f
f
e
c
t
i
ve
pr
e
pr
o
c
e
s
s
i
ng
m
e
t
h
o
ds
.
C
o
m
m
o
n
a
ppr
o
a
c
h
s
uc
h
a
s
n
o
r
m
a
li
z
a
t
i
o
n
,
c
o
n
t
r
a
s
t
i
m
pr
o
v
e
m
e
n
t
,
a
n
d
hi
s
to
gr
a
m
e
qua
l
i
z
a
t
i
o
n
im
pr
o
v
e
s
im
a
ge
u
ni
f
o
r
mi
t
y
a
n
d
a
m
p
li
f
y
e
s
s
e
n
t
i
a
l
vi
s
ua
l
a
tt
r
i
b
ut
e
s
.
T
hi
s
f
a
c
il
i
t
a
t
e
s
be
tt
e
r
f
e
a
t
ur
e
e
x
t
r
a
c
t
i
o
n
a
n
d
kn
o
w
l
e
dge
tr
a
n
s
f
e
r
,
whi
c
h
i
s
c
r
i
t
i
c
a
l
f
o
r
di
f
f
e
r
e
n
t
i
a
t
i
n
g
wh
e
a
t
f
r
o
m
we
e
d
s
pe
c
i
e
s
.
C
h
o
i
c
e
s
r
e
ga
r
d
i
n
g
da
t
a
s
e
t
s
e
l
e
c
t
i
o
n
,
pr
e
pr
o
c
e
s
s
i
n
g
wo
r
kf
l
o
ws
,
m
o
de
l
de
s
i
g
n
,
a
n
d
t
r
a
i
ni
ng
pr
oto
c
o
l
s
s
i
g
nif
i
c
a
n
t
ly
i
n
f
l
u
e
n
c
e
t
h
e
f
i
na
l
c
l
a
s
s
if
i
c
a
t
i
o
n
pe
r
f
o
r
m
a
n
c
e
.
P
ubli
c
ly
a
v
a
il
a
bl
e
r
e
s
o
ur
c
e
s
l
i
k
e
t
h
e
wh
e
a
t
-
we
e
d
da
t
a
s
e
t
e
n
a
bl
e
r
e
s
e
a
r
c
he
r
s
to
tr
a
i
n
a
n
d
e
v
a
l
ua
t
e
m
a
c
hi
ne
l
e
a
r
ni
ng
m
o
de
l
s
.
T
h
e
s
e
da
t
a
s
e
t
s
t
y
p
i
c
a
ll
y
i
nc
l
ude
R
GB
f
i
e
l
d
i
m
a
ge
s
c
a
pt
ur
e
d
un
de
r
d
i
v
e
r
s
e
we
a
t
h
e
r
c
o
n
d
i
t
i
o
n
s
,
r
e
f
l
e
c
t
i
n
g
t
h
e
c
o
m
p
l
e
xi
t
y
o
f
r
e
a
l
-
w
o
r
l
d
a
gr
i
c
u
l
t
ur
a
l
e
nvi
r
o
nm
e
n
t
s
.
Ne
v
e
r
t
h
e
l
e
s
s
,
s
uc
h
da
t
a
s
e
t
s
o
f
t
e
n
s
u
f
f
e
r
f
r
o
m
c
l
a
s
s
im
ba
l
a
n
c
e
s
—
f
o
r
i
n
s
t
a
n
c
e
,
c
o
n
t
a
i
ni
n
g
f
a
r
m
o
r
e
we
e
d
i
m
a
ge
s
t
h
a
n
h
e
a
l
t
hy
c
r
o
p
s
a
m
p
l
e
s
.
A
ddr
e
s
s
i
ng
t
hi
s
d
i
s
pa
r
i
t
y
i
s
e
s
s
e
n
t
i
a
l
t
o
pr
e
v
e
n
t
bi
a
s
e
d
m
o
de
l
o
u
t
pu
t
s
a
n
d
e
n
s
ur
e
r
e
l
i
a
b
l
e
c
l
a
s
s
i
f
i
c
a
t
i
o
n
a
c
r
o
s
s
a
l
l
c
a
t
e
go
r
i
e
s
.
H
y
pe
r
pa
r
a
m
e
t
e
r
s
l
i
ke
l
e
a
r
ni
ng
r
a
t
e
s
,
dr
o
p
o
u
t
r
a
t
i
o
s
,
a
c
t
i
v
a
t
i
o
n
f
u
n
c
t
i
o
n
s
,
a
n
d
o
pt
i
m
ize
r
t
y
p
e
s
n
e
e
d
c
a
r
e
f
u
l
t
uni
ng
f
o
r
e
a
c
h
m
o
de
l
.
T
he
A
da
m
o
pt
i
mi
z
e
r
i
s
b
r
o
a
d
l
y
pr
e
f
e
r
r
e
d
f
o
r
c
l
a
s
s
if
i
c
a
t
i
o
n
t
a
s
ks
due
to
i
t
s
e
f
f
i
c
i
e
n
t
l
e
a
r
ni
ng
a
n
d
r
a
p
i
d
c
o
nv
e
r
ge
n
c
e
.
Ac
t
i
v
a
t
i
o
n
f
u
nc
t
i
o
ns
s
uc
h
a
s
r
e
c
t
i
f
i
e
d
li
ne
a
r
uni
t
(
R
e
L
U
)
e
nh
a
nc
e
t
r
a
i
ni
ng
by
i
n
t
r
o
duc
i
n
g
n
o
n
-
l
i
ne
a
r
i
t
y
a
n
d
mi
t
i
ga
t
i
n
g
t
h
e
v
a
ni
s
hi
ng
gr
a
d
i
e
n
t
pr
o
b
l
e
m
.
Out
pu
t
-
s
t
a
ge
s
o
f
t
m
a
x
l
a
y
e
r
s
c
o
nv
e
r
t
r
a
w
s
c
or
e
s
i
n
t
o
p
r
o
b
a
bil
i
t
y
d
i
s
t
r
i
b
ut
i
o
ns
,
e
n
a
bli
ng
m
u
l
t
i
-
c
l
a
s
s
c
a
t
e
g
o
r
i
z
a
t
i
o
n
.
L
o
s
s
f
u
nc
t
i
o
ns
s
uc
h
a
s
c
a
t
e
go
r
i
c
a
l
c
r
o
s
s
-
e
n
t
r
o
py
gu
i
de
o
pt
i
m
i
z
a
t
i
o
n
by
pe
na
l
i
z
i
ng
pr
e
d
i
c
t
i
o
ns
a
c
c
o
r
di
n
g
to
t
h
e
i
r
de
vi
a
t
i
o
n
f
r
o
m
tr
ue
l
a
b
e
l
s
.
A
r
i
go
r
o
us
da
t
a
-
s
p
l
i
t
t
i
n
g
s
t
r
a
t
e
g
y
,
t
y
p
i
c
a
ll
y
i
nv
o
l
v
i
ng
s
t
r
a
t
i
f
i
e
d
tr
a
i
n
–
t
e
s
t
di
vi
s
i
o
n
s
,
i
s
e
s
s
e
n
t
i
a
l
f
o
r
t
h
o
r
o
ughl
y
va
l
idat
i
n
g
m
o
de
l
e
nh
a
nc
e
m
e
n
t
a
n
d
e
ns
ur
i
n
g
ge
n
e
r
a
li
z
a
t
i
o
n
.
2.
L
I
T
E
RA
T
UR
E
RE
VI
E
W
T
h
i
s
s
e
c
t
i
o
n
di
s
c
us
s
e
s
t
h
e
m
o
s
t
pe
r
t
i
n
e
n
t
e
a
r
l
i
e
r
s
tudy
r
e
l
a
ted
to
th
i
s
s
ubj
e
c
t
,
a
s
s
u
m
m
a
r
i
z
e
d
i
n
T
a
bl
e
1
.
S
i
n
g
h
s
ugge
s
t
s
c
o
m
pr
i
s
e
s
upp
l
e
m
e
n
t
a
r
y
c
l
im
a
t
i
c
a
n
d
to
p
o
g
r
a
phi
c
a
l
i
n
f
o
r
m
a
t
i
o
n
a
l
o
n
gs
i
de
hi
g
h
-
r
e
s
o
l
ut
i
o
n
s
a
t
e
l
l
i
t
e
i
m
a
ge
s
to
f
o
r
e
c
a
s
t
wh
e
a
t
y
i
e
l
d
d
i
f
f
e
r
e
n
c
e
s
a
t
t
h
e
f
a
r
m
l
e
v
e
l
u
s
i
n
g
a
m
a
c
hi
ne
l
e
a
r
ni
ng
t
e
c
hni
que
.
T
r
a
i
ni
ng
a
n
d
va
l
i
d
a
t
i
o
n
o
f
t
h
e
m
o
de
l
r
e
li
e
d
o
n
mo
r
e
t
h
a
n
10,
000
da
t
a
po
i
n
t
s
c
o
l
l
e
c
t
e
d
f
r
o
m
45
f
a
r
m
s
i
n
t
h
e
F
e
r
ga
n
a
Va
l
l
e
y
i
n
c
e
n
t
r
a
l
a
n
a
ly
s
e
s
m
o
ni
t
o
r
i
n
g
a
gr
i
c
u
l
t
ur
a
l
r
e
s
o
ur
c
e
s
(
M
A
R
S
)
a
gr
i
c
u
l
t
ur
a
l
y
i
e
l
d
f
o
r
e
c
a
s
t
i
n
g
s
y
s
t
e
m
da
t
a
to
a
s
s
e
s
s
n
e
ur
a
l
n
e
t
wo
r
k
m
o
de
l
s
f
o
r
a
gr
i
c
u
l
t
ur
a
l
yi
e
l
d
f
o
r
e
c
a
s
t
i
n
g.
W
e
l
o
o
ke
d
a
t
h
o
w
g
r
a
di
e
n
t
-
b
oo
s
t
e
d
de
c
i
s
i
o
n
t
r
e
e
s
(
GB
DT
s
)
a
n
d
l
o
n
g
s
h
o
r
t
-
t
e
r
m
m
e
m
o
r
y
(
L
S
T
M
)
m
o
de
l
s
d
i
d
c
o
m
pa
r
e
d
to
1D
c
o
n
v
o
l
ut
i
o
n
a
l
ne
ur
a
l
ne
t
wor
ks
(
1D
C
NN
s
)
.
L
S
T
M
wa
s
b
e
tt
e
r
a
t
p
r
e
d
i
c
t
i
n
g
Ge
r
m
a
ny
's
s
o
f
t
wh
e
a
t
y
i
e
l
d
s
t
h
a
n
GB
DT
,
b
ut
th
e
f
i
nd
i
ng
s
we
r
e
d
i
f
f
e
r
e
n
t
f
o
r
e
a
c
h
c
a
s
e
.
T
h
e
r
e
s
e
a
r
c
h
hi
g
hli
g
h
t
s
t
h
e
n
e
e
d
o
f
i
nv
o
l
vi
ng
s
t
a
ke
h
o
l
de
r
s
,
m
a
ke
m
o
de
l
s
e
a
s
y
t
o
un
de
r
s
t
a
n
d,
a
n
d
s
h
o
w
h
o
w
we
l
l
d
e
e
p
l
e
a
r
ni
ng
wo
r
ks
f
o
r
a
u
to
m
a
t
e
d
f
e
a
t
ur
e
e
x
t
r
a
c
t
i
o
n
a
nd
pr
e
di
c
t
i
o
n
.
I
t
a
l
s
o
t
a
l
ks
a
b
o
ut
h
o
w
m
a
c
hi
ne
l
e
a
r
ni
ng
m
a
y
b
e
us
e
d
to
gue
s
s
h
o
w
m
uc
h
f
i
ve
c
r
o
ps
i
n
R
a
j
a
s
t
ha
n
,
I
n
d
i
a
w
i
ll
yi
e
l
d.
R
a
n
do
m
f
o
r
e
s
t
o
u
t
pe
r
f
o
r
m
e
d
s
uppor
t
v
e
c
t
o
r
m
a
c
hi
ne
(
S
VM
)
,
gr
a
di
e
n
t
de
s
c
e
n
t
,
L
S
T
M
,
a
n
d
L
a
s
s
o
r
e
gr
e
s
s
i
o
n
.
R
e
c
e
n
t
l
y
,
r
e
s
e
a
r
c
h
e
r
s
h
a
v
e
c
r
e
a
t
e
d
a
l
o
t
o
f
de
e
p
l
e
a
r
ni
ng
(
DL
)
-
b
a
s
e
d
s
o
l
ut
i
o
ns
t
h
a
t
us
e
di
f
f
e
r
e
n
t
r
e
m
o
t
e
s
e
n
s
i
ng
(
R
S
)
da
t
a
s
e
t
s
a
n
d
DL
a
r
c
hi
t
e
c
t
ur
e
s
.
T
h
e
s
e
t
e
c
h
n
o
l
o
g
i
e
s
pr
o
duc
e
p
r
e
c
i
s
e
o
u
t
c
o
m
e
s
f
o
r
c
r
o
p
m
a
pp
i
ng
a
n
d
y
i
e
l
d
f
o
r
e
c
a
s
t
i
n
g.
T
h
e
s
e
s
o
l
ut
i
o
ns
a
l
s
o
n
e
e
d
to
f
i
gur
e
o
u
t
h
o
w
to
de
a
l
w
i
t
h
t
h
e
d
i
f
f
i
c
u
l
t
ies
o
f
n
o
t
h
a
vi
ng
e
n
o
ugh
t
r
a
i
ni
ng
da
t
a
,
b
u
i
l
d
i
ng
m
o
de
l
s
t
h
a
t
a
r
e
h
e
l
p
f
u
l
,
e
f
f
i
c
i
e
n
t
,
a
n
d
c
a
n
b
e
ut
i
li
z
e
d
o
n
a
w
i
de
r
a
n
ge
o
f
s
i
t
e
s
a
n
d
c
r
o
ps
.
T
h
e
s
e
a
r
e
a
l
l
im
po
r
t
a
n
t
f
o
r
p
r
o
vi
n
g
t
h
a
t
t
h
e
y
wo
r
k
a
n
d
c
a
n
b
e
us
e
d
i
n
d
if
f
e
r
e
n
t
f
a
r
m
i
n
g
s
i
t
ua
t
i
o
ns
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
tt
e
nti
on
-
e
nhanc
e
d
V
GG
-
16
ar
c
hit
e
c
tur
e
f
or
pr
e
c
is
ion
w
e
e
d
de
tec
ti
on
in
w
h
e
at
…
(
A
k
ank
s
ha
B
odhale
)
1305
T
a
bl
e
1.
L
i
t
e
r
a
t
u
r
e
o
n
s
t
a
t
e
-
o
f
a
r
t
r
e
s
e
a
r
c
h
A
ut
ho
r
(
s
)
M
e
th
o
d
ol
o
g
y
D
a
ta
s
e
t
A
c
c
u
r
a
c
y
(
%
)
K
a
una
la
ki
s
e
t
al
.
(
2019)
A
le
x
N
e
t,
V
G
G
-
F
, V
G
G
-
V
D
16, I
n
c
e
pt
i
o
n
-
V
I
,
R
e
s
N
e
t5
0, R
e
s
N
e
t1
01
M
o
n
o
c
h
r
o
m
e
C
a
me
r
a
R
e
s
N
e
t
-
50 w
it
h
96.10
L
a
m
e
t.
al
.
(
2021)
V
G
G
16
D
J
I
P
ha
nt
o
m 3 a
nd 4
92.1
L
iu
e
t
al
.
(
2021)
A
le
x
N
e
t,
V
G
G
16,
G
oo
gl
e
N
e
t
V
a
r
ia
bl
e
r
a
t
e
s
pr
a
y
e
r
VGG
-
16
a
c
hi
e
ve
d 96.00
M
il
io
t
o
e
t
al
.
(
2018)
C
N
N
+
F
ul
ly
C
o
n
vo
lu
t
i
o
na
l
N
e
tw
o
r
k
(
F
C
N
)
B
o
nn D
a
ta
s
e
t
(
R
G
B
+
N
I
R
)
91.88
O
ls
e
n
e
t
al
. (
2019)
I
ne
pt
i
o
n
-
V
3 a
nd R
e
s
N
e
t
-
50
D
a
ni
s
h W
e
e
d D
a
ta
s
e
t
95.1
D
y
r
ma
nn
e
t
al
.
(
2016)
H
O
G
+
S
V
M
c
la
s
s
if
i
e
r
D
a
ni
s
h A
gr
ic
ul
tu
r
a
l
D
a
ta
s
e
t
86.2
Y
u
e
t
al
.
(
2019)
D
e
e
p R
e
s
id
ua
l
N
e
tw
o
r
k
(
R
e
s
N
e
t
-
50)
C
us
to
m whe
a
t
f
i
e
ld
da
ta
s
e
t
(
C
hi
na
)
93.2
L
ot
te
s
e
t
al
.
(
2017
)
E
nc
o
d
e
r
–
D
e
c
o
de
r
C
N
N
(
S
e
gN
e
t
-
ba
s
e
d)
S
uga
r
B
e
e
t
D
a
ta
s
e
t
90.3
D
. S
. F
e
r
r
e
ir
a
e
t
al
.
(
2017)
C
N
N
w
it
h H
a
ndc
r
a
f
t
e
d P
r
e
pr
oc
e
s
s
in
g
R
e
a
l
f
ie
ld
i
ma
ge
s
i
n
B
r
a
z
il
88.4
P
r
o
p
o
s
e
d M
o
d
e
l
V
G
G
16
W
he
a
t
-
W
e
e
d D
a
ta
s
e
t
99.07
3.
M
E
T
HO
D
T
h
e
da
t
a
s
e
t
c
o
n
s
i
s
t
s
o
f
a
r
o
un
d
10,
046
i
m
a
ge
s
f
r
om
w
h
e
a
t
f
i
e
l
d
s
i
n
R
a
i
s
e
n
R
o
a
d,
M
a
dhy
a
P
r
a
de
s
h
,
I
n
d
i
a
(
23°
15
'
40.
5"
N
77°
38
'
60.
0"
E
)
,
c
a
p
t
u
r
e
d
dur
i
n
g
t
h
e
De
c
2024
–
J
a
n
2025
c
r
o
p
s
e
a
s
o
n
.
T
h
e
wh
e
a
t
-
we
e
d
da
t
a
s
e
t
i
s
a
gr
e
a
t
pl
a
c
e
t
o
s
t
a
r
t
wh
e
n
c
r
e
a
t
i
n
g
a
n
d
t
e
s
t
i
n
g
m
a
c
hi
ne
l
e
a
r
ni
ng
m
o
de
l
s
t
h
a
t
c
a
n
t
e
l
l
t
h
e
d
i
f
f
e
r
e
nc
e
b
e
t
we
e
n
h
e
a
l
t
hy
w
h
e
a
t
f
i
e
l
d
s
a
n
d
we
e
d
y
o
n
e
s
(
F
i
gur
e
1)
.
T
hi
s
c
o
l
l
e
c
t
i
o
n
h
a
s
10,
046
R
GB
pi
c
t
ur
e
s
,
wi
t
h
t
h
e
"
No
r
m
a
l
"
a
n
d
"
W
e
e
d"
s
e
g
m
e
n
t
s
e
a
c
h
h
a
vi
ng
5,
023
ph
o
to
s
.
T
h
e
"
No
r
m
a
l
"
c
l
a
s
s
h
a
s
2,
024
ph
o
to
g
r
a
ph
s
o
f
wh
e
a
t
c
r
o
ps
t
h
a
t
d
o
n
'
t
h
a
v
e
a
ny
w
e
e
ds
i
n
t
h
e
m
,
a
n
d
t
h
e
"
W
e
e
d"
c
l
a
s
s
h
a
s
8,
022
p
i
c
t
ur
e
s
s
h
o
w
i
n
g
d
i
f
f
e
r
e
n
t
k
i
n
ds
o
f
we
e
d
s
gr
o
wi
n
g
a
m
o
n
g
wh
e
a
t
pl
a
n
t
s
.
T
h
e
r
e
a
l
i
s
t
i
c
r
e
l
e
v
a
n
c
e
o
f
t
h
e
a
n
a
l
y
s
i
s
i
s
s
t
r
e
n
gt
h
e
ne
d
by
t
h
e
c
o
n
s
i
de
r
a
bl
e
v
a
r
i
a
t
i
o
n
be
t
we
e
n
t
h
e
t
w
o
c
l
a
s
s
e
s
,
whi
c
h
m
i
r
r
o
r
s
a
f
r
e
que
n
t
c
h
a
l
l
e
nge
f
o
un
d
i
n
r
e
a
l
-
wo
r
l
d
f
a
r
m
i
ng
e
nvi
r
o
nm
e
n
t
s
.
T
hi
s
da
t
a
pl
a
y
s
a
c
r
i
t
i
c
a
l
r
o
l
e
i
n
t
h
e
t
r
a
i
ni
n
g,
e
v
a
l
ua
t
i
o
n
,
a
n
d
v
a
li
da
t
i
o
n
o
f
t
h
e
d
i
f
f
e
r
e
n
t
s
e
g
m
e
n
t
a
t
i
o
n
a
n
d
c
l
a
s
s
i
f
i
c
a
t
i
o
n
t
e
c
h
ni
que
s
i
n
t
r
o
du
c
e
d
i
n
t
hi
s
wo
r
k
—
l
i
ke
C
NN
s
,
R
NN
s
,
a
n
d
VG
G
-
1
6.
T
h
e
s
e
a
l
go
r
i
t
hm
s
w
i
ll
c
o
n
t
r
i
b
ut
e
to
i
m
pr
o
vi
ng
t
h
e
pr
e
c
i
s
i
o
n
o
f
we
e
d
l
o
c
a
l
i
z
a
t
i
o
n
a
n
d
whe
a
t
f
i
e
l
d
m
o
ni
t
o
r
i
n
g,
t
h
e
r
e
by
s
uppo
r
t
i
n
g
a
d
v
a
n
c
e
m
e
n
t
s
i
n
pr
e
c
i
s
i
o
n
a
gr
i
c
u
l
t
ur
e
.
F
i
gur
e
1.
W
e
e
d
p
i
c
t
ur
e
o
f
da
t
a
3.
1.
P
r
e
-
p
r
oc
e
s
s
in
g
I
n
t
h
e
pr
e
pr
o
c
e
s
s
i
n
g
p
h
a
s
e
o
f
t
h
e
wh
e
a
t
-
we
e
d
c
l
a
s
s
if
i
c
a
t
i
o
n
wo
r
kf
l
o
w,
c
o
n
tr
a
s
t
s
tr
e
t
c
hi
ng
a
n
d
hi
s
t
o
gr
a
m
e
qua
l
i
z
a
t
i
o
n
s
e
r
v
e
a
s
t
w
o
f
un
d
a
m
e
n
t
a
l
c
o
n
t
r
a
s
t
i
m
pr
o
v
e
m
e
n
t
t
e
c
h
ni
que
s
.
T
h
e
s
e
m
e
t
h
o
ds
s
i
g
nif
i
c
a
n
t
l
y
b
o
o
s
t
i
m
a
ge
qua
l
i
t
y
a
n
d
s
i
m
p
li
f
y
f
e
a
t
ur
e
di
s
c
r
i
mi
na
t
i
o
n
w
i
t
hi
n
d
e
e
p
l
e
a
r
ni
ng
m
o
de
l
s
.
F
o
r
t
h
e
pr
e
s
e
n
t
,
hi
s
t
o
gr
a
m
e
qua
li
z
a
t
i
o
n
f
u
n
c
t
i
o
n
s
a
s
a
n
o
nl
i
ne
a
r
e
nh
a
n
c
e
m
e
n
t
a
ppr
o
a
c
h
t
h
a
t
i
nc
r
e
a
s
e
s
o
v
e
r
a
l
l
c
o
n
t
r
a
s
t
by
r
e
d
i
s
t
r
i
b
ut
i
n
g
p
i
xe
l
i
n
t
e
ns
i
t
y
l
e
v
e
l
s
.
T
h
e
s
e
m
e
t
h
o
ds
us
e
d
t
o
ge
t
h
e
r
n
o
r
m
a
l
i
z
e
im
a
ge
i
n
t
e
n
s
i
t
y
a
n
d
c
o
n
t
r
a
s
t
,
h
e
n
c
e
a
l
l
o
w
i
ng
m
o
de
l
s
l
i
ke
VG
G16
a
n
d
E
f
f
i
c
i
e
n
t
Ne
t
to
ga
i
n
s
t
r
o
n
ge
r
a
n
d
m
o
r
e
di
s
c
r
im
i
na
t
i
v
e
pr
o
pe
r
t
i
e
s
.
C
o
n
t
r
a
s
t
s
t
r
e
t
c
hi
n
g
e
nh
a
n
c
e
s
i
m
a
ge
qua
l
i
t
y
by
e
m
p
h
a
s
i
z
i
ng
s
u
b
t
l
e
de
t
a
i
l
s
i
n
l
e
a
f
m
o
r
p
h
o
l
o
g
y
t
h
a
t
m
a
y
a
ddr
e
s
s
c
e
r
t
a
i
n
de
f
i
c
i
e
nc
i
e
s
a
n
d
by
e
x
pa
n
d
in
g
t
h
e
r
a
n
ge
o
f
i
n
t
e
ns
i
t
y
va
l
u
e
s
.
H
i
s
t
o
gr
a
m
e
qua
l
i
z
a
t
i
o
n
im
pr
o
v
e
s
t
h
e
a
ppe
a
r
a
n
c
e
o
f
l
e
a
f
i
m
a
ge
s
t
h
r
o
ugh
t
h
e
a
dj
u
s
t
m
e
n
t
o
f
pi
x
e
l
i
n
t
e
ns
i
t
y
l
e
v
e
l
s
.
T
hi
s
m
e
t
h
o
d
c
r
e
a
t
e
s
a
m
o
r
e
uni
f
o
r
m
d
i
s
t
r
i
b
ut
i
o
n
o
f
i
n
t
e
n
s
i
t
i
e
s
a
c
r
o
s
s
t
h
e
i
m
a
ge
,
t
h
e
r
e
by
im
pr
o
vi
ng
t
h
e
vi
s
i
b
il
i
t
y
o
f
f
e
a
t
ur
e
s
l
i
nk
e
d
to
n
ut
r
i
e
n
t
di
s
o
r
de
r
s
,
s
uc
h
a
s
c
hl
o
r
o
s
i
s
o
r
n
e
c
r
o
s
i
s
.
F
i
gur
e
2
di
s
p
l
a
y
s
e
nh
a
n
c
e
d
i
m
a
g
e
s
o
f
b
o
t
h
h
e
a
l
t
hy
p
l
a
n
t
s
a
n
d
we
e
d
s
,
s
h
o
w
i
n
g
gr
e
a
t
e
r
c
l
a
r
i
t
y
a
n
d
f
e
a
t
ur
e
di
s
t
i
n
c
t
i
o
n
f
o
l
l
o
w
i
ng
t
h
e
a
pp
l
i
c
a
t
i
o
n
o
f
c
o
n
t
r
a
s
t
s
tr
e
t
c
hi
n
g
m
e
t
h
o
ds
(
a
s
i
l
l
u
s
t
r
a
t
e
d
i
n
F
i
gur
e
3)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
3
,
S
e
p
te
m
b
e
r
20
2
6
:
1303
-
1312
1306
F
i
gur
e
2.
N
o
r
m
a
l
a
n
d
we
e
d
p
h
o
to
gr
a
ph
s
a
f
t
e
r
a
pp
l
yi
ng
c
o
n
t
r
a
s
t
s
tr
e
t
c
hi
n
g
F
i
gur
e
3.
N
o
r
m
a
l
a
n
d
we
e
d
im
a
ge
s
a
f
t
e
r
a
ppl
yi
ng
hi
s
t
o
gr
a
m
e
qua
li
z
a
t
i
o
n
4.
P
ROP
OS
E
D
M
ODE
L
I
nh
e
r
e
n
t
l
y
a
pp
lyi
ng
t
h
e
de
s
i
g
n
a
n
d
a
s
s
e
s
s
m
e
n
t
of
f
i
ve
D
L
m
o
de
l
s
,
t
hi
s
pa
p
e
r
o
f
f
e
r
s
a
t
h
o
r
o
ugh
m
e
t
h
o
d
to
wh
e
a
t
-
we
e
d
c
a
t
e
g
or
i
z
a
t
i
o
n
.
T
he
da
t
a
c
a
m
e
f
r
o
m
gr
a
y
s
c
a
l
e
p
h
o
to
s
t
a
ke
n
i
n
a
gr
i
c
u
l
t
ur
a
l
f
i
e
lds
.
T
h
e
y
we
r
e
pr
e
pr
o
c
e
s
s
e
d
by
c
ha
n
g
i
ng
t
h
e
m
t
o
R
GB
f
o
r
m
a
t
a
n
d
s
c
a
l
i
ng
t
h
e
m
t
o
256×
256×
3
pi
xe
l
s
to
m
e
e
t
th
e
n
e
e
d
s
o
f
C
NN
-
b
a
s
e
d
a
r
c
hi
t
e
c
t
ur
e
s
.
B
a
t
c
h
-
wi
s
e
a
ug
m
e
n
t
a
t
i
o
n
a
n
d
n
o
r
m
a
li
z
a
t
i
o
n
we
r
e
t
w
o
m
o
r
e
p
r
e
pr
o
c
e
s
s
i
n
g
m
e
t
h
o
ds
t
h
a
t
h
e
l
pe
d
m
a
k
e
t
h
e
m
o
de
l
m
o
r
e
ge
n
e
r
a
li
z
a
bl
e
.
M
o
de
l
1
us
e
s
E
f
f
i
c
i
e
n
t
Ne
t
V2B
0
to
e
x
t
r
a
c
t
f
e
a
t
ur
e
s
.
T
hi
s
m
o
de
l
h
a
s
c
o
n
v
o
l
ut
i
o
n
a
l
l
a
y
e
r
s
a
n
d
a
c
us
tom
a
t
t
e
n
t
i
o
n
s
t
r
a
t
e
g
y
to
m
a
ke
i
t
b
e
tt
e
r
.
M
o
de
l
2
i
m
pr
o
v
e
s
l
e
a
r
ni
ng
s
pa
t
i
a
l
f
e
a
t
ur
e
s
by
a
dd
i
ng
m
o
r
e
C
o
n
v
2D
l
a
y
e
r
s
a
n
d
a
tt
e
n
t
i
o
n
to
a
p
r
e
tr
a
i
n
e
d
VG
G16
b
a
s
e
.
M
o
de
l
3
i
s
b
a
s
e
d
o
n
R
e
s
Ne
t
50
a
n
d
h
a
s
r
e
s
i
dua
l
bl
o
c
k
s
a
n
d
g
l
o
b
a
l
a
v
e
r
a
ge
poo
l
i
ng
to
e
x
t
r
a
c
t
s
t
r
o
n
g
hi
e
r
a
r
c
hi
c
a
l
f
e
a
t
ur
e
s
.
M
o
de
l
4
s
h
o
ws
a
hy
br
i
d
I
C
N
A
-
C
N
N
a
n
d
L
S
T
M
a
r
c
hi
t
e
c
t
ur
e
i
n
w
hi
c
h
t
h
e
L
S
T
M
l
a
y
e
r
s
a
c
t
l
i
ke
s
e
que
n
t
i
a
l
de
pe
n
de
nc
i
e
s
whil
e
t
h
e
C
NN
l
a
y
e
r
s
c
o
l
l
e
c
t
s
pa
t
i
a
l
i
n
f
o
r
m
a
t
i
o
n
.
Us
i
n
g
I
n
c
e
pt
i
o
nR
e
s
N
e
t
V2,
M
o
de
l
5
hi
g
hli
g
h
t
s
c
r
uc
i
a
l
vi
s
u
a
l
a
r
e
a
s
.
C
NN
a
n
d
a
t
t
e
n
ti
o
n
l
a
y
e
r
s
dr
i
ve
n
by
Que
r
y
-
K
e
y
t
r
a
n
s
f
o
r
m
a
t
i
o
n
m
a
ke
t
his
m
o
de
l
m
o
r
e
r
e
f
i
ne
d.
T
r
a
i
n
e
d
f
o
r
100,
e
a
c
h
i
t
e
r
a
t
i
o
n
e
m
p
l
o
y
s
t
h
e
Ada
m
o
p
t
i
mi
z
e
r
to
a
c
c
o
un
t
f
o
r
c
a
t
e
g
o
r
i
c
a
l
c
r
o
s
s
-
e
n
t
r
o
py
l
o
s
s
;
a
ll
m
o
de
l
s
-
e
m
p
l
o
y
R
e
L
U
a
c
t
i
va
t
i
o
n
,
b
a
t
c
h
n
o
r
m
a
l
i
z
a
t
i
o
n
,
M
a
x
P
o
o
l
i
ng2D,
&
dr
o
p
o
u
t
(
0.
5)
.
W
e
c
h
e
c
ke
d
t
h
e
m
o
de
l
's
a
c
c
ur
a
c
y
,
pr
e
c
i
s
i
o
n,
a
n
d
r
e
c
a
l
l
t
o
m
a
ke
s
ur
e
i
t
wa
s
goo
d
a
t
r
e
c
o
gni
z
in
g
p
l
a
n
t
s
pe
c
i
e
s
.
T
a
bl
e
2
r
e
pr
e
s
e
n
t
s
t
h
e
s
t
e
ps
o
f
t
h
e
pr
o
p
o
s
e
d
m
o
de
l
a
l
go
r
i
t
hm
a
n
d
F
i
gur
e
4
s
h
o
ws
t
h
e
pr
o
p
o
s
e
d
m
o
de
l
f
l
o
wc
ha
r
t
.
T
a
bl
e
2
.
P
r
o
p
o
s
e
d
m
o
de
l
a
l
go
r
i
t
hm
-
1
S
te
p
D
e
s
c
r
ip
ti
o
n
S
te
p 1
W
he
a
t
-
W
e
e
d D
a
ta
s
e
t
S
te
p 2
W
he
a
t
-
W
e
e
d
D
a
ta
s
e
t
W
he
a
t
-
W
e
e
d D
a
ta
s
e
t
W
he
a
t
-
W
e
e
d D
a
ta
s
e
t.
S
te
p 3
M
o
de
l
S
e
le
c
ti
o
n,
E
f
f
i
c
i
e
nt
N
e
tV2 +
C
N
N
+
A
tt
e
nt
i
o
n,
V
G
G
16
+
C
o
nv
2D
+
A
tt
e
nt
i
o
n, C
us
to
m R
e
s
N
e
t5
0, I
C
N
A
-
C
N
N
+
L
S
T
M
, I
n
c
e
pt
i
o
nR
e
s
N
e
tV2 +
C
N
N
+
A
tt
e
nt
io
n
S
te
p 4
A
dj
us
t
th
e
h
y
pe
r
pa
r
a
me
t
e
r
s
of
th
e
m
o
d
e
l,
s
u
c
h a
s
t
he
R
e
L
U
a
c
t
iv
a
ti
o
n, ba
t
c
h n
o
r
ma
li
z
a
ti
o
n, ma
x
p
oo
li
ng, a
nd
dr
o
p
o
ut
la
y
e
r
s
S
te
p 5
V
a
li
da
te
m
o
de
l
p
e
r
f
o
r
ma
n
c
e
o
n a
v
a
li
da
ti
o
n s
pl
it
of
t
he
da
ta
s
e
t
S
te
p 6
T
e
s
ti
ng a
nd
G
r
o
upi
ng U
s
e
w
h
e
a
t
-
w
e
e
d ph
ot
o
s
t
ha
t
th
e
t
r
a
in
e
d
mo
d
e
ls
ha
ve
n'
t
s
e
e
n b
e
f
or
e
t
o
t
e
s
t
th
e
m. Us
e
c
la
s
s
if
i
c
a
ti
o
n t
o
t
e
ll
t
h
e
di
f
f
e
r
e
nc
e
b
e
tw
e
e
n w
h
e
a
t
a
nd w
e
e
d
c
la
s
s
e
s
S
te
p 7
U
s
e
me
t
r
ic
s
, R
O
C
, a
nd t
he
C
o
n
ve
r
ge
n
c
e
c
ur
ve
t
o
r
a
t
e
e
a
c
h m
ode
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
tt
e
nti
on
-
e
nhanc
e
d
V
GG
-
16
ar
c
hit
e
c
tur
e
f
or
pr
e
c
is
ion
w
e
e
d
de
tec
ti
on
in
w
h
e
at
…
(
A
k
ank
s
ha
B
odhale
)
1307
4
.
1.
M
od
e
l
1:
e
f
f
icie
n
t
Ne
t
V2
f
or
wheat
-
we
e
d
c
las
s
if
i
c
at
ion
T
h
e
i
n
put
ph
oto
s
,
whi
c
h
we
r
e
i
ni
t
i
a
ll
y
i
n
bl
a
c
k
a
n
d
whi
t
e
,
a
r
e
c
h
a
n
ge
d
to
R
GB
a
n
d
r
e
s
i
z
e
d
to
256×
256×
3
to
f
i
t
t
h
e
m
o
de
l
's
i
n
put
r
e
qui
r
e
m
e
n
t
s
.
Af
t
e
r
f
e
a
t
ur
e
e
x
t
r
a
c
t
i
o
n
,
s
e
v
e
r
a
l
c
o
nv
o
l
ut
i
o
n
a
l
l
a
ye
r
s
a
n
d
a
s
pe
c
i
f
i
c
a
t
t
e
n
t
i
o
n
m
e
c
h
a
ni
s
m
a
r
e
us
e
d
to
e
n
h
a
n
c
e
t
h
e
r
e
l
e
va
n
t
f
e
a
t
ur
e
s
.
T
hi
s
h
e
l
ps
t
h
e
m
o
de
l
f
o
c
u
s
o
n
t
h
e
m
o
s
t
i
m
po
r
t
a
n
t
e
l
e
m
e
n
t
s
o
f
t
h
e
i
n
put
da
t
a
,
whi
c
h
m
a
ke
s
i
t
b
e
tt
e
r
a
t
t
hi
n
gs
l
i
ke
c
l
a
s
s
i
f
yi
ng
p
i
c
t
ur
e
s
or
f
i
n
d
i
ng
o
bj
e
c
t
s
.
T
h
e
m
o
de
l
a
r
c
hi
t
e
c
t
ur
e
i
n
c
l
u
de
s
R
e
L
U
a
c
t
i
va
t
i
o
n
s
a
n
d
M
a
x
P
o
o
l
i
ng2D
l
a
y
e
r
s
t
o
m
a
k
e
i
t
n
o
n
-
li
ne
a
r
o
r
l
o
we
r
i
t
s
d
i
m
e
ns
i
o
n
a
li
t
y
.
Dr
o
po
u
t
i
s
us
e
d
a
t
a
r
a
t
e
o
f
0.
5
i
n
c
o
m
p
l
e
t
e
l
y
c
o
nn
e
c
t
e
d
l
a
y
e
r
s
t
o
s
to
p
o
v
e
r
f
i
t
t
i
n
g.
R
e
c
a
ll
,
pr
e
c
i
s
i
o
n
,
a
n
d
a
c
c
ur
a
c
y
a
r
e
a
l
l
wa
y
s
t
o
m
e
a
s
ur
e
pe
r
f
o
r
m
a
n
c
e
.
F
i
gur
e
5
s
h
o
ws
t
h
e
l
a
y
e
r
e
d
a
r
c
hi
t
e
c
t
ur
e
o
f
t
h
e
m
o
de
l
.
F
i
gur
e
4.
P
r
o
p
o
s
e
d
f
l
o
wc
ha
r
t
F
i
gur
e
5.
L
a
y
e
r
e
d
a
r
c
hi
t
e
c
t
ur
e
f
o
r
m
o
de
l
s
4
.
2
.
M
od
e
l
2:
VGG16
f
or
wheat
-
we
e
d
c
l
as
s
if
ica
t
ion
T
o
e
x
t
r
a
c
t
s
t
r
o
n
g
hi
e
r
a
r
c
hi
c
a
l
a
t
t
r
i
b
ut
e
s
f
r
o
m
a
gr
i
c
u
l
t
ur
a
l
ph
o
to
s
,
t
h
e
VG
G16
-
b
a
s
e
d
a
ppr
o
a
c
h
us
e
s
a
pr
e
tr
a
i
ne
d
VG
G16
ne
t
wor
k.
I
n
o
r
de
r
to
m
a
ke
t
h
e
gr
e
y
s
c
a
l
e
p
h
oto
gr
a
ph
s
wo
r
k
wi
t
h
VG
G16,
we
h
a
d
to
r
e
s
i
z
e
t
h
e
m
a
n
d
c
h
a
n
g
e
t
h
e
i
r
f
o
r
m
a
t
to
R
GB
.
T
h
r
e
e
m
o
r
e
C
o
n
v
2
D
l
a
y
e
r
s
w
i
t
h
R
e
L
U
a
c
t
i
v
a
t
i
o
n
a
n
d
M
a
x
P
oo
l
i
ng
a
r
e
s
upe
r
i
m
po
s
e
d
o
n
to
p
o
f
t
h
e
pr
e
tr
a
i
n
e
d
b
a
s
e
to
m
a
ke
t
h
e
f
e
a
t
ur
e
m
a
ps
e
ve
n
b
e
t
t
e
r
.
M
e
tr
i
c
s
f
o
r
pe
r
f
o
r
m
a
n
c
e
i
nc
l
ude
r
e
c
a
ll
,
pr
e
c
i
s
i
o
n
,
a
n
d
a
c
c
ur
a
c
y
.
4
.
3
.
M
od
e
l
3:
Re
s
Ne
t
50
f
or
wheat
-
we
e
d
c
l
as
s
i
f
i
c
at
ion
Af
t
e
r
e
v
e
r
y
c
o
nv
o
l
ut
i
o
n
a
l
l
a
y
e
r
,
t
h
e
m
o
de
l
e
m
p
l
o
ys
b
a
t
c
h
n
o
r
m
a
li
z
a
t
i
o
n
,
a
n
d
f
o
r
s
pa
t
i
a
l
r
e
duc
t
i
o
n
,
i
t
m
a
ke
s
us
e
o
f
M
a
x
P
o
o
l
i
ng2D.
I
t
i
s
c
o
m
m
o
n
pr
a
c
t
i
c
e
t
o
c
o
m
pr
e
s
s
t
h
e
f
e
a
t
ur
e
m
a
p
s
us
i
n
g
g
l
o
b
a
l
a
v
e
r
a
ge
poo
l
i
ng
b
e
f
o
r
e
to
a
ddi
n
g
t
h
e
l
a
s
t
t
hi
c
k
l
a
y
e
r
s
.
T
o
b
u
il
d
t
h
e
m
o
de
l
,
we
us
e
d
c
a
t
e
g
or
i
c
a
l
c
r
o
s
s
-
e
n
t
r
o
p
y
f
o
r
l
o
s
s
a
n
d
t
h
e
a
da
m
-
o
pt
i
mi
z
e
r
,
c
h
e
c
k
e
d
i
t
s
pe
r
f
o
r
m
a
n
c
e
us
i
n
g
r
e
c
a
ll
,
pr
e
c
i
s
i
o
n
,
a
n
d
a
c
c
ur
a
c
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
3
,
S
e
p
te
m
b
e
r
20
2
6
:
1303
-
1312
1308
4
.
4
.
M
od
e
l
4:
I
CN
A
-
CN
N
wit
h
L
S
T
M
-
b
as
e
d
R
NN
f
o
r
wheat
-
we
e
d
c
l
as
s
i
f
icat
ion
T
o
e
x
t
r
a
c
t
s
pa
t
i
a
l
f
e
a
t
ur
e
s
,
t
hi
s
de
e
p
l
e
a
r
ni
ng
m
o
de
l
us
e
s
a
n
I
C
N
A
-
C
NN
,
a
n
d
t
o
c
a
p
t
ur
e
s
e
que
n
t
i
a
l
r
e
l
a
t
i
o
n
s
hi
p
s
,
i
t
us
e
s
a
n
L
S
T
M
,
o
r
l
o
n
g
s
h
o
r
t
-
t
e
r
m
m
e
m
o
r
y
,
n
e
t
wo
r
k
.
T
h
e
t
h
r
e
e
c
o
n
v
o
l
ut
i
o
n
a
l
l
a
y
e
r
s
t
h
a
t
o
pe
r
a
t
e
wi
t
h
i
nc
r
e
a
s
i
ng
f
il
t
e
r
s
(
32,
64,
128)
p
r
o
c
e
s
s
gr
e
y
s
c
a
l
e
i
nput
i
m
a
ge
s
w
i
t
h
d
i
m
e
ns
i
o
n
s
o
f
256
×
256×
1.
L
a
y
e
r
s
e
m
p
l
o
y
R
e
L
U
a
c
t
i
v
a
t
i
o
n
/M
a
x
P
o
o
l
i
ng2D
to
r
e
duc
e
d
i
m
e
n
s
i
o
n
s
w
hil
e
e
x
t
r
a
c
t
i
n
g
a
n
d
c
o
n
de
ns
i
ng
s
pa
t
i
a
l
da
t
a
.
Us
i
n
g
C
NN
s
f
o
r
s
pa
t
i
a
l
l
e
a
r
ni
ng
a
n
d
l
o
n
g
s
h
o
r
t
-
t
e
r
m
m
e
m
o
r
i
e
s
(
L
S
T
M
s
)
f
o
r
s
e
que
n
t
i
a
l
a
n
a
ly
s
i
s
,
t
hi
s
hy
b
r
i
d
m
e
t
h
o
d
i
m
pr
o
v
e
s
c
l
a
s
s
if
i
c
a
t
i
o
n
pe
r
f
o
r
m
a
n
c
e
.
4
.
5
.
M
od
e
l
5:
I
n
c
e
p
t
ion
Re
s
Ne
t
V2
wit
h
CN
N
a
n
d
at
t
e
n
t
ion
f
or
c
l
as
s
if
yin
g
wheat
an
d
we
e
d
s
To
i
m
pr
o
v
e
t
h
e
pe
r
f
o
r
m
a
n
c
e
o
f
c
a
t
e
g
o
r
i
z
i
ng
,
o
ur
c
utt
i
n
g
-
e
dge
de
e
p
l
e
a
r
ni
ng
m
o
de
l
b
r
i
n
g
s
to
ge
t
h
e
r
a
pr
e
tr
a
i
ne
d
I
n
c
e
pt
i
o
n
-
R
e
s
Ne
t
f
o
r
pul
l
i
ng
o
u
t
f
e
a
t
ur
e
s
,
a
l
o
n
g
w
i
t
h
C
NN
l
a
y
e
r
s
a
n
d
a
n
a
tt
e
n
t
i
o
n
m
e
c
h
a
ni
s
m
.
T
h
e
to
p
-
l
a
y
e
r
-
l
e
s
s
I
n
c
e
pt
i
o
n
R
e
s
Ne
t
V2
a
pp
l
i
e
s
r
e
s
i
du
a
l
a
n
d
I
n
c
e
pt
i
o
n
m
o
du
l
e
s
t
o
c
a
pt
u
r
e
hi
e
r
a
r
c
hi
c
a
l
a
n
d
m
u
l
t
i
-
s
c
a
l
e
pr
o
pe
r
t
i
e
s
.
W
e
a
dd
a
b
e
s
po
ke
a
tt
e
n
t
i
o
n
l
a
y
e
r
t
h
a
t
n
o
r
m
a
li
z
e
s
t
h
e
m
o
de
l
us
i
ng
S
o
f
t
m
a
x
a
f
t
e
r
Que
r
y
a
n
d
K
e
y
t
r
a
n
s
f
o
r
m
a
t
i
o
n
s
t
o
c
o
m
put
e
a
tt
e
n
t
i
o
n
we
i
g
h
t
s
,
whi
c
h
f
ur
t
h
e
r
i
m
pr
o
v
e
s
m
o
de
l
f
o
c
us
.
T
o
ke
e
p
t
hi
n
gs
f
r
o
m
o
v
e
r
f
i
t
t
i
n
g,
we
a
dde
d
t
w
o
de
n
s
e
l
a
y
e
r
s
-
o
n
e
w
i
t
h
512
uni
t
s
a
n
d
a
n
o
t
h
e
r
wi
t
h
256
-
a
nd
f
i
na
ll
y
e
x
pe
r
im
e
n
t
e
d
w
i
t
h
r
e
c
a
ll
a
n
d
a
c
c
ur
a
c
y.
H
e
r
e
's
a
c
o
n
s
o
l
i
da
t
e
d
hy
p
e
r
pa
r
a
m
e
t
e
r
de
t
a
i
l
s
t
a
bl
e
c
o
m
pa
r
i
n
g
a
l
l
f
i
ve
m
o
de
l
s
(
E
f
f
i
c
i
e
n
t
Ne
t
V2,
VG
G16
,
R
e
s
Ne
t
50,
I
C
NA
-
C
NN
+
R
NN
,
a
n
d
I
n
c
e
pt
i
o
nR
e
s
N
e
t
V2)
us
e
d
f
o
r
W
h
e
a
t
-
W
e
e
d
C
l
a
s
s
if
i
c
a
t
i
o
n
(
F
i
gur
e
6)
,
whi
c
h
s
h
o
ws
t
h
e
H
y
pe
r
pa
r
a
m
e
t
e
r
de
t
a
i
l
s
.
F
i
gur
e
6.
Hy
p
e
r
pa
r
a
m
e
t
e
r
d
e
t
a
i
l
s
5.
RE
S
UL
T
S
AN
D
DI
S
CU
S
S
I
ON
T
a
bl
e
3
s
h
o
ws
t
h
e
e
v
a
l
ua
t
i
o
n
m
e
t
r
i
c
s
f
o
r
f
i
v
e
di
f
f
e
r
e
n
t
de
e
p
l
e
a
r
ni
n
g
m
o
de
l
s
t
h
a
t
we
r
e
us
e
d
to
c
l
a
s
s
i
f
y
t
hi
n
gs
.
T
h
e
m
o
de
l
s
c
o
m
pa
r
e
d
i
n
c
l
u
de
c
us
to
m
R
e
s
Ne
t
50
,
R
NN
(
I
C
NA
-
C
NN
+
L
S
T
M
)
,
I
n
c
e
p
t
i
o
n
R
e
s
Ne
t
V2
,
E
f
f
i
c
i
e
n
t
Ne
tV2
,
a
n
d
t
h
e
p
r
o
p
o
s
e
d
-
VG
G1
6.
T
h
e
p
r
o
p
o
s
e
d
VG
G16
m
o
de
l
pr
e
s
e
n
t
s
t
h
e
b
e
s
t
o
v
e
r
a
l
l
e
f
f
i
c
i
e
n
c
y
w
i
t
h
t
h
e
hi
gh
e
s
t
pr
e
c
i
s
i
o
n
(
99.
0
7)
,
F
1
s
c
or
e
(
9
9.
20%
)
,
a
n
d
b
a
l
a
n
c
e
d
l
o
w
e
r
r
or
r
a
tes
(
F
NR
:
0.
6
951
%
;
F
P
R
:
0
.
269
4%
)
.
T
h
i
s
i
n
d
i
c
a
t
e
s
a
s
tr
o
n
g
g
e
n
e
r
a
l
i
z
a
t
i
o
n
c
a
pa
bi
l
i
t
y
a
n
d
m
i
nim
a
l
m
i
s
c
l
a
s
s
i
f
i
c
a
t
i
o
n
.
T
a
bl
e
3
.
C
o
m
pa
r
i
n
g
t
h
e
pe
r
f
o
r
m
a
n
c
e
o
f
o
ur
p
r
o
p
o
s
e
d
m
o
de
l
s
N
a
me
of
t
h
e
m
o
d
e
l
A
c
c
u
r
a
c
y
%
P
r
e
c
is
i
o
n %
R
e
c
a
ll
%
F1
-
S
c
o
r
e
%
F
N
R
%
F
P
R
%
C
us
to
m R
e
s
N
e
t5
0
90.58
90.57
89.11
89.84
10.887
8.127
R
N
N
(
I
C
N
A
-
C
N
N
+
L
S
T
M
)
93.40
93.25
92.47
92.86
7.531
5.797
I
nc
e
pt
i
o
nR
e
s
N
e
tV2
97.98
97.09
99.50
98.28
0.4975
0.411
E
f
f
i
c
i
e
nt
N
e
tV2
98.27
99.01
98.04
98.52
0.9608
0.4085
P
r
o
p
o
s
e
d
-
V
G
G
16
99.07
99.10
99.30
99.20
0.6951
0.2694
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
tt
e
nti
on
-
e
nhanc
e
d
V
GG
-
16
ar
c
hit
e
c
tur
e
f
or
pr
e
c
is
ion
w
e
e
d
de
tec
ti
on
in
w
h
e
at
…
(
A
k
ank
s
ha
B
odhale
)
1309
F
i
gur
e
7
di
s
t
i
n
c
t
l
y
de
m
o
ns
t
r
a
t
e
s
t
h
a
t
t
h
e
pr
o
p
o
s
e
d
VG
G16
m
o
de
l
i
s
t
h
e
b
e
s
t
o
p
t
i
o
n
f
o
r
r
e
a
l
-
wo
r
l
d
a
pp
l
i
c
a
t
i
o
n
b
e
c
a
us
e
i
t
i
s
m
o
r
e
a
c
c
ur
a
t
e
a
n
d
r
e
l
i
a
bl
e
t
h
a
n
a
ll
t
h
e
ot
h
e
r
s
.
M
o
de
l
s
s
uc
h
a
s
E
f
f
i
c
i
e
n
t
Ne
t
V2
a
n
d
I
n
c
e
pt
i
o
n
R
e
s
Ne
t
V2
a
l
s
o
pr
e
s
e
n
t
s
t
r
o
n
g
pot
e
n
t
i
a
l
,
whil
e
c
u
s
t
o
m
R
e
s
N
e
t
50
m
i
g
h
t
n
e
e
d
f
ur
t
h
e
r
o
p
t
i
mi
z
a
t
i
o
n
.
No
w,
F
i
gur
e
8
s
h
o
ws
us
wh
a
t
'
s
go
i
n
g
o
n
i
n
s
i
de
.
T
h
e
r
e
a
r
e
f
o
ur
l
i
ne
p
l
o
t
s
t
h
a
t
s
h
o
w
h
o
w
t
h
e
pr
o
p
o
s
e
d
m
o
de
l
wo
r
ks
t
h
r
o
ug
h
o
u
t
100
t
r
a
i
ni
ng
c
y
c
l
e
s
.
T
h
e
s
e
p
l
o
t
s
s
h
o
w
a
c
c
ur
a
c
y
,
l
o
s
s
,
pr
e
c
i
s
i
o
n
,
a
n
d
r
e
c
a
l
l
a
l
o
n
g
t
h
e
wa
y
.
F
i
na
ll
y
,
F
i
gur
e
9
s
h
o
ws
t
h
e
R
OC
c
ur
v
e
s
f
o
r
f
i
ve
d
i
f
f
e
r
e
n
t
de
e
p
l
e
a
r
ni
ng
m
o
de
l
s
.
O
n
c
e
m
o
r
e
,
t
h
e
pr
o
p
o
s
e
d
VG
G
-
16
m
o
de
l
s
t
a
n
ds
o
ut
,
wi
t
h
i
t
s
c
ur
ve
r
unni
n
g
n
e
a
r
t
h
e
to
p
-
l
e
f
t
c
o
r
n
e
r
—
a
c
l
e
a
r
i
n
d
i
c
a
t
o
r
o
f
i
t
s
s
t
r
o
n
g
pe
r
f
o
r
m
a
n
c
e
i
n
c
l
a
s
s
if
yi
ng
o
bj
e
c
t
s
.
F
i
gur
e
7.
P
r
e
s
e
n
t
a
c
o
m
pa
r
i
s
o
n
b
e
t
we
e
n
t
h
e
pr
o
p
o
s
e
d
m
o
de
l
a
n
d
w
i
t
h
ot
h
e
r
e
xi
t
i
n
g
m
o
de
l
s
F
i
gur
e
8.
P
r
o
p
o
s
e
d
m
o
de
l
c
o
n
ve
r
ge
nc
e
c
ur
v
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
3
,
S
e
p
te
m
b
e
r
20
2
6
:
1303
-
1312
1310
F
i
gur
e
9.
C
o
m
pa
r
i
s
o
n
o
f
R
OC
c
ur
v
e
—
pr
o
po
s
e
d
m
o
de
l
a
n
d
t
h
e
t
e
s
t
e
d
m
o
de
l
s
6.
CONC
L
USI
ON
I
n
c
o
n
c
l
us
i
o
n
,
t
hi
s
r
e
s
e
a
r
c
h
o
f
f
e
r
s
a
s
o
l
i
d
b
a
s
i
s
f
o
r
wh
e
a
t
-
we
e
d
c
l
a
s
s
i
f
i
c
a
t
i
o
n
ut
i
l
i
z
i
ng
a
gr
i
c
u
l
t
ur
a
l
im
a
g
i
ng
da
t
a
t
h
a
t
i
s
ba
s
e
d
o
n
DL
.
W
e
de
v
e
l
o
pe
d
a
n
d
e
v
a
l
ua
t
e
f
i
ve
d
i
s
t
i
n
c
t
m
o
de
l
s
to
de
t
e
r
m
i
ne
t
h
e
m
o
s
t
e
f
f
e
c
t
i
v
e
o
n
e
f
o
r
i
d
e
n
t
i
f
yi
ng
we
e
ds
i
n
w
h
e
a
t
f
i
e
l
ds
:
E
f
f
i
c
i
e
n
t
Ne
t
V2,
VG
G16
,
a
c
us
to
m
i
z
e
d
R
e
s
Ne
t
50,
a
hy
br
i
d
R
NN
i
n
c
o
r
p
o
r
a
t
e
I
C
NA
-
C
NN
w
i
t
h
L
S
T
M
,
a
n
d
I
n
c
e
pt
i
o
n
R
e
s
Ne
t
V2.
T
o
e
n
h
a
nc
e
i
m
a
ge
qua
li
t
y
a
n
d
e
ns
ur
e
m
o
r
e
r
o
b
us
t
f
e
a
t
ur
e
e
x
t
r
a
c
t
i
o
n
,
a
t
h
o
r
o
ugh
pr
e
pr
o
c
e
s
s
i
n
g
wo
r
kf
l
o
w
wa
s
i
m
p
l
e
m
e
n
t
e
d.
T
hi
s
p
i
pe
li
ne
i
nv
o
l
ve
d
s
e
ve
r
a
l
s
t
e
ps
,
i
n
c
l
ud
i
ng
c
o
nv
e
r
t
i
n
g
im
a
ge
s
f
r
o
m
gr
a
y
s
c
a
l
e
to
R
GB
,
r
e
s
i
z
i
ng,
n
o
r
m
a
li
z
a
t
i
o
n
,
c
o
n
t
r
a
s
t
s
t
r
e
t
c
hi
n
g,
hi
s
t
o
gr
a
m
e
qua
li
z
a
t
i
o
n
,
a
n
d
da
t
a
a
ug
m
e
n
t
a
t
i
o
n
.
Am
o
n
g
a
ll
t
h
e
a
r
c
hi
t
e
c
t
ur
e
s
e
v
a
l
ua
t
e
d,
o
ur
VG
G16
-
b
a
s
e
d
m
o
de
l
a
c
hi
e
ve
d
a
tr
a
i
ni
ng
a
c
c
ur
a
c
y
o
f
99.
07%
,
wi
t
h
pr
e
c
i
s
i
o
n
a
n
d
r
e
c
a
l
l
b
o
t
h
r
e
a
c
hi
ng
i
de
n
t
i
c
a
l
l
e
v
e
l
s
-
o
ut
pe
r
f
o
r
m
i
ng
e
v
e
r
y
o
t
h
e
r
m
o
de
l
.
T
hi
s
s
t
ud
y
c
o
n
t
r
i
b
ut
e
s
to
t
h
e
gr
o
wi
n
g
b
o
d
y
o
f
k
no
wl
e
dg
e
o
n
h
o
w
de
e
p
l
e
a
r
ni
ng
c
a
n
s
uppo
r
t
i
m
pr
o
v
e
d
de
c
i
s
i
o
n
-
m
a
k
i
ng
i
n
pr
e
c
i
s
i
o
n
a
gr
i
c
u
l
t
ur
e
.
T
h
e
u
l
t
i
m
a
t
e
a
i
m
i
s
to
m
a
xim
i
z
e
c
r
o
p
pr
o
duc
t
i
vi
t
y
w
hil
e
m
i
n
im
i
z
i
ng
pe
s
t
i
c
i
de
a
pp
l
i
c
a
t
i
o
n
a
n
d
pr
o
m
o
t
i
n
g
eco
-
f
r
i
e
n
d
l
y
f
a
r
m
i
ng
t
e
c
h
ni
que
s
.
F
UN
DI
NG
I
N
F
ORM
AT
I
ON
No
pa
r
t
i
c
u
l
a
r
gr
a
n
t
f
r
o
m
publi
c
,
pr
i
v
a
t
e
,
o
r
n
o
n
pr
o
f
i
t
f
u
n
d
i
ng
or
ga
ni
z
a
t
i
o
n
s
wa
s
r
e
c
e
i
v
e
d
f
o
r
t
hi
s
s
t
udy
.
T
h
e
a
ut
h
o
r
s
a
f
f
i
r
m
t
h
a
t
t
hi
s
wo
r
k
wa
s
n
ot
s
u
ppo
r
t
e
d
by
a
ny
f
u
n
d
i
ng
.
AU
T
HO
R
CONT
RI
B
UT
I
ONS
S
T
AT
E
M
E
NT
T
hi
s
j
o
ur
n
a
l
u
s
e
s
t
h
e
C
o
n
t
r
i
b
ut
or
R
o
l
e
s
T
a
x
o
n
o
m
y
(
C
R
e
d
i
T
)
to
r
e
c
o
gni
z
e
i
n
d
i
v
i
dua
l
a
ut
h
o
r
c
o
n
t
r
i
b
ut
i
o
ns
,
r
e
duc
e
a
ut
h
or
s
hi
p
d
i
s
put
e
s
,
a
n
d
f
a
c
il
i
t
a
t
e
c
o
l
l
a
b
o
r
a
t
i
o
n
.
Nam
e
of
Aut
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
A
ka
n
k
s
h
a
B
o
dh
a
l
e
S
e
e
m
a
Ve
r
m
a
Ai
s
hwa
r
y
B
o
dh
a
l
e
C
:
C
o
n
c
e
pt
ua
li
z
a
ti
o
n
M
:
M
e
th
o
d
o
l
o
g
y
So
:
So
f
twa
r
e
Va
:
Va
li
da
ti
o
n
Fo
:
Fo
r
ma
l
a
na
l
y
s
is
I
:
I
nve
s
ti
ga
ti
o
n
R
:
R
e
s
o
u
r
c
e
s
D
:
D
a
ta
C
ur
a
ti
o
n
O
:
W
r
it
in
g
-
O
r
ig
in
a
l
D
r
a
f
t
E
:
W
r
it
in
g
-
R
e
v
i
e
w
&
E
di
ti
ng
Vi
:
Vi
s
ua
li
z
a
ti
o
n
Su
:
Su
pe
r
v
is
io
n
P
:
P
r
o
j
e
c
t
a
dmi
ni
s
tr
a
ti
o
n
Fu
:
Fu
ndi
ng a
c
qui
s
it
i
o
n
CONF
L
I
CT
OF
I
NT
E
RE
S
T
S
T
AT
E
M
E
NT
T
h
e
a
ut
h
o
r
s
a
f
f
i
r
m
t
h
a
t
t
h
e
wor
k
de
s
c
r
i
b
e
d
i
n
t
hi
s
pu
bl
i
c
a
t
i
o
n
wa
s
n
o
t
i
m
pa
c
t
e
d
by
a
ny
k
n
o
wn
c
o
m
pe
t
i
n
g
f
i
na
nc
i
a
l
i
n
t
e
r
e
s
t
s
o
r
pe
r
s
o
n
a
l
r
e
l
a
t
i
o
ns
hi
ps
.
T
h
e
a
ut
h
o
r
s
a
l
s
o
a
f
f
i
r
m
t
h
a
t
t
hi
s
pa
pe
r
i
s
f
r
e
e
o
f
a
n
y
n
o
n
-
f
i
na
nc
i
a
l
c
o
nf
li
c
t
i
n
g
i
n
t
e
r
e
s
t
s
,
s
uc
h
a
s
t
h
o
s
e
t
h
a
t
a
r
e
a
c
a
de
m
i
c
,
i
n
t
e
l
l
e
c
t
ua
l
,
r
e
l
i
g
i
o
u
s
,
po
l
it
i
c
a
l
,
o
r
pe
r
s
o
n
a
l
.
No
c
o
nf
li
c
t
s
o
f
i
n
t
e
r
e
s
t
a
r
e
di
s
c
l
o
s
e
d
by
t
h
e
a
ut
h
o
r
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
tt
e
nti
on
-
e
nhanc
e
d
V
GG
-
16
ar
c
hit
e
c
tur
e
f
or
pr
e
c
is
ion
w
e
e
d
de
tec
ti
on
in
w
h
e
at
…
(
A
k
ank
s
ha
B
odhale
)
1311
DA
T
A
AV
AI
L
AB
I
L
I
T
Y
T
h
e
pr
i
m
a
r
y
da
t
a
us
e
d
i
n
t
hi
s
s
t
ud
y
wa
s
ga
t
h
e
r
e
d
s
t
r
a
i
g
h
t
f
r
o
m
w
h
e
a
t
f
i
e
l
ds
.
A
s
pa
r
t
o
f
t
h
e
f
i
e
l
d
e
x
pe
r
im
e
n
t
s
c
a
r
r
i
e
d
o
u
t
f
o
r
t
hi
s
s
t
ud
y
,
t
h
e
wr
i
t
e
r
s
c
r
e
a
t
e
d
a
n
d
d
o
c
um
e
n
t
e
d
t
h
e
da
t
a
.
T
h
e
a
s
s
o
c
i
a
t
e
d
a
ut
h
o
r
c
a
n
pr
o
vi
de
t
h
e
da
t
a
s
e
t
up
o
n
r
e
a
s
o
n
a
bl
e
r
e
que
s
t
.
RE
F
E
R
E
NC
E
S
[
1]
D
.
P
a
ude
l,
A
.
d
e
W
it
,
H
.
B
oo
ga
a
r
d,
D
.
M
a
r
c
o
s
,
S
.
O
s
in
ga
,
a
nd
I
.
N
.
A
th
a
na
s
ia
di
s
,
“
I
nt
e
r
pr
e
ta
bi
li
t
y
of
de
e
p
le
a
r
ni
ng
m
ode
ls
f
or
c
r
o
p
y
ie
ld
f
o
r
e
c
a
s
ti
ng,”
C
om
put
.
E
le
c
tr
on.
A
gr
ic
.
,
v
ol
.
206,
no
.
A
ugus
t
2022,
p.
107663,
2
023,
do
i:
10.1016/j
.
c
o
mpa
g.2023.107663.
[
2]
G
.
S
ha
o
,
W
.
H
a
n,
H
.
Z
ha
ng,
L
.
Z
ha
ng,
Y
.
W
a
ng,
a
nd
Y
.
Z
ha
ng,
“
P
r
e
di
c
ti
o
n
of
ma
i
z
e
c
r
o
p
c
o
e
f
f
i
c
i
e
nt
f
r
o
m
U
A
V
mul
ti
s
e
ns
or
r
e
m
o
t
e
s
e
ns
in
g
us
in
g
ma
c
hi
n
e
l
e
a
r
ni
ng
m
e
th
o
ds
,”
A
gr
ic
.
W
at
e
r
M
anag.
,
vo
l.
276,
n
o
.
J
ul
y
2021,
p.
108064,
20
23,
do
i:
10.1016/j
.a
gw
a
t.
2022.108064.
[
3]
Y
.
W
a
ng,
W
.
S
hi
,
a
nd
T
.
W
e
n,
“
P
r
e
di
c
ti
o
n
of
w
in
t
e
r
w
he
a
t
yi
e
ld
a
nd
dr
y
ma
tt
e
r
in
N
o
r
th
C
hi
na
P
la
in
us
in
g
ma
c
hi
n
e
l
e
a
r
n
in
g
a
lg
o
r
it
h
ms
f
o
r
o
pt
im
a
l
w
a
t
e
r
a
nd
ni
tr
o
ge
n
a
ppl
i
c
a
ti
o
n,”
A
gr
ic
.
W
at
e
r
M
anag.
,
vo
l.
277,
n
o
.
D
e
c
e
mbe
r
2022,
p.
108140,
20
23,
do
i:
10.1016/j
.a
gw
a
t.
2023.108140.
[
4]
J
.
L
iu
,
K
.
Y
a
ng,
A
.
T
a
r
iq
,
L
.
L
u,
W
.
S
o
u
f
a
n,
a
nd
A
.
E
l
S
a
b
a
gh,
“
I
nt
e
r
a
c
ti
o
n
of
c
li
ma
t
e
,
to
p
o
g
r
a
ph
y
a
nd
s
o
il
pr
o
p
e
r
ti
e
s
w
it
h
c
r
o
pl
a
nd
a
nd
c
r
o
ppi
ng
pa
tt
e
r
n
us
in
g
r
e
m
o
t
e
s
e
ns
in
g
da
ta
a
nd
m
a
c
hi
ne
l
e
a
r
ni
ng
m
e
th
o
ds
,”
E
gy
pt
.
J
.
R
e
m
ot
e
Se
n
s
.
Sp.
S
c
i.
,
v
o
l.
26,
no
.
3, pp. 415
–
426, 2023, d
o
i:
10.1016/j
.
e
j
r
s
.2023.05.005.
[
5]
K
.
J
ha
jh
a
r
ia
,
P
.
M
a
th
ur
,
S
.
J
a
in
,
a
nd
S
.
N
ij
ha
w
a
n,
“
C
r
o
p
Y
ie
ld
P
r
e
di
c
ti
o
n
us
in
g
M
a
c
hi
ne
L
e
a
r
ni
ng
a
nd
D
e
e
p
L
e
a
r
ni
ng
T
e
c
hni
qu
e
s
,”
P
r
oc
e
di
a C
om
put
. Sc
i.
, v
o
l.
218, pp. 406
–
417, 20
23, do
i:
10.1016/j
.p
r
oc
s
.2023.01.023.
[
6]
B
.
P
a
ni
g
r
a
hi
,
K
.
C
.
R
.
K
a
th
a
la
,
a
nd
M
.
S
uj
a
th
a
,
“
A
M
a
c
hi
n
e
L
e
a
r
ni
ng
-
B
a
s
e
d
C
o
mpa
r
a
ti
ve
A
ppr
o
a
c
h
t
o
P
r
e
di
c
t
th
e
C
r
o
p
Y
ie
ld
U
s
in
g
S
upe
r
v
is
e
d
L
e
a
r
n
in
g
W
it
h
R
e
gr
e
s
s
io
n
M
o
de
ls
,”
P
r
o
c
e
di
a
C
om
put
.
Sc
i.
,
v
o
l.
218,
n
o
.
2022,
pp.
2684
–
2693,
2
023
,
do
i:
10.1016/j
.p
r
oc
s
.2023.01.241.
[
7]
S
.
A
kba
r
ia
n,
M
.
R
.
J
a
mna
ni
,
C
.
X
u,
W
.
W
a
ng,
a
nd
S
.
L
im
,
“
P
lo
t
l
e
v
e
l
s
uga
r
c
a
n
e
y
i
e
ld
e
s
ti
ma
ti
o
n
b
y
ma
c
hi
n
e
l
e
a
r
ni
ng
o
n
mul
ti
s
pe
c
tr
a
l
im
a
g
e
s
:
a
c
a
s
e
s
tu
d
y
of
B
unda
be
r
g, A
us
tr
a
li
a
,”
I
n
f
. P
r
oc
e
s
s
. A
gr
ic
., 2023, do
i:
10.
1016/j
.i
npa
.2023.06.004.
[
8]
D
. B
. S
h
e
lk
e
, M
. R
. C
ha
mbha
r
e
,
G
. C
. N
ik
a
lj
e
, a
nd
T
. D
. N
ik
a
m, “
I
mpr
ove
me
n
t
of
S
oy
b
e
a
n C
r
o
p
f
or
Y
i
e
ld
, S
t
r
e
s
s
T
o
l
e
r
a
n
c
e
,
a
nd
V
a
lu
e
-
A
dde
d
P
r
o
du
c
ts
U
s
in
g a
T
r
a
ns
ge
ni
c
A
pp
r
o
a
c
h,”
A
dv
.
A
g
r
ic
.
, v
ol
. 2023, 2023, d
o
i:
10.1155/2023/
816
6928.
[
9]
D
.
P
a
ude
l
e
t
al
.
,
“
M
a
c
hi
ne
l
e
a
r
ni
ng
f
o
r
r
e
gi
o
na
l
c
r
o
p
y
i
e
ld
f
or
e
c
a
s
ti
ng
in
E
ur
o
p
e
,”
F
.
C
r
op.
R
e
s
.
,
v
o
l.
276,
no
.
D
e
c
e
mb
e
r
20
21,
p. 108377, 2022, do
i:
10.1016/j
.
f
c
r
.2021.108377.
[
10]
A
.
T
r
ip
a
th
i,
R
.
K
.
T
iwa
r
i,
a
nd
S
.
P
.
T
iwa
r
i,
“
A
de
e
p
le
a
r
ni
ng
m
ul
ti
-
l
a
y
e
r
pe
r
c
e
pt
r
o
n
a
nd
r
e
mo
te
s
e
ns
i
ng
a
ppr
o
a
c
h
f
o
r
s
o
il
he
a
lt
h
b
a
s
e
d
c
r
o
p
y
i
e
ld
e
s
ti
ma
t
io
n,”
I
nt
.
J
.
A
ppl
.
E
ar
t
h
O
bs
.
G
e
oi
n
f.
,
v
o
l.
113,
no
.
A
pr
i
l,
p
.
10
2959
,
2
022
,
do
i:
10.10
16/
j.
j
a
g.
2022
.102
959.
[
11]
R
.
A
w
o
r
ka
e
t
al
.
,
“
A
gr
i
c
ul
tu
r
a
l
d
e
c
is
i
o
n
s
y
s
te
m
b
a
s
e
d
o
n
a
d
v
a
nc
e
d
ma
c
hi
n
e
l
e
a
r
ni
ng
m
o
de
ls
f
or
y
i
e
ld
p
r
e
di
c
ti
o
n:
C
a
s
e
of
E
a
s
t
A
f
r
ic
a
n
c
o
unt
r
i
e
s
,”
S
ma
r
t
A
gr
i
c
.
T
e
c
hn
o
l.
,
vo
l.
2,
no
. J
a
nua
r
y
, p. 100048, 2022, d
oi
:
10.1016/j
.a
te
c
h.2022.10004
8.
[
12]
L
.
S
.
C
e
dr
ic
e
t
al
.
,
“
C
r
o
ps
y
i
e
ld
pr
e
di
c
ti
o
n
ba
s
e
d
o
n
ma
c
hi
ne
l
e
a
r
ni
ng
m
o
de
ls
:
C
a
s
e
o
f
W
e
s
t
A
f
r
i
c
a
n
c
o
unt
r
i
e
s
,”
Sm
ar
t
A
gr
ic
. T
e
c
hnol
.
, v
o
l.
2, n
o
. D
e
c
e
mbe
r
2021, p. 100049, 2
022, do
i:
10.1016/j
.a
t
e
c
h.2022.100049.
[
13]
A
.
I
kr
a
m
e
t
al
.
,
“
C
r
o
p
Y
ie
ld
M
a
x
im
i
z
a
ti
o
n
U
s
in
g
a
n
I
oT
-
B
a
s
e
d
S
ma
r
t
D
e
c
is
io
n,”
J
.
Se
ns
or
s
,
v
o
l.
2022,
2
022,
do
i:
10.1155/2022/
2022923.
[
14]
S
.
V
e
nka
te
s
a
n,
J
.
L
im
,
a
nd
Y
.
C
ho
,
“
A
C
r
o
p
G
r
o
w
th
P
r
e
di
c
ti
on
M
o
de
l
U
s
in
g
E
ne
r
g
y
D
a
ta
B
a
s
e
d
o
n
M
a
c
hi
ne
L
e
a
r
ni
ng
in
S
ma
r
t
F
a
r
ms
,”
C
om
put
. I
nt
e
ll
. N
e
ur
o
s
c
i.
, v
o
l.
2022, 2022,
d
oi
:
10.115
5/
2022/2648695.
[
15]
S
.
K
a
ma
l
e
t
al
.
,
“
F
C
N
N
e
tw
o
r
k
-
B
a
s
e
d
W
e
e
d
a
nd
C
r
o
p
S
e
gm
e
nt
a
ti
o
n
f
or
I
oT
-
A
id
e
d
A
gr
i
c
ul
tu
r
e
A
ppl
i
c
a
ti
ons
,”
W
ir
e
l.
C
om
m
un. M
ob. C
om
put
.,
vo
l.
2022, 2022, d
oi
:
10.1155/
2022/2770706.
[
16]
Q
.
Z
ha
ng,
B
.
L
i,
Y
.
Z
ha
ng,
a
nd
S
.
W
a
ng,
“
S
ui
ta
bi
li
t
y
E
v
a
lu
a
ti
o
n
of
C
r
o
p
V
a
r
i
e
t
y
v
ia
G
r
a
ph
N
e
ur
a
l
N
e
tw
o
r
k,”
C
om
put
.
I
n
te
ll
.
N
e
ur
os
c
i.
,
v
ol
. 2022, 2022, d
o
i:
10.1155/2022/
5614974.
[
17]
A
. B
ha
nda
r
i,
R
. J
o
s
hi
, M
. S
.
T
ha
pa
, R
. P
. S
ha
r
ma
, a
nd S
.
K
. R
a
uni
y
a
r
, “
L
a
nd
C
ove
r
C
ha
nge
a
nd I
ts
I
mpa
c
t
in
C
r
o
p Y
i
e
ld
:
A
C
a
s
e
S
tu
d
y
f
r
o
m W
e
s
te
r
n N
e
pa
l,
”
Sc
i.
W
or
ld
J
.
, vo
l.
2022, 2022, d
o
i
:
10.1155/2022/
5129423.
[
18]
S
.
G
upt
a
e
t
al
.
,
“
M
a
c
hi
ne
L
e
a
r
ni
ng
-
a
nd
F
e
a
tu
r
e
S
e
l
e
c
t
i
o
n
-
E
na
bl
e
d
F
r
a
m
e
w
o
r
k
f
o
r
A
c
c
u
r
a
te
C
r
o
p
Y
ie
ld
P
r
e
di
c
ti
o
n,”
J
. F
ood Qual.
,
vo
l.
2022, 2022, d
o
i
:
10.
1155/2022/
6293985.
[
19]
R
.
H
a
be
e
b,
I
.
H
us
s
a
in
,
N
.
A
l
-
A
ns
a
r
i,
a
nd
S
.
S
.
S
a
mm
e
n,
“
A
P
r
o
p
o
s
e
d
C
o
mpa
r
a
ti
ve
A
lg
or
it
hm
f
o
r
R
e
gi
o
na
l
C
r
o
p
Y
i
e
ld
A
s
s
e
s
s
me
nt
:
A
n
A
ppl
ic
a
ti
o
n
of
C
ha
r
a
c
t
e
r
is
t
ic
O
bj
e
c
ts
M
e
th
o
d,”
M
at
h.
P
r
ob
l.
E
ng.
,
vo
l.
2022,
20
22,
do
i:
10.1155/2022/
8
224953.
[
20]
D
.
P
a
ude
l
e
t
al
.
,
“
M
a
c
hi
n
e
le
a
r
ni
ng
f
o
r
la
r
ge
-
s
c
a
l
e
c
r
o
p
y
i
e
ld
f
or
e
c
a
s
ti
ng,”
A
gr
i
c
.
S
y
s
t.
,
vo
l.
187,
n
o
.
O
c
t
o
be
r
2020,
p.
1030
16,
2021, do
i:
10.1016/j
.a
gs
y
.2020.103016.
[
21]
Y
.
W
a
ng,
H
.
W
a
ng,
C
.
G
a
o
,
P
.
A
.
S
e
gl
a
h,
a
nd
Y
.
B
i,
“
U
r
e
a
A
ppl
ic
a
ti
o
n
R
a
te
f
or
C
r
o
p
S
t
r
a
w
D
e
c
o
mp
o
s
it
i
o
n
in
T
e
mpe
r
a
te
C
hi
na
,”
A
ppl
. E
nv
ir
on. Soil
Sc
i.
, vo
l.
2021, 2021, d
o
i:
10.1155/
2021/2240807.
[
22]
I
. N
ur
A
r
in
a
, M
. Y
. M
a
r
ti
ni
,
S
. S
u
r
di
a
na
, R
. M
o
hd
F
a
uz
i,
a
nd
S
. Z
ul
k
e
f
l
y
, “
R
a
di
a
ti
o
n D
y
na
mi
c
s
o
n
C
r
o
p P
r
o
du
c
ti
v
i
t
y
i
n D
i
f
f
e
r
e
nt
C
r
o
ppi
ng S
y
s
t
e
ms
,”
I
nt
. J
. A
gr
on.
, vol
. 2021, 2021, d
o
i:
10.115
5/
2021/4570616
[2
3
]
M
. B
. H
o
s
s
a
in
e
t
al
.
, “
R
e
s
po
ns
e
of
C
r
o
p W
a
te
r
R
e
qui
r
e
m
e
nt
a
n
d Y
ie
ld
of
I
r
r
ig
a
t
e
d R
ic
e
t
o
E
l
e
v
a
te
d
T
e
mpe
r
a
tu
r
e
i
n B
a
ngl
a
d
e
s
h,”
I
nt
. J
. A
gr
on.
, vo
l.
2021, 2021, d
o
i
:
10.1155/2021/
9963201.
[
24]
T
.
v
a
n
K
l
o
mp
e
nbur
g, A
.
K
a
s
s
a
hun, a
nd C
. C
a
ta
l,
“
C
r
o
p
y
ie
ld
p
r
e
di
c
ti
o
n us
in
g ma
c
h
in
e
l
e
a
r
ni
ng
:
A
s
y
s
te
ma
ti
c
l
i
t
e
r
a
tu
r
e
r
e
v
i
e
w
,”
C
om
put
. E
le
c
tr
on. A
gr
ic
.
, v
o
l.
177, n
o
. J
ul
y
, p. 105709, 2020, d
o
i:
10.1016/j
.
c
o
mpa
g.2020.105709.
[
25]
W
. D
o
ng, C
.
L
i,
Q
.
H
u, F
. P
a
n, J
. B
ha
nda
r
i,
a
nd Z
.
S
un, “
P
o
te
n
ti
a
l
E
v
a
po
t
r
a
ns
pi
r
a
ti
o
n R
e
du
c
ti
o
n a
nd I
ts
I
n
f
lu
e
n
c
e
o
n C
r
o
p Y
i
e
ld
in
t
he
N
o
r
th
C
hi
na
P
la
in
i
n 1961
-
2014,”
A
dv
. M
e
te
or
ol
.
, vo
l.
20
20, 2020, do
i:
10.1155/2020/
3691421.
[
26]
R
. K
a
ma
th
, M
. B
a
la
c
ha
nd
r
a
, a
nd S
. P
r
a
bhu, “
P
a
dd
y
C
r
o
p a
nd
W
e
e
d D
is
c
r
im
in
a
ti
o
n:
A
M
ul
ti
pl
e
C
la
s
s
if
i
e
r
S
y
s
t
e
m Appr
o
a
c
h,
”
I
nt
.
J
. A
gr
on.
, vo
l.
2020, 2020, d
o
i:
10.1155/2020/
6474536.
[
27]
H
.
S
un,
X
.
Z
ha
ng,
Z
.
Y
u,
a
nd
G
.
X
i,
“
F
e
a
tu
r
e
r
e
c
o
gni
t
i
o
n
of
c
r
o
p
gr
o
w
th
in
f
o
r
ma
ti
o
n
in
pr
e
c
is
i
o
n
f
a
r
mi
ng,”
C
om
pl
e
x
it
y
,
vo
l.
2018, 2018, d
o
i:
10.1155/2018
/9
250832.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
3
,
S
e
p
te
m
b
e
r
20
2
6
:
1303
-
1312
1312
B
I
OG
RA
P
HI
E
S
OF
AU
T
HO
RS
M
s
.
A
k
a
nk
s
h
a
Bo
dh
a
l
e
re
cei
v
ed
h
e
r
BE
D
eg
r
ee
i
n
CSE
fro
m
SG
SI
T
S
I
n
d
o
r
e
an
d
M
E
D
e
g
r
ee
i
n
I
T
Sp
eci
al
i
zat
i
o
n
w
i
t
h
I
n
f
o
r
m
at
i
o
n
S
ec
u
r
i
t
y
fr
o
m
IE
T
D
A
V
V
I
n
d
o
r
e
I
n
d
i
a
an
d
cu
rr
e
n
t
l
y
p
u
rs
u
i
n
g
a
Ph
.
D
.
S
h
e
i
s
p
re
s
e
n
t
l
y
r
e
s
e
arc
h
s
ch
o
l
ar
o
f
C
o
m
p
u
t
e
r
S
ci
e
n
ce
an
d
E
n
g
i
n
ee
r
i
n
g
fr
o
m
Ba
n
as
t
h
al
i
V
i
d
y
ap
i
t
h
Raj
as
t
h
an
In
d
i
a.
H
e
r
m
ai
n
r
e
s
e
arch
ar
e
a
Mach
i
n
e
L
e
ar
n
i
n
g
,
A
rt
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
ce,
Cr
y
p
t
o
g
rap
h
y
.
S
h
e
h
a
s
au
t
h
o
r
e
d
a
n
d
co
-
au
t
h
o
r
e
d
n
u
me
r
o
u
s
re
s
e
ar
ch
p
ap
e
rs
i
n
r
e
p
u
t
e
d
i
n
t
e
rn
at
i
o
n
a
l
j
o
u
rn
a
l
s
a
n
d
c
o
n
f
e
r
e
n
ce
s
an
d
8
p
at
e
n
t
s
h
a
v
e
b
ee
n
p
u
b
l
i
s
h
ed
.
Sh
e
c
a
n
b
e
c
o
n
t
ac
t
e
d
at
em
ai
l
:
ak
a
n
s
h
a2
6
1
1
@
g
m
ai
l
.
co
m
.
D
r.
See
m
a
V
er
m
a
i
s
w
o
r
k
i
n
g
as
Pro
f
e
s
s
o
r
o
f
E
l
ec
t
ro
n
i
c
s
E
n
g
i
n
ee
ri
n
g
at
N
I
T
T
T
R,
Bh
o
p
al
.
B
e
f
o
r
e
j
o
i
n
i
n
g
N
I
T
T
T
R,
s
h
e
w
as
w
o
rk
i
n
g
as
Pr
o
f
e
s
s
o
r
&
D
e
a
n
,
S
ch
o
o
l
o
f
A
v
i
at
i
o
n
,
Ba
n
as
t
h
al
i
V
i
d
y
ap
i
t
h
,
I
n
d
i
a.
Sh
e
h
as
au
t
h
o
r
ed
mo
r
e
t
h
a
n
t
o
1
0
0
p
ap
e
rs
i
n
v
ari
o
u
s
p
e
e
r
-
r
ev
i
ew
ed
J
o
u
r
n
al
s
an
d
Co
n
f
e
r
e
n
ce
s
.
S
h
e
h
as
a
u
t
h
o
re
d
6
b
o
o
k
s
w
i
t
h
r
en
o
w
n
ed
p
u
b
l
i
s
h
e
rs
.
Sh
e
h
as
b
e
en
g
ran
t
e
d
5
f
u
l
l
y
fu
n
d
e
d
p
ro
j
ec
t
s
fro
m
D
ST
,
U
G
C
an
d
A
I
CT
E
.
Sh
e
h
as
b
een
g
ran
t
ed
w
i
t
h
t
h
r
ee
In
d
i
an
Pat
en
t
s
an
d
h
as
p
u
b
l
i
s
h
ed
8
p
at
e
n
t
s
an
d
h
as
r
ecei
v
ed
1
5
co
p
y
ri
g
h
t
s
.
H
e
r
ar
e
as
o
f
r
e
s
e
ar
c
h
ar
e
W
i
r
e
l
e
s
s
Co
mmu
n
i
c
at
i
o
n
,
A
I
i
n
V
L
S
I,
a
n
d
s
emi
c
o
n
d
u
c
t
o
r
t
e
ch
n
o
l
o
g
y
.
Sh
e
c
a
n
b
e
c
o
n
t
ac
t
e
d
at
em
ai
l
:
s
v
e
r
m
a@
n
i
t
t
t
rb
p
l
.
ac
.
i
n
.
M
r.
A
i
s
h
w
a
ry
Bo
dh
a
l
e
r
ece
i
v
e
d
h
i
s
BE
i
n
E
l
ec
t
ri
c
al
E
n
g
i
n
ee
r
i
n
g
fr
o
m
RG
PV
U
n
i
v
e
rs
i
t
y
a
n
d
h
i
s
MS
i
n
Bu
s
i
n
e
s
s
A
n
al
y
t
i
c
s
an
d
Pro
j
ec
t
Ma
n
ag
eme
n
t
fro
m
U
n
i
v
e
rs
i
t
y
o
f
Co
n
n
ec
t
i
cu
t
,
U
S
A
.
H
e
i
s
cu
rr
e
n
t
l
y
p
u
rs
u
i
n
g
MB
A
i
n
Bu
s
i
n
e
s
s
A
n
a
l
y
t
i
c
s
fro
m
U
n
i
v
e
rs
i
t
y
o
f
Cu
m
b
e
rl
a
n
d
s
,
U
SA
.
H
e
i
s
p
re
s
e
n
t
l
y
w
o
rk
i
n
g
as
a
S
en
i
o
r
Man
a
g
e
r
at
Cr
e
d
i
t
O
n
e
Ba
n
k
,
U
S
A
,
w
h
e
r
e
h
e
l
e
a
d
s
d
at
a
-
d
r
i
v
en
i
n
i
t
i
at
i
v
e
s
i
n
c
r
ed
i
t
ri
s
k
,
A
I
an
al
y
t
i
c
s
,
an
d
e
n
t
e
rp
ri
s
e
d
at
a
s
t
rat
e
g
y
.
H
i
s
p
ri
m
ar
y
r
e
s
e
ar
c
h
a
n
d
p
ro
fe
s
s
i
o
n
al
i
n
t
e
r
e
s
t
s
i
n
cl
u
d
e
m
a
ch
i
n
e
l
e
ar
n
i
n
g
,
art
i
fi
ci
al
i
n
t
el
l
i
g
e
n
ce
,
d
at
a
an
al
y
t
i
c
s
,
c
r
ed
i
t
ri
s
k
mo
d
el
i
n
g
,
an
d
fi
n
an
ci
al
t
ec
h
n
o
l
o
g
y
(F
i
n
T
ec
h
)
.
Mr.
Bo
d
h
a
l
e
h
as
a
u
t
h
o
r
ed
a
n
d
co
-
au
t
h
o
red
s
ev
e
ral
r
es
e
arch
p
ap
e
rs
i
n
r
e
p
u
t
e
d
i
n
t
e
rn
at
i
o
n
al
j
o
u
rn
al
s
an
d
co
n
f
e
r
e
n
ce
s
an
d
h
as
c
o
n
t
ri
b
u
t
ed
t
o
i
n
n
o
v
at
i
v
e
s
o
l
u
t
i
o
n
s
i
n
d
at
a
a
n
d
an
a
l
y
t
i
c
s
.
H
e
c
an
b
e
co
n
t
a
c
t
ed
at
em
a
i
l
:
a
i
s
h
w
ar
y
b
o
d
h
al
e
3
@
g
m
ai
l
.
co
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.