I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
15
,
No
.
3
,
J
u
n
e
20
25
,
p
p
.
2834
~
2
8
4
4
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
3
.
pp
2
8
3
4
-
2
8
4
4
2834
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
A predi
ction o
f
coco
nut
a
nd coco
nu
t
lea
f
disea
se usin
g
M
o
bileNetV2 ba
s
ed
cla
ss
ificatio
n
K
a
v
it
ha
M
a
g
a
di G
o
pa
la
k
ris
hn
a
,
Ra
v
ipra
k
a
s
h M
a
denu
r
L
ing
a
ra
j
u
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
,
K
a
l
p
a
t
a
r
u
I
n
st
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
,
T
i
p
t
u
r
-
a
f
f
i
l
i
a
t
e
d
t
o
V
i
sv
e
s
v
a
r
a
y
a
Te
c
h
n
o
l
o
g
i
c
a
l
U
n
i
v
e
r
s
i
t
y
,
B
e
l
a
g
a
v
i
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 1
2
,
2
0
2
4
R
ev
is
ed
Ma
r
1
,
2
0
2
5
Acc
ep
ted
Ma
r
2
0
,
2
0
2
5
Th
is
re
se
a
rc
h
is
a
ime
d
a
t
e
ffe
c
ti
v
e
ly
p
re
d
icti
n
g
c
o
c
o
n
u
t
a
n
d
c
o
c
o
n
u
t
lea
f
d
ise
a
se
u
sin
g
e
n
h
a
n
c
e
d
M
o
b
i
leN
e
tV2
a
n
d
Re
sN
e
t5
0
m
e
th
o
d
s.
Th
e
sta
g
e
s
in
v
o
lv
e
d
in
t
h
is
imp
lem
e
n
ted
m
e
th
o
d
a
re
d
a
ta
c
o
l
lec
ti
o
n
,
p
re
-
p
r
o
c
e
ss
in
g
,
fe
a
tu
re
e
x
trac
ti
o
n
,
a
n
d
c
las
sifica
ti
o
n
.
At
first,
d
a
ta
is
c
o
l
lec
ted
fr
o
m
c
o
c
o
n
u
t
a
n
d
c
o
c
o
n
u
t
lea
f
d
a
tas
e
ts.
G
a
u
ss
ia
n
fil
ter
a
n
d
d
a
ta
a
u
g
m
e
n
tatio
n
t
e
c
h
n
iq
u
e
s
a
re
a
p
p
li
e
d
o
n
t
h
e
se
ima
g
e
s
to
e
li
m
in
a
te
n
o
ise
d
u
ri
n
g
th
e
p
re
-
p
ro
c
e
ss
in
g
p
h
a
se
.
Th
e
n
,
fe
a
tu
re
s
a
re
e
x
trac
ted
u
sin
g
Re
sN
e
t5
0
tec
h
n
iq
u
e
,
wh
il
e
th
e
d
ise
a
se
s
a
re
c
las
sified
u
sin
g
M
o
b
il
e
Ne
tV2
a
p
p
r
o
a
c
h
.
I
n
c
o
m
p
a
ri
so
n
t
o
th
e
e
x
isti
n
g
m
e
th
o
d
s
n
a
m
e
ly
,
Eff
ici
e
n
tDe
t
-
D2
,
DL
-
a
ss
isted
wh
it
e
fly
d
e
tec
ti
o
n
m
o
d
e
l
(DL
-
WD
M
),
a
n
d
m
o
d
ifi
e
d
i
n
c
e
p
ti
o
n
n
e
t
-
b
a
se
d
h
y
p
e
r
t
u
n
i
n
g
su
p
p
o
r
t
v
e
c
to
r
m
a
c
h
in
e
(M
IN
-
S
VM)
,
th
e
p
ro
p
o
se
d
m
e
th
o
d
a
c
h
iev
e
s
su
p
e
rio
r
c
las
sifica
ti
o
n
v
a
lu
e
s
with
9
9
.
9
9
%
a
n
d
9
9
.
2
%
a
c
c
u
ra
c
y
fo
r
c
o
c
o
n
u
t
lea
f
a
n
d
fo
r
c
o
c
o
n
u
ts,
re
sp
e
c
ti
v
e
ly
.
K
ey
w
o
r
d
s
:
C
o
co
n
u
t f
r
u
it
C
o
co
n
u
t le
af
Dis
ea
s
e
p
r
ed
ictio
n
Mo
b
ileNetV2
R
esNet5
0
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
:
Kav
ith
a
Ma
g
ad
i G
o
p
ala
k
r
is
h
n
a
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
Kalp
atar
u
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
,
T
ip
t
u
r
-
a
f
f
iliated
to
Vis
v
esv
ar
ay
a
T
ec
h
n
o
lo
g
ical
Un
iv
er
s
ity
B
elag
av
i
,
I
n
d
ia
E
m
ail:
k
av
ith
a.
m
g
@
k
itti
p
tu
r
.
a
c.
in
1.
I
NT
RO
D
UCT
I
O
N
I
n
r
ec
en
t
y
e
ar
s
,
co
co
n
u
t
h
as
em
er
g
ed
as
a
k
e
y
cr
o
p
in
tr
o
p
ical
r
eg
io
n
s
,
p
la
y
in
g
a
v
ital
r
o
le
in
th
e
ec
o
n
o
m
ies
o
f
s
ev
er
al
c
o
u
n
tr
ie
s
o
f
th
e
s
o
u
th
an
d
s
o
u
th
ea
s
t
Asi
an
r
eg
io
n
s
s
u
ch
as
Ph
ilip
p
in
es,
I
n
d
o
n
esia,
an
d
I
n
d
ia
[
1
]
.
R
ec
en
t
o
b
s
er
v
atio
n
s
h
av
e
s
h
o
wn
th
at
a
m
ajo
r
ity
o
f
co
co
n
u
t
tr
ee
s
ar
e
af
f
licted
with
d
is
ea
s
es,
wh
ich
p
r
o
g
r
ess
iv
ely
wea
k
e
n
an
d
r
estrict
th
e
am
o
u
n
t
o
f
co
co
n
u
t
p
r
o
d
u
ce
d
.
C
o
co
n
u
t
tr
ee
s
ar
e
r
ef
er
r
ed
to
as
th
e
‘
tr
ee
s
o
f
h
ea
v
en
’
as
all
p
ar
ts
o
f
th
is
p
lan
t
p
r
o
v
id
e
a
s
o
u
r
ce
o
f
in
co
m
e
f
o
r
f
ar
m
er
s
[
2
]
.
Mo
r
e
th
an
9
3
co
u
n
tr
ies
cu
ltiv
ate
co
co
n
u
ts
o
n
1
2
m
illi
o
n
h
ec
tar
es
o
f
lan
d
,
p
r
o
d
u
cin
g
5
9
.
9
8
m
illi
o
n
to
n
s
o
f
n
u
ts
an
n
u
ally
.
W
ith
1
0
.
5
6
m
illi
o
n
to
n
s
o
f
an
n
u
al
c
o
co
n
u
t
p
r
o
d
u
c
tio
n
,
I
n
d
ia
r
an
k
s
th
ir
d
in
th
e
wo
r
ld
[
3
]
.
T
h
e
I
n
d
ian
s
tates
th
at
p
r
o
d
u
ce
co
c
o
n
u
ts
ar
e
T
am
il
Nad
u
,
Ker
ala,
Kar
n
atak
a,
Ass
am
,
Or
is
s
a,
Go
a,
an
d
W
est
B
en
g
al
[
4
]
.
T
h
ese
tr
ee
s
p
r
o
v
id
e
a
v
ar
iety
o
f
v
alu
a
b
le
p
r
o
d
u
cts
th
at
in
cl
u
d
e
ed
ib
le
f
r
u
its
,
co
co
n
u
t
d
r
i
n
k
s
,
f
ib
er
f
o
r
c
o
m
m
er
cial
u
s
e
,
ed
ib
le
co
co
n
u
t
o
il,
an
d
co
co
n
u
t
s
h
ells
f
o
r
f
u
el
an
d
in
d
u
s
tr
ial
p
u
r
p
o
s
es
[
5
]
.
Sig
n
if
ican
t
d
am
ag
e
to
co
co
n
u
t
p
alm
s
f
r
o
m
co
co
n
u
t
leaf
d
is
ea
s
es
r
esu
lt
in
lo
wer
cr
o
p
y
ield
s
an
d
f
in
an
cial
lo
s
s
es
f
o
r
g
r
o
wer
s
.
Ad
d
itio
n
ally
,
n
u
tr
ie
n
t
d
ef
icien
cies
a
n
d
in
s
ec
t
in
f
estatio
n
s
h
ar
m
m
an
y
tr
ee
leav
es.
E
ar
ly
d
etec
tio
n
an
d
p
r
ed
ictio
n
a
r
e
ess
en
tial
f
o
r
im
p
lem
e
n
t
in
g
ef
f
ec
tiv
e
co
n
tr
o
l
m
ea
s
u
r
es
an
d
to
r
ed
u
ce
th
e
im
p
ac
t
o
f
th
es
e
d
is
ea
s
es
[
6
]
,
[
7
]
.
T
h
er
ef
o
r
e,
th
er
e
is
a
n
ee
d
f
o
r
s
izab
le
d
ataset
s
to
id
en
tify
t
h
e
s
y
m
p
to
m
s
o
f
d
is
ea
s
e
s
an
d
p
est in
f
ec
tio
n
s
u
s
in
g
v
ar
i
o
u
s
d
ee
p
lear
n
in
g
(
DL
)
an
d
m
ac
h
in
e
lear
n
in
g
(
ML
)
m
o
d
el
s
[
8
]
.
Ho
wev
er
,
d
u
e
to
th
e
d
if
f
icu
lt
an
d
tim
e
-
co
n
s
u
m
i
n
g
n
atu
r
e
o
f
d
ata
g
ath
er
in
g
o
n
all
d
is
ea
s
es
an
d
p
ests
,
th
is
s
tu
d
y
h
as
b
ee
n
r
estricte
d
to
t
h
e
d
etec
tio
n
o
f
o
n
ly
th
r
ee
d
is
ea
s
es:
s
tem
b
leed
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
p
r
ed
ictio
n
o
f c
o
co
n
u
t a
n
d
c
o
co
n
u
t le
a
f d
is
ea
s
e
u
s
in
g
Mo
b
ileN
etV
2
…
(
K
a
vith
a
Ma
g
a
d
i
Go
p
a
la
kris
h
n
a
)
2835
d
is
ea
s
e,
leaf
b
lig
h
t
d
is
ea
s
e,
an
d
r
ed
p
alm
wee
v
il
p
est
in
f
ec
ti
o
n
[
9
]
,
[
1
0
]
.
C
u
r
r
en
t
tr
e
n
d
s
d
e
m
o
n
s
tr
ate
th
at
ML
an
d
DL
ap
p
r
o
ac
h
es e
x
h
ib
it g
r
ea
t p
r
o
m
is
es in
p
r
e
d
ictin
g
an
d
m
an
ag
in
g
p
lan
t
d
is
ea
s
es
[
1
1
]
.
C
o
c
o
n
u
t
p
a
l
m
s
a
r
e
v
u
l
n
e
r
a
b
l
e
t
o
v
a
r
i
o
u
s
d
i
s
e
as
e
s
s
u
c
h
a
s
l
e
af
b
r
i
g
h
t
n
e
s
s
,
l
e
a
f
s
p
o
t
,
b
u
d
r
o
t
,
a
n
d
l
e
t
h
a
l
y
e
l
l
o
w
i
n
g
.
T
h
es
e
d
is
e
a
s
e
s
ca
u
s
e
s
i
g
n
i
f
i
c
a
n
t
r
e
d
u
c
t
i
o
n
i
n
y
i
eld
s
a
n
d
e
c
o
n
o
m
i
c
u
n
c
e
r
t
a
i
n
t
y
f
o
r
c
o
c
o
n
u
t
f
a
r
m
e
r
s
[
1
2
]
.
H
e
n
c
e
,
t
h
e
e
a
r
l
y
d
e
t
e
c
ti
o
n
o
f
c
o
c
o
n
u
t
l
e
a
f
d
i
s
e
as
e
s
is
c
r
u
c
i
a
l
d
u
e
t
o
v
a
r
i
o
u
s
f
a
ct
o
r
s
i
n
c
l
u
d
i
n
g
l
i
m
i
t
e
d
s
o
u
r
ce
a
c
c
e
s
s
i
b
il
i
t
y
a
n
d
t
h
e
h
i
g
h
t
i
m
e
c
o
n
s
u
m
p
t
i
o
n
o
f
t
h
e
e
x
is
t
i
n
g
m
o
d
e
l
s
[
1
3
]
.
M
L
a
l
g
o
r
it
h
m
s
s
u
ch
a
s
r
a
n
d
o
m
f
o
r
es
ts
,
d
e
c
i
s
i
o
n
t
r
e
e
s
,
a
n
d
n
e
u
r
a
l
n
e
tw
o
r
k
s
a
r
e
e
f
f
e
ct
i
v
e
i
n
c
o
c
o
n
u
t
l
e
a
f
d
is
e
a
s
e
p
r
e
d
i
c
ti
o
n
[
1
4
]
.
I
t
i
s
n
o
w
p
o
s
s
i
b
l
e
t
o
d
e
l
i
v
e
r
a
c
c
u
r
a
t
e
a
n
d
r
e
l
i
a
b
le
p
r
e
d
i
c
t
i
o
n
s
f
o
r
c
o
c
o
n
u
t
le
a
f
d
i
s
e
a
s
es
b
y
e
f
f
i
c
i
e
n
t
l
y
t
r
a
i
n
i
n
g
t
h
e
s
e
a
l
g
o
r
it
h
m
s
[
1
5
]
.
P
l
a
n
t
d
is
e
as
e
s
a
n
d
p
e
s
ts
h
a
v
e
c
o
n
s
i
s
t
e
n
t
l
y
i
n
c
r
ea
s
e
d
,
c
a
u
s
i
n
g
a
n
a
d
v
e
r
s
e
e
f
f
e
c
t
o
n
c
r
o
p
s
,
l
o
w
e
r
i
n
g
y
i
e
l
d
s
,
f
o
o
d
q
u
a
l
i
t
y
,
f
i
b
e
r
c
o
n
t
e
n
t
,
a
n
d
b
i
o
f
u
e
l
y
i
e
l
d
s
.
S
ta
t
e
-
of
-
t
h
e
-
a
r
t
m
o
d
e
l
s
l
i
k
e
I
n
c
e
p
ti
o
n
v
3
,
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
(
C
N
N
)
,
a
n
d
v
is
u
a
l
g
e
o
m
e
t
r
y
g
r
o
u
p
-
1
6
(
V
G
G
1
6
)
a
r
e
u
t
i
l
i
z
e
d
t
o
o
p
t
i
m
i
z
e
d
i
s
ea
s
e
d
et
e
c
ti
o
n
,
i
m
p
r
o
v
e
e
a
r
l
y
d
i
a
g
n
o
s
i
s
,
a
n
d
e
n
h
a
n
c
e
c
r
o
p
o
u
t
p
u
t
i
n
t
h
e
c
o
c
o
n
u
t
i
n
d
u
s
t
r
y
w
i
th
r
e
a
l
-
t
i
m
e
/
h
i
s
t
o
r
i
c
a
l
d
at
a
[
1
6
]
.
R
ec
en
t
o
b
s
er
v
atio
n
s
h
av
e
s
h
o
wn
th
at
a
m
ajo
r
ity
o
f
co
co
n
u
t
tr
ee
s
ar
e
af
f
licted
with
d
is
ea
s
es
wh
ich
p
r
o
g
r
ess
iv
ely
wea
k
en
an
d
lim
it
th
e
am
o
u
n
t
o
f
co
co
n
u
t
s
p
r
o
d
u
ce
d
[
1
7
]
.
I
n
o
r
d
er
to
m
a
x
im
ize
th
e
b
en
ef
its
o
f
co
co
n
u
t
p
r
o
d
u
ctio
n
,
t
h
e
p
r
im
a
r
y
f
o
cu
s
is
o
n
i
m
p
r
o
v
in
g
th
e
liv
elih
o
o
d
o
f
co
c
o
n
u
t
leav
es
th
r
o
u
g
h
ea
r
ly
d
is
ea
s
e
d
etec
tio
n
.
Ho
wev
e
r
,
m
a
n
y
f
a
r
m
er
s
s
tr
u
g
g
le
t
o
id
en
tif
y
d
i
s
ea
s
es
,
as
m
u
ltip
le
p
lan
ts
o
f
ten
ex
h
i
b
it
s
im
ilar
s
y
m
p
to
m
s
o
f
n
u
tr
ien
t
d
e
f
icien
cies.
Sin
ce
d
if
f
er
en
t
f
er
tili
ze
r
s
in
th
e
m
ar
k
et
c
o
n
tain
v
ar
y
in
g
lev
els
o
f
n
u
tr
ien
ts
,
s
elec
tin
g
th
e
r
ig
h
t
f
er
tili
ze
r
i
s
cr
u
cial
[
1
8
]
.
L
a
k
s
h
m
i
an
d
Sav
ar
im
u
th
u
[
1
9
]
in
tr
o
d
u
ce
d
an
E
f
f
icien
tDet
-
D2
m
eth
o
d
,
a
DL
tech
n
iq
u
e
d
esig
n
ed
to
d
etec
t
leaf
d
is
ea
s
e
r
eg
io
n
s
.
T
h
is
m
eth
o
d
o
f
f
er
ed
i
m
p
r
o
v
e
d
d
etec
tio
n
ac
cu
r
ac
y
ev
en
u
n
d
er
ad
v
e
r
s
e
co
n
d
itio
n
s
s
u
c
h
as
s
ig
n
if
ican
t
in
ter
class
v
ar
iatio
n
s
,
an
d
s
m
al
l
in
f
ec
ted
ar
ea
s
o
n
d
is
ea
s
ed
lea
v
es.
T
h
e
d
ev
el
o
p
ed
d
ee
p
lea
r
n
in
g
ap
p
r
o
a
ch
p
r
o
v
id
e
d
s
ig
n
if
ican
t
o
u
t
p
u
t
with
m
in
im
al
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
an
d
p
ar
am
eter
s
.
I
ts
ad
v
an
ta
g
e
s
in
clu
d
ed
co
s
t
-
ef
f
icien
cy
an
d
m
in
im
al
s
to
r
ag
e
r
eq
u
ir
em
e
n
ts
,
m
ak
in
g
it
s
u
i
tab
le
f
o
r
in
teg
r
atio
n
in
t
o
s
m
ar
tp
h
o
n
es
o
r
d
r
o
n
es
with
lim
ited
r
eso
u
r
ce
s
.
Kav
ith
am
an
i
an
d
Um
aM
a
h
eswar
i
[
2
0
]
im
p
lem
en
ted
a
DL
-
ass
is
ted
wh
itef
ly
d
etec
tio
n
m
o
d
el
(
DL
-
W
DM
)
f
o
r
th
e
id
en
tific
atio
n
o
f
wh
itef
lies
in
co
co
n
u
t
tr
ee
leav
es.
T
h
is
im
p
lem
en
ted
m
eth
o
d
was
em
p
lo
y
ed
to
d
iag
n
o
s
e
is
s
u
es
lik
e
b
lad
e
p
o
llu
tio
n
,
in
s
ec
t
in
f
ec
tio
n
,
an
d
r
o
o
t
b
leed
in
g
in
co
co
n
u
t
t
r
ee
s
.
T
h
e
d
e
v
elo
p
ed
m
o
d
el
im
p
r
o
v
ed
th
e
tr
an
s
lo
ca
tio
n
ef
f
icie
n
cy
o
f
co
co
n
u
t
tr
ee
leav
es;
h
o
we
v
er
,
th
e
DL
-
W
DM
m
o
d
el
r
eq
u
ir
ed
m
o
r
e
r
eso
u
r
ce
s
in
its
n
o
n
-
p
r
ess
u
r
ized
f
o
r
m
w
h
en
co
m
p
ar
e
d
to
th
e
p
r
ess
u
r
ized
f
o
r
m
.
Me
g
alin
g
am
et
a
l.
[
2
1
]
i
m
p
lem
en
ted
a
MI
N
-
SVM
m
o
d
el
f
o
r
th
e
class
if
icatio
n
o
f
c
o
co
n
u
t
tr
ee
s
.
T
h
e
p
r
e
-
p
r
o
ce
s
s
ed
im
ag
e
f
ea
t
u
r
es
o
f
th
e
co
co
n
u
t
tr
ee
wer
e
ex
tr
ac
ted
u
s
in
g
f
o
u
r
C
NN
m
o
d
els:
I
n
ce
p
tio
n
Net,
R
esN
e
t,
VGG)
an
d
MI
N
-
SVM.
T
h
e
clu
s
ter
in
g
an
aly
s
is
r
ev
ea
led
th
at
d
wa
r
f
an
d
tall a
c
ce
s
s
io
n
s
wer
e
s
ep
ar
ated
in
to
d
is
tin
ct
clu
s
ter
s
.
De
Sil
v
a
et
a
l.
[
2
2
]
d
ev
elo
p
ed
a
n
ested
p
o
ly
m
er
ase
ch
ain
r
e
ac
tio
n
(
PC
R
)
-
b
ased
m
eth
o
d
f
o
r
d
etec
tin
g
co
co
n
u
t
lea
f
d
is
ea
s
es
in
Sri
L
an
k
a.
T
h
e
m
o
d
el
also
tr
ac
k
ed
th
e
s
y
s
tem
atic
m
o
v
em
en
t
o
f
p
at
h
o
g
e
n
s
b
y
ev
alu
atin
g
th
e
s
u
itab
ilit
y
o
f
v
ar
io
u
s
PC
R
co
m
b
in
atio
n
s
,
d
is
tin
g
u
is
h
in
g
b
etwe
en
tis
s
u
e
ty
p
es
an
d
p
ath
o
g
en
m
o
v
em
en
ts
.
W
h
ile
th
e
m
eth
o
d
h
as a
lo
w
s
u
cc
ess
r
ate
in
p
ath
o
g
en
d
etec
tio
n
,
it su
cc
ess
f
u
lly
id
en
tifie
d
d
is
ea
s
e
-
f
r
ee
r
eg
io
n
s
in
th
e
s
am
p
le
d
ata.
Pam
m
it
et
a
l.
[
2
3
]
d
e
v
elo
p
ed
a
tr
an
s
cr
ip
t
ass
em
b
ly
o
f
r
e
f
er
en
ce
-
aid
e
d
f
u
ll
-
len
g
th
,
m
o
lec
u
lar
ch
ar
ac
te
r
izatio
n
an
d
cDN
A
clo
n
i
n
g
o
f
co
r
o
n
atin
e
-
in
s
en
s
itiv
e
1
b
g
en
e
(
C
OI
1
b
)
in
co
co
n
u
t
leav
es.
T
h
e
f
u
ll
-
len
g
th
C
OI
1
b
-
1
o
f
cDN
A
h
ad
7
9
1
9
b
p
with
O
R
F
o
f
1
1
7
6
b
p
wh
ich
was
en
c
o
d
e
d
f
o
r
an
in
f
er
r
ed
p
r
o
tein
o
f
3
9
1
ac
i
d
s
.
On
th
e
o
t
h
er
h
an
d
,
th
e
C
OI
1
b
-
2
h
ad
2
3
6
0
b
p
with
cDN
A
o
f
OR
F
1
7
4
3
b
p
wh
ich
en
co
d
e
d
an
in
f
e
r
r
ed
p
r
o
tein
o
f
5
8
0
ac
i
d
s
.
T
h
is
an
aly
s
is
p
r
o
v
ed
th
at
th
e
is
o
f
o
r
m
s
we
r
e
in
v
o
lv
e
d
in
v
ar
io
u
s
d
e
v
elo
p
m
e
n
t
p
r
o
ce
d
u
r
es,
in
clu
d
in
g
th
e
p
lan
t’
s
d
ef
en
s
e
m
ec
h
an
is
m
to
p
a
th
o
g
en
s
an
d
in
s
ec
ts
.
Pre
v
io
u
s
m
eth
o
d
s
s
tr
u
g
g
le
d
with
d
etec
tin
g
p
ests
d
u
e
to
lim
itatio
n
s
in
f
ac
to
r
s
s
u
ch
as
s
p
ac
e,
ed
g
es,
lig
h
tin
g
,
r
o
tatio
n
,
an
d
s
p
atial
v
ar
iatio
n
s
,
wh
ich
wer
e
r
etr
iev
e
d
f
o
r
ef
f
ici
en
t
p
est
d
etec
tio
n
.
W
ith
tech
n
o
lo
g
ical
a
d
v
an
ce
m
en
ts
,
ML
an
d
DL
tech
n
iq
u
es h
av
e
b
ec
o
m
e
cr
u
cial
f
o
r
ef
f
icien
t
co
co
n
u
t
d
is
ea
s
e
d
etec
tio
n
an
d
p
r
ev
en
ti
o
n
.
Fro
m
th
e
o
v
er
all
liter
atu
r
e
an
aly
s
is
,
it
is
s
ee
n
th
at
s
ev
er
al
DL
ap
p
r
o
ac
h
es
em
p
lo
y
ed
f
o
r
co
c
o
n
u
t
an
d
c
o
co
n
u
t
leaf
d
is
ea
s
e
id
en
tific
a
tio
n
ex
p
e
r
ien
ce
ce
r
tain
s
etb
a
ck
s
.
T
h
is
is
d
u
e
to
in
ef
f
icien
t
f
ea
tu
r
e
lea
r
n
in
g
o
f
t
h
e
ex
tr
ac
ted
f
ea
tu
r
es,
wh
ich
m
ad
e
it
h
ar
d
er
f
o
r
th
e
m
o
d
el
to
d
i
f
f
er
en
tiate
b
etwe
e
n
d
is
ea
s
ed
an
d
h
ea
lth
y
im
ag
es,
r
esu
ltin
g
in
in
ac
cu
r
ate
p
r
ed
icti
o
n
.
T
h
is
m
o
tiv
ates
th
e
p
r
esen
t
s
tu
d
y
th
at
p
r
o
p
o
s
es
DL
-
b
ased
p
r
e
d
ictio
n
m
o
d
el
f
o
r
ac
cu
r
ate
p
r
ed
ictio
n
o
f
c
o
co
n
u
t
an
d
c
o
co
n
u
t
leaf
im
ag
es
d
a
taset.
T
h
u
s
,
i
n
th
is
r
esear
ch
,
th
e
Mo
b
ileNetV2
m
eth
o
d
is
u
tili
ze
d
to
p
r
ed
ict
co
co
n
u
t
leaf
d
is
ea
s
es
with
th
e
p
r
esen
ce
o
f
ca
ter
p
illar
s
,
leaf
lets
,
d
r
y
in
g
o
f
leaf
lets
,
f
lac
cid
ity
,
an
d
y
ello
win
g
,
f
o
r
th
e
c
lass
if
icatio
n
o
f
d
is
ea
s
es
with
b
u
d
r
o
t
an
d
n
u
t
f
all.
T
h
e
m
ajo
r
c
o
n
tr
ib
u
tio
n
s
o
f
t
h
is
s
tu
d
y
ar
e
lis
ted
b
elo
w:
−
I
n
th
is
s
tu
d
y
,
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
p
r
o
p
o
s
ed
Mo
b
ileNetV2
is
an
aly
ze
d
o
n
co
co
n
u
t
an
d
co
c
o
n
u
t
lea
f
im
a
g
e
d
atasets
.
−
Pre
-
p
r
o
ce
s
s
in
g
is
ca
r
r
ied
o
u
t
u
s
in
g
a
Gau
s
s
ian
f
ilter
f
o
r
n
o
is
e
elim
in
atio
n
,
wh
ile
d
ata
au
g
m
en
tatio
n
is
d
ep
lo
y
e
d
f
o
r
im
p
r
o
v
in
g
th
e
m
o
d
el’
s
p
er
f
o
r
m
an
ce
.
−
T
h
en
,
th
e
R
esNet5
0
is
em
p
lo
y
ed
f
o
r
t
h
e
ex
tr
ac
tio
n
o
f
o
p
tim
a
l f
ea
tu
r
es f
r
o
m
t
h
e
p
r
e
-
p
r
o
ce
s
s
ed
im
ag
e,
af
ter
wh
ich
class
if
icatio
n
is
ca
r
r
ied
o
u
t u
s
in
g
Mo
b
ileNetV2
.
I
n
th
e
f
in
al
s
tag
e,
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
p
r
o
p
o
s
ed
Mo
b
ileNetV2
is
an
aly
ze
d
in
te
r
m
s
o
f
p
r
ec
is
io
n
,
ac
cu
r
ac
y
,
f
-
s
co
r
e,
an
d
r
ec
all.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
3
,
J
u
n
e
20
25
:
2
8
3
4
-
2
8
4
4
2836
T
h
is
r
est
o
f
th
is
m
an
u
s
cr
ip
t
i
s
o
r
g
an
ized
as
f
o
llo
ws:
th
e
p
r
o
p
o
s
ed
m
eth
o
d
is
d
etailed
in
s
ec
tio
n
2
,
class
if
icatio
n
u
s
in
g
Mo
b
ileN
etV2
is
ex
p
lain
ed
in
s
ec
tio
n
3
,
t
h
e
co
m
p
ar
ativ
e
an
aly
s
is
is
ca
r
r
ied
o
u
t
i
n
s
ec
tio
n
4
,
wh
ile
th
e
co
n
clu
s
io
n
o
f
th
is
r
esear
ch
is
s
p
ec
if
ied
in
s
ec
tio
n
5.
2.
M
E
T
H
O
D
T
h
e
p
r
o
p
o
s
ed
Mo
b
ileNetV2
m
eth
o
d
is
im
p
lem
e
n
ted
f
o
r
t
h
e
c
lass
if
icatio
n
o
f
co
co
n
u
t
an
d
c
o
co
n
u
t
leaf
d
is
ea
s
es.
T
h
e
co
co
n
u
t
a
n
d
co
c
o
n
u
t
lea
f
im
ag
e
d
atasets
ar
e
u
t
ilized
in
th
is
p
ap
er
f
o
r
g
ath
er
in
g
im
ag
es,
f
o
llo
wed
b
y
Gau
s
s
ian
f
ilter
an
d
d
ata
a
u
g
m
en
tatio
n
em
p
lo
y
ed
f
o
r
i
m
ag
e
p
r
e
p
r
o
ce
s
s
in
g
f
o
r
n
o
is
e
elim
in
atio
n
.
T
h
e
n
,
R
esNet5
0
is
em
p
lo
y
ed
f
o
r
f
e
atu
r
e
ex
tr
ac
tio
n
f
r
o
m
th
e
p
r
e
-
p
r
o
ce
s
s
ed
im
ag
es,
an
d
f
in
all
y
,
Mo
b
ileNetV2
is
u
tili
ze
d
f
o
r
d
is
ea
s
e
class
if
icat
i
o
n
.
T
h
e
b
lo
ck
d
iag
r
a
m
o
f
t
h
e
o
v
er
all
p
r
o
ce
s
s
is
g
iv
en
in
Fig
u
r
e
1
.
Fig
u
r
e
1
.
Ov
e
r
all
b
lo
ck
d
iag
r
a
m
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
2
.
1
.
Da
t
a
s
et
s
des
cr
iptio
n
T
h
is
r
esear
ch
u
tili
ze
s
co
co
n
u
t
an
d
c
o
co
n
u
t
leaf
im
ag
e
d
atasets
f
o
r
a
n
aly
s
is
.
T
h
e
i
m
ag
es
ar
e
ca
p
tu
r
e
d
u
s
in
g
d
i
g
ital
im
ag
in
g
d
ev
ices
lik
e
d
ig
ital
ca
m
er
as,
ce
ll
p
h
o
n
es,
d
ig
ital
s
in
g
le
-
len
s
r
ef
lex
es
,
an
d
o
t
h
er
s
im
ilar
d
ev
ices.
T
h
e
im
a
g
es
ar
e
ca
p
tu
r
ed
in
v
ar
io
u
s
n
atu
r
al
e
n
v
ir
o
n
m
en
ts
,
at
d
if
f
e
r
en
t
tim
es
o
f
th
e
d
ay
,
with
v
ar
y
in
g
lig
h
t
in
ten
s
ities
an
d
an
g
les
to
a
v
o
id
r
e
d
u
n
d
an
cy
in
im
a
g
es.
T
h
e
co
co
n
u
t
im
ag
es
ar
e
co
llected
f
r
o
m
in
an
d
ar
o
u
n
d
th
e
f
ar
m
s
o
f
T
ip
tu
r
,
T
u
m
ak
u
r
u
Dis
tr
ict,
an
d
Kar
n
atak
a
State.
T
h
e
co
co
n
u
t
im
a
g
e
d
ataset
co
n
tain
s
3
1
5
im
a
g
es
wh
ich
ar
e
s
p
lit
in
to
two
ca
te
g
o
r
ies:
1
5
0
im
ag
es
o
f
h
ea
lth
y
co
co
n
u
ts
,
an
d
1
6
5
im
ag
es
o
f
in
f
ec
ted
co
co
n
u
ts
.
Similar
ly
,
th
e
co
co
n
u
t
leaf
im
ag
e
d
ataset
[
2
4
]
co
n
s
is
ts
o
f
a
to
tal
o
f
5
,
0
3
6
im
ag
es,
wh
ic
h
ar
e
ca
teg
o
r
ized
i
n
to
5
class
es
as
7
9
5
im
ag
es
o
f
leaf
lets
,
9
9
0
im
ag
es
o
f
ca
ter
p
illar
s
,
1
0
8
4
i
m
ag
es
o
f
leaf
o
r
f
r
u
it
y
ello
win
g
,
1
0
8
8
im
ag
es
with
d
r
ied
leaf
lets
,
an
d
1
0
7
9
im
ag
es
o
f
f
lacc
id
ity
.
Fig
u
r
e
2
r
e
p
r
esen
ts
th
e
s
am
p
l
e
im
ag
es
o
f
co
c
o
n
u
t
im
ag
e
d
ataset,
in
Fig
u
r
e
2
(
a)
r
ep
r
esen
ts
h
ea
lth
y
c
o
co
n
u
t
s
a
m
p
le
an
d
Fig
u
r
e
2
(
b
)
r
ep
r
esen
ts
in
f
ec
ted
co
c
o
n
u
t
s
am
p
le.
Fig
u
r
e
3
r
ep
r
esen
ts
th
e
s
am
p
le
im
ag
es
o
f
co
co
n
u
t
leaf
im
ag
e
d
ataset,
in
Fig
u
r
e
3
(
a
)
r
e
p
r
ese
n
ts
ca
ter
p
illar
s
class
s
am
p
le
im
ag
e,
Fig
u
r
e
3
(
b
)
r
ep
r
esen
ts
lea
f
lets
class
s
am
p
le
im
ag
e,
Fig
u
r
e
3
(
c)
r
ep
r
esen
ts
d
r
y
in
g
o
f
leaf
lets
class
s
am
p
le
im
ag
e,
Fig
u
r
e
4
(
d
)
r
ep
r
esen
ts
f
lacc
id
ity
class
s
am
p
le
im
ag
e
an
d
Fig
u
r
e
5
(
e)
r
ep
r
esen
ts
y
ello
win
g
class
s
am
p
le
im
ag
e.
(
a)
(
b
)
Fig
u
r
e
2
.
Sam
p
le
im
a
g
es o
f
th
e
co
co
n
u
t
im
ag
e
d
ataset
(
a
)
h
e
alth
y
co
co
n
u
t a
n
d
(
b
)
in
f
ec
ted
co
co
n
u
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
p
r
ed
ictio
n
o
f c
o
co
n
u
t a
n
d
c
o
co
n
u
t le
a
f d
is
ea
s
e
u
s
in
g
Mo
b
ileN
etV
2
…
(
K
a
vith
a
Ma
g
a
d
i
Go
p
a
la
kris
h
n
a
)
2837
(
a)
(
b
)
(
c)
(
d
)
(
e)
Fig
u
r
e
3
.
Sam
p
le
im
a
g
es o
f
c
o
co
n
u
t
leaf
im
a
g
e
d
ataset
(
a)
ca
ter
p
illar
s
,
(
b
)
leaf
lets
,
(
c)
d
r
y
i
n
g
o
f
leaf
lets
,
(
d
)
f
lacc
id
ity
,
an
d
(
e)
y
ell
o
win
g
2
.
2
.
P
re
-
pro
ce
s
s
ing
Pre
p
r
o
ce
s
s
in
g
is
p
er
f
o
r
m
ed
af
ter
co
llectin
g
d
ata
f
r
o
m
c
o
co
n
u
t
a
n
d
co
c
o
n
u
t
lea
f
im
ag
es.
T
h
e
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es,
Gau
s
s
ian
f
ilter
an
d
d
ata
au
g
m
en
tat
io
n
ar
e
u
tili
ze
d
to
co
n
v
e
r
t
r
aw
d
ata
in
to
a
u
s
ab
le
f
o
r
m
at
th
r
o
u
g
h
n
o
is
e
r
em
o
v
al
an
d
d
ata
b
alan
cin
g
f
o
r
all
cla
s
s
es.
B
o
th
g
au
s
s
ian
f
ilter
an
d
d
ata
au
g
m
en
tatio
n
tech
n
iq
u
es a
r
e
e
x
p
lain
ed
b
elo
w
in
d
etail.
2
.
2
.
1
.
G
a
us
s
ia
n
f
ilte
r
Gau
s
s
ian
f
ilter
is
ap
p
lied
to
f
ilter
n
o
is
e
p
r
esen
t
in
co
c
o
n
u
t
an
d
co
c
o
n
u
t
leaf
im
a
g
es
b
ef
o
r
e
th
e
class
if
icatio
n
p
r
o
ce
s
s
.
B
a
s
ed
o
n
th
e
Gau
s
s
ian
f
u
n
ctio
n
’
s
s
h
ap
e,
th
is
ap
p
r
o
ac
h
s
elec
ts
a
li
n
ea
r
f
ilter
th
r
o
u
g
h
a
weig
h
ted
v
alu
e
f
o
r
ea
ch
co
m
p
o
n
en
t.
T
h
e
f
ilter
i
s
ch
o
s
en
b
ec
au
s
e
it
ef
f
ec
tiv
ely
r
ef
in
es
th
e
im
ag
es
wh
ile
co
n
s
id
er
in
g
th
e
k
e
r
n
el
ce
n
t
er
.
T
h
e
v
alu
es
o
f
ea
ch
co
m
p
o
n
e
n
t
in
th
e
Gau
s
s
ian
s
m
o
o
th
in
g
f
ilter
ar
e
m
ath
em
atica
lly
co
m
p
u
te
d
u
s
in
g
(
1
)
,
wh
er
e
,
th
e
n
o
r
m
aliza
tio
n
co
n
s
tan
t
is
d
en
o
ted
as
,
an
d
th
e
Gau
s
s
ia
n
k
er
n
el’
s
s
tan
d
ar
d
d
ev
iatio
n
is
r
ep
r
esen
ted
as
.
ℎ
(
,
)
=
1
2
+
2
2
2
(
1
)
2
.
2
.
2
.
Da
t
a
a
ug
m
ent
a
t
io
n
T
h
e
d
e
n
o
is
ed
im
ag
e
is
th
en
p
ass
ed
o
n
to
d
ata
au
g
m
en
tatio
n
f
o
r
g
en
e
r
atin
g
m
o
d
if
ie
d
v
e
r
s
io
n
s
o
f
th
e
im
ag
es
to
ar
tific
ially
in
cr
ea
s
e
t
h
e
s
ize
o
f
th
e
d
ataset.
T
h
is
ad
d
itio
n
ally
h
elp
s
o
v
er
co
m
e
th
e
is
s
u
e
o
f
lack
o
f
d
ata,
alo
n
g
with
o
v
er
f
itti
n
g
,
an
d
also
im
p
r
o
v
es
th
e
m
o
d
els’
ca
p
ac
i
ty
f
o
r
g
en
er
aliza
tio
n
.
Af
ter
d
at
a
au
g
m
en
tatio
n
,
th
e
co
o
r
d
in
ates
o
f
a
p
o
in
t
a
r
e
o
b
t
ain
ed
u
s
in
g
th
e
tr
an
s
f
o
r
m
atio
n
m
atr
ix
.
T
h
e
p
r
esen
t
s
tu
d
y
e
v
alu
ates
u
s
in
g
t
h
e
im
ag
e
au
g
m
en
tatio
n
tech
n
iq
u
es
o
f
f
lip
p
in
g
,
s
h
if
tin
g
,
c
r
o
p
p
in
g
,
a
n
d
co
lo
r
ch
a
n
g
e
[
2
5
]
.
T
h
ese
p
r
e
-
p
r
o
ce
s
s
ed
im
ag
es
ar
e
f
u
r
th
er
tr
an
s
f
o
r
m
e
d
in
to
f
ea
tu
r
es
b
y
b
ein
g
f
ed
in
to
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
p
h
ase
wh
ich
h
elp
s
in
th
e
ac
cu
r
ate
class
if
icatio
n
o
f
d
is
ea
s
ed
im
ag
es.
2
.
3
.
F
e
a
t
ure
ex
t
r
a
ct
io
n
us
ing
ResNet
5
0
Af
ter
p
r
e
-
p
r
o
ce
s
s
in
g
,
th
e
au
g
m
en
ted
im
ag
es
ar
e
f
ed
as
in
p
u
t
in
to
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
p
h
ase
wh
ich
is
an
im
p
o
r
tan
t
s
tep
in
th
e
co
co
n
u
t
leaf
d
is
ea
s
e
p
r
ed
ictio
n
f
o
r
ef
f
icien
t
f
ea
tu
r
e
e
x
tr
ac
tio
n
.
Featu
r
e
ex
tr
ac
tio
n
h
elp
s
ex
tr
ac
t
s
ig
n
if
ica
n
t
in
f
o
r
m
ativ
e
an
d
d
is
cr
im
in
ativ
e
f
ea
t
u
r
es
th
at
in
c
r
ea
s
e
th
e
p
r
ed
ictio
n
p
er
f
o
r
m
a
n
ce
o
f
th
e
m
o
d
el.
Fu
r
th
er
m
o
r
e,
t
h
e
R
esN
et5
0
is
o
n
e
o
f
th
e
d
ee
p
n
eu
r
al
n
etwo
r
k
s
,
wh
er
e
th
e
ter
m
“
r
esid
u
al
n
etwo
r
k
s
”
is
u
s
ed
,
an
d
th
e
r
e
ar
e
5
0
lay
er
s
p
r
esen
t
in
th
e
R
esNet5
0
ar
ch
itectu
r
e
wh
ich
is
co
m
m
o
n
l
y
u
s
ed
f
o
r
f
ea
tu
r
e
ex
tr
ac
tio
n
.
T
h
e
r
ea
s
o
n
R
esNet
is
s
e
lecte
d
in
th
is
ca
s
e
is
th
at
it
p
r
esen
ts
th
e
id
ea
o
f
r
esid
u
al
u
n
its
,
wh
ich
m
ak
e
it
p
o
s
s
ib
le
f
o
r
d
ee
p
lay
e
r
s
t
o
d
ir
ec
tly
lear
n
f
r
o
m
s
h
allo
w
lay
er
s
,
th
er
eb
y
ea
s
in
g
th
e
p
r
o
ce
s
s
o
f
n
etwo
r
k
co
n
v
er
g
en
ce
.
I
t
also
p
o
s
s
ess
e
s
b
etter
lear
n
in
g
ca
p
ac
ity
with
im
p
r
o
v
is
ed
im
ag
e
r
ec
o
g
n
itio
n
ca
p
ab
ilit
ies
.
Mo
r
eo
v
er
,
it
f
ac
ilit
ates
th
e
tr
a
in
in
g
o
f
d
ee
p
n
etwo
r
k
s
b
y
u
s
in
g
s
k
ip
co
n
n
ec
tio
n
s
wh
ich
m
i
tig
ate
th
e
is
s
u
e
o
f
v
an
is
h
in
g
g
r
ad
ien
t.
I
n
r
elatio
n
to
Den
s
eNe
t,
th
e
R
esNet5
0
ar
c
h
itectu
r
e
is
s
im
p
ler
an
d
co
m
p
u
tatio
n
ally
ef
f
icien
t
.
T
h
e
R
esNet5
0
[
2
6
]
is
p
ar
tic
u
lar
ly
ef
f
ec
tiv
e
at
ex
tr
ac
tin
g
r
o
b
u
s
t
f
ea
tu
r
es
f
r
o
m
im
a
g
es
an
d
lear
n
in
g
th
ei
r
u
n
d
er
ly
i
n
g
s
tr
u
ctu
r
e.
Similar
t
o
th
e
s
tan
d
ar
d
C
NN,
th
e
R
esNet5
0
in
clu
d
es
p
o
o
lin
g
lay
er
s
,
co
n
v
o
lu
tio
n
lay
e
r
s
an
d
r
esid
u
al
b
lo
ck
s
,
wh
ich
ar
e
ce
n
tr
al
to
th
e
R
esNet
ar
ch
itectu
r
e
[
2
7
]
.
T
h
ese
r
esid
u
al
b
lo
c
k
s
aim
to
estab
lis
h
co
n
n
ec
tio
n
s
b
etwe
en
th
e
p
r
ed
i
ctio
n
s
an
d
th
e
ac
tu
al
i
n
p
u
t.
T
h
e
R
esNet5
0
ar
ch
itectu
r
e
is
ex
p
r
ess
ed
in
(
2
)
.
=
(
,
)
+
(
2
)
w
h
er
e
,
th
e
r
esid
u
al
b
lo
c
k
i
n
p
u
t
is
d
en
o
ted
as
,
co
n
v
o
lu
tio
n
al
lay
er
s
with
in
b
lo
c
k
weig
h
ts
a
r
e
d
e
n
o
ted
as
,
an
d
th
e
r
esid
u
al
b
lo
c
k
o
u
tp
u
t i
s
d
en
o
ted
as
.
T
h
e
f
u
n
ctio
n
o
f
c
o
n
v
o
lu
ti
o
n
al
la
y
er
s
with
in
th
e
b
lo
ck
is
r
ef
er
r
ed
to
as
(
,
)
.
I
t
tak
es
in
p
u
t
(
)
an
d
weig
h
t
(
)
as
ar
g
u
m
en
ts
an
d
ap
p
lies
d
if
f
er
en
t
o
p
e
r
atio
n
s
in
clu
d
i
n
g
b
at
c
h
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
3
,
J
u
n
e
20
25
:
2
8
3
4
-
2
8
4
4
2838
n
o
r
m
aliza
tio
n
,
ac
tiv
atio
n
s
,
an
d
co
n
v
o
lu
tio
n
s
.
T
h
e
R
esNet5
0
h
as
a
th
r
ee
-
ch
an
n
el
in
p
u
t
o
f
t
h
e
s
ize
2
2
4
×
2
2
4
.
I
t
b
eg
in
s
with
a
m
ax
-
p
o
o
lin
g
an
d
co
n
v
o
lu
tio
n
al
lay
er
with
3
×
3
an
d
7
×
7
k
er
n
el
s
izes,
r
esp
ec
tiv
ely
.
E
ac
h
co
n
v
o
l
u
tio
n
al
b
lo
c
k
in
th
e
R
esNet5
0
ar
ch
itectu
r
e
h
as
th
r
ee
co
n
v
o
l
u
tio
n
lay
er
s
,
an
d
ea
ch
i
d
en
tifie
d
b
lo
c
k
also
in
clu
d
es
th
r
ee
c
o
n
v
o
lu
tio
n
lay
er
s
.
T
h
e
f
u
lly
lin
k
ed
lay
er
wit
h
1
0
0
0
n
e
u
r
o
n
s
in
th
e
f
in
al
la
y
er
is
f
o
llo
wed
b
y
av
er
ag
e
p
o
o
lin
g
lay
e
r
an
d
th
e
p
r
ev
io
u
s
f
iv
e
s
tag
es.
T
h
e
n
etw
o
r
k
also
in
clu
d
es
a
f
u
lly
c
o
n
n
e
cted
lay
er
with
1
0
0
0
n
eu
r
o
n
s
in
th
e
f
in
al
lay
e
r
,
f
o
llo
wed
b
y
a
n
av
e
r
ag
e
p
o
o
lin
g
lay
er
an
d
th
e
p
r
o
ce
ed
i
n
g
f
i
v
e
s
tag
es.
I
n
th
is
r
esear
ch
,
R
esNet5
0
s
er
v
es
a
s
th
e
b
ase
m
o
d
el,
with
ad
d
itio
n
al
f
u
lly
c
o
n
n
ec
ted
lay
er
s
b
u
ilt
o
n
to
p
.
Fo
r
th
e
f
in
al
lay
er
,
1024
×
7
f
u
lly
co
n
n
ec
ted
lay
e
r
s
ar
e
s
u
b
s
titu
ted
.
Usi
n
g
th
is
R
esNet5
0
m
o
d
el,
2
0
4
8
f
ea
tu
r
es
ar
e
ex
tr
ac
ted
f
r
o
m
ea
ch
p
r
e
p
r
o
ce
s
s
ed
c
o
co
n
u
t
a
n
d
co
c
o
n
u
t
leaf
im
a
g
es.
R
esNet5
0
o
f
f
er
s
a
c
o
m
m
en
d
ab
l
e
p
er
f
o
r
m
an
ce
f
o
r
r
ec
o
g
n
izin
g
co
c
o
n
u
t
leaf
im
a
g
es.
T
h
en
,
o
p
tim
al
f
ea
tu
r
es
ar
e
p
ass
ed
as
in
p
u
t
to
th
e
class
if
icatio
n
s
tag
e,
a
n
d
th
ese
o
p
tim
al
f
ea
tu
r
es a
r
e
u
s
ed
to
in
cr
ea
s
e
th
e
class
if
icatio
n
o
u
tp
u
ts
.
3.
CL
AS
SI
F
I
CAT
I
O
N
US
I
NG
M
O
B
I
L
E
NE
T
V2
Fo
llo
win
g
th
e
f
ea
tu
r
e
ex
tr
ac
ti
o
n
,
t
h
e
Mo
b
ileNetV2
is
em
p
l
o
y
ed
to
ef
f
ec
tiv
ely
class
if
y
t
h
e
co
co
n
u
t
f
r
u
it
an
d
co
co
n
u
t
leaf
d
is
ea
s
es
b
y
u
tili
zin
g
th
e
ex
tr
ac
ted
s
ig
n
if
ican
t
f
ea
tu
r
es.
Mo
b
ileNetV2
is
u
tili
ze
d
f
o
r
p
r
ed
ictin
g
c
o
co
n
u
t
an
d
co
c
o
n
u
t
leaf
d
is
ea
s
es
an
d
is
ch
o
s
en
f
o
r
its
lig
h
tweig
h
t
a
r
ch
itectu
r
e
,
allo
win
g
ef
f
icien
t
im
p
lem
en
tatio
n
f
o
r
ac
cu
r
ate
i
m
ag
e
class
if
icatio
n
.
Mo
b
ileNetV2
[
2
8
]
p
o
s
s
ess
es
a
s
tr
o
n
g
an
d
p
o
r
tab
le
m
o
d
el
f
o
r
im
ag
e
class
if
icatio
n
task
s
.
T
h
e
co
m
b
in
atio
n
o
f
t
h
e
Mo
b
i
leNe
tV2
’s
ex
ce
llen
t
ac
cu
r
ac
y
an
d
its
ef
f
ec
tiv
e
ar
ch
itectu
r
e
m
a
k
e
s
it
an
id
ea
l
ch
o
ice
f
o
r
d
ev
ices
with
lim
ite
d
r
eso
u
r
ce
s
.
Ad
d
itio
n
ally
,
it
in
co
r
p
o
r
ates
s
h
o
r
tcu
t
co
n
n
ec
tio
n
s
b
etwe
en
b
o
ttlen
ec
k
s
to
im
p
r
o
v
e
g
r
ad
ien
t
f
lo
w
a
n
d
en
h
an
ce
t
r
ain
in
g
ef
f
icien
c
y
.
T
h
e
ar
c
h
itectu
r
e
o
f
Mo
b
ileNetV2
[
2
9
]
is
s
im
ilar
to
th
at
o
f
th
e
o
r
ig
in
al
Mo
b
ileNet,
with
th
e
m
ain
d
if
f
er
e
n
ce
b
ein
g
th
e
r
e
m
o
v
al
o
f
n
o
n
lin
ea
r
ities
in
n
ar
r
o
w
lay
er
s
,
an
d
th
e
ad
o
p
tio
n
o
f
in
v
er
ted
r
esid
u
al
b
lo
ck
s
with
b
o
ttlen
ec
k
in
g
ch
ar
ac
ter
is
tics
.
C
o
m
p
ar
ed
to
th
e
o
r
ig
i
n
al
Mo
b
ileNet,
Mo
b
ileNets h
as
f
ewe
r
p
ar
am
eter
s
an
d
h
an
d
les all
in
p
u
t sizes lar
g
er
th
an
32
×
3
2
p
ix
els
m
o
r
e
ef
f
icien
tly
,
p
er
f
o
r
m
in
g
b
etter
with
lar
g
e
i
m
ag
e
s
izes.
T
h
e
Mo
b
ileNetV2
co
n
tain
s
two
ty
p
es
o
f
b
lo
ck
s
:
a
r
esid
u
al
b
lo
ck
wi
th
a
s
tr
id
e
o
f
1
,
an
d
a
b
lo
c
k
with
a
s
tr
id
e
o
f
2
f
o
r
d
o
wn
s
i
z
in
g
.
T
h
er
e
ar
e
th
r
ee
lay
er
s
f
o
r
ea
ch
b
lo
ck
,
in
clu
d
i
n
g
th
e
f
i
r
s
t
lay
er
wh
ic
h
is
a
1
×
1
co
n
v
o
lu
tio
n
al
lay
er
with
R
eL
U6
,
f
o
llo
we
d
b
y
a
d
ep
th
-
wis
e
co
n
v
o
lu
tio
n
an
d
a
f
in
al
co
n
v
o
lu
tio
n
1
×
1
lay
e
r
,
w
ith
o
u
t a
n
y
ac
tiv
atio
n
f
u
n
ctio
n
.
T
h
e
Mo
b
ileNetV2
m
eth
o
d
is
u
s
ed
to
en
h
a
n
ce
th
e
s
y
s
tem
’
s
ca
p
ab
ilit
y
f
o
r
co
co
n
u
t
an
d
le
af
d
is
ea
s
e
class
if
icatio
n
.
I
n
r
elatio
n
to
th
e
ex
is
tin
g
m
eth
o
d
s
,
th
e
Mo
b
ileNetV2
en
ab
les
r
ea
l
-
tim
e
d
is
ea
s
e
d
etec
tio
n
in
r
eso
u
r
ce
-
co
n
s
tr
ain
ed
en
v
ir
o
n
m
en
ts
,
an
d
p
r
o
v
id
es
an
ef
f
ec
ti
v
e
b
ala
n
ce
b
etwe
en
c
o
m
p
u
tati
o
n
al
ef
f
icien
cy
an
d
ac
cu
r
ac
y
.
T
h
e
ac
cu
r
ac
y
o
f
M
o
b
ileNetV2
in
d
etec
tio
n
o
f
co
co
n
u
t
leaf
d
is
ea
s
es
is
f
u
r
th
er
im
p
r
o
v
e
d
th
r
o
u
g
h
o
p
tim
izatio
n
.
Giv
en
th
at
th
e
Mo
b
ileNetV2
h
as
a
co
m
p
licated
s
tr
u
ctu
r
e
an
d
m
u
ltip
le
p
ar
a
m
eter
s
,
o
v
er
f
itti
n
g
is
co
m
m
o
n
is
s
u
e,
esp
ec
ially
w
h
e
n
th
e
tr
ain
in
g
d
ataset
is
s
m
all.
T
o
m
itig
ate
o
v
er
f
itti
n
g
,
Mo
b
il
eNe
tV2
is
o
p
tim
ized
u
s
in
g
th
e
L
2
r
eg
u
lar
izatio
n
m
eth
o
d
,
wh
ich
h
elp
s
p
r
ev
en
t
o
v
er
f
itti
n
g
,
r
ed
u
ce
s
th
e
im
p
ac
t
o
f
co
m
p
lex
b
ac
k
g
r
o
u
n
d
f
ilter
n
o
is
e,
an
d
e
n
s
u
r
es
th
at
th
e
m
o
d
el
ca
n
ef
f
e
ctiv
ely
h
an
d
le
t
h
e
ca
teg
o
r
izatio
n
o
f
n
ew
d
ata.
3
.
1
.
L
2
re
g
ula
riza
t
io
n
L
2
r
eg
u
la
r
izatio
n
ad
d
s
a
r
eg
u
lar
izatio
n
item
(
p
en
alty
item
)
to
th
e
lo
s
s
eq
u
atio
n
f
o
r
p
r
ev
en
tin
g
th
e
m
o
d
el
f
r
o
m
o
v
er
f
itti
n
g
to
n
o
is
e,
in
clu
d
in
g
c
o
m
p
lex
b
ac
k
g
r
o
u
n
d
s
in
th
e
tr
ain
i
n
g
s
et.
I
t
also
r
ed
u
ce
s
th
e
m
o
d
el
co
m
p
lex
ity
b
y
lim
itin
g
th
e
w
eig
h
t
with
a
m
ax
im
u
m
n
u
m
b
er
o
f
p
a
r
am
eter
s
.
T
h
e
m
ea
n
s
q
u
ar
e
er
r
o
r
lo
s
s
f
u
n
ctio
n
an
d
th
e
o
r
i
g
in
al
lo
s
s
f
u
n
ctio
n
ar
e
em
p
lo
y
ed
d
u
r
i
n
g
th
e
tr
ain
in
g
p
r
o
ce
s
s
.
T
h
e
lo
s
s
f
u
n
ctio
n
is
d
en
o
ted
as
0
(
,
)
,
an
d
is
m
ath
em
atica
lly
ex
p
r
ess
ed
in
(
3
)
.
0
(
,
)
=
1
∑
(
′
(
)
,
(
)
)
=
1
(
3
)
w
h
er
e
,
th
e
co
s
t
f
u
n
ctio
n
is
r
ep
r
esen
t
ed
as
,
an
d
th
e
n
u
m
b
er
o
f
elem
en
ts
in
th
e
s
am
p
le
d
ataset
is
d
en
o
ted
as
.
T
h
e
am
o
u
n
t
o
f
b
ias
d
u
r
i
n
g
n
eu
r
o
n
tr
an
s
m
is
s
io
n
is
d
en
o
ted
as
.
T
h
e
n
eu
r
o
n
’
s
d
esire
d
o
u
tp
u
t
v
alu
e
is
r
ep
r
esen
t
ed
as
(
)
,
wh
ile
th
e
r
ea
l
v
alu
e
is
′
(
)
.
Fo
llo
win
g
th
e
r
eg
u
lar
izatio
n
ad
d
itio
n
,
0
(
,
)
is
n
o
t
o
p
tim
i
ze
d
d
ir
ec
tly
,
but
is
p
r
im
ar
ily
o
p
tim
ized
as
0
(
,
)
+
(
)
.
T
h
e
m
o
d
el
’s
co
m
p
lex
ity
is
d
escr
ib
ed
b
y
th
e
p
e
n
alty
ter
m
o
r
r
eg
u
lar
izatio
n
ter
m
d
en
o
ted
as
(
)
,
wh
ich
is
ca
lcu
lated
u
s
in
g
(
4
)
.
W
h
er
e
,
th
e
n
u
m
b
er
o
f
f
u
lly
co
n
n
ec
ted
lay
e
r
s
is
r
ep
r
esen
ted
as
,
an
d
th
e
th
n
eu
r
o
n
’
s
weig
h
t
is
d
en
o
te
d
a
s
.
T
h
e
e
x
p
r
ess
io
n
o
f
th
e
r
eg
u
lar
ized
l
o
s
s
f
u
n
ctio
n
is
r
e
p
r
esen
ted
in
(
5
)
.
(
)
=
|
|
|
|
2
=
∑
2
=
1
(
4
)
(
,
)
=
1
∑
(
′
(
)
,
(
)
)
=
1
+
2
∑
2
=
1
(
5
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
p
r
ed
ictio
n
o
f c
o
co
n
u
t a
n
d
c
o
co
n
u
t le
a
f d
is
ea
s
e
u
s
in
g
Mo
b
ileN
etV
2
…
(
K
a
vith
a
Ma
g
a
d
i
Go
p
a
la
kris
h
n
a
)
2839
w
h
er
e
,
th
e
r
eg
u
lar
izatio
n
f
ac
to
r
is
in
d
icate
d
b
y
.
I
n
th
e
e
x
p
er
i
m
en
t,
th
e
o
u
t
p
u
t
a
n
d
t
h
e
all
th
e
co
n
n
e
ctio
n
la
y
er
s
ar
e
o
p
tim
ized
u
s
in
g
th
e
L
2
r
eg
u
lar
izatio
n
tech
n
iq
u
e
.
Af
ter
cr
o
s
s
-
v
alid
atio
n
,
is
as
s
u
m
ed
to
b
e
0
.
0
0
4
[
3
0
]
.
T
h
e
r
esu
lts
an
d
an
aly
s
is
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
ar
e
g
iv
en
b
elo
w
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
f
o
llo
win
g
s
y
s
tem
s
etu
p
is
u
s
ed
f
o
r
p
r
o
ce
s
s
in
g
th
e
s
im
u
la
tio
n
d
ata
f
o
r
class
if
icatio
n
an
d
p
r
ed
ictio
n
of
co
co
n
u
t
leaf
d
is
ea
s
e
s
.
T
h
e
An
ac
o
n
d
a
Nav
ig
at
o
r
v
er
s
io
n
o
f
3
.
5
.
2
.
0
is
u
s
ed
with
Py
th
o
n
3
.
1
0
,
with
th
e
s
y
s
tem
s
p
ec
if
icatio
n
s
b
ein
g
:
1
6
GB
R
AM
,
W
in
d
o
ws
1
0
(
6
4
-
b
it)
o
p
er
atio
n
s
y
s
tem
,
an
d
I
n
tel
C
o
r
e
i7
p
r
o
ce
s
s
o
r
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
’s
ef
f
ec
ti
v
e
n
ess
is
an
aly
ze
d
in
ter
m
s
o
f
ac
cu
r
ac
y
,
F
1
-
s
co
r
e,
p
r
ec
is
io
n
,
an
d
r
ec
all.
T
h
e
m
ath
em
atica
l
r
ep
r
esen
tatio
n
s
o
f
th
ese
p
ar
am
ete
r
s
ar
e
r
ep
r
esen
ted
in
(
6
)
-
(
9
)
.
=
+
+
+
+
(
6
)
1
−
=
2
2
+
+
(
7
)
=
+
(
8
)
=
+
(
9
)
w
h
er
e,
d
en
o
tes
tr
u
e
p
o
s
itiv
e,
tr
u
e
n
eg
ativ
e
r
ep
r
esen
ts
,
f
al
s
e
p
o
s
itiv
e
r
ef
er
s
to
F
,
an
d
f
al
s
e
n
eg
ativ
e
d
en
o
tes
.
4
.
1
.
Q
ua
ntit
a
t
iv
e
a
nd
qu
a
lita
t
iv
e
a
na
ly
s
is
T
h
e
p
er
f
o
r
m
an
ce
ev
alu
atio
n
o
f
Mo
b
ileNetV2
is
co
m
p
u
ted
in
ter
m
s
o
f
r
ec
all,
p
r
ec
is
io
n
,
F1
-
s
co
r
e,
an
d
ac
cu
r
ac
y
w
h
ich
a
r
e
d
is
p
lay
e
d
in
T
ab
les
1
an
d
2
.
T
h
e
im
p
le
m
en
ted
Mo
b
ileNetV2
m
o
d
el
i
s
co
m
p
ar
e
d
with
t
h
e
ex
is
tin
g
s
tate
-
of
-
th
e
-
ar
t
m
eth
o
d
s
n
am
ely
,
C
NN,
R
s
Net5
0
,
an
d
VGG1
9
.
T
h
e
R
esNet5
0
m
eth
o
d
is
co
m
p
a
r
ed
with
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
m
eth
o
d
s
:
I
n
ce
p
tio
n
v
3
,
C
NN
an
d
VGG1
6
.
T
h
e
an
aly
s
is
o
f
R
esNet5
0
m
eth
o
d
in
ter
m
s
o
f
r
ec
all,
p
r
ec
is
io
n
,
F1
-
s
co
r
e,
an
d
ac
cu
r
ac
y
ar
e
r
e
p
r
esen
ted
in
T
ab
les
3
an
d
4
.
T
h
e
Mo
b
ileNetV2
au
to
m
ati
ca
lly
p
r
ed
ict
s
co
co
n
u
t
an
d
co
co
n
u
t
leaf
d
is
ea
s
e
im
ag
es
,
as
o
p
p
o
s
ed
to
o
t
h
er
m
eth
o
d
s
.
T
h
e
im
p
lem
en
ted
Mo
b
ileNetV2
p
er
f
o
r
m
s
ex
ce
p
tio
n
ally
well
in
th
e
class
if
icati
o
n
o
f
p
lan
t
leaf
d
is
ea
s
es
d
u
e
to
its
ad
ap
tab
ilit
y
to
ex
ten
s
iv
e
im
ag
e
v
ar
iab
ilit
y
,
a
n
d
ef
f
icien
t
ex
tr
ac
tio
n
an
d
p
ar
am
eter
m
an
a
g
em
en
t
ca
p
ab
il
ities
.
Ad
d
itio
n
ally
,
f
aster
tr
ain
in
g
an
d
less
m
em
o
r
y
co
n
s
u
m
p
tio
n
ar
e
also
m
ad
e
p
o
s
s
ib
le
d
u
e
to
th
e
ex
ce
p
tio
n
al
ac
cu
r
ac
y
o
f
th
e
m
o
d
el.
T
h
e
M
o
b
il
eNe
tV2
is
c
o
m
m
en
d
a
b
le
at
ca
p
tu
r
i
n
g
elem
en
ts
n
ec
ess
ar
y
f
o
r
d
is
ea
s
e
c
ate
g
o
r
izatio
n
at
b
o
th
lo
w
an
d
h
ig
h
lev
els.
Ad
v
a
n
ce
d
r
eg
u
la
r
izatio
n
m
eth
o
d
s
a
r
e
in
co
r
p
o
r
ated
t
o
p
r
e
v
en
t
o
v
er
f
itti
n
g
with
th
e
g
u
ar
an
tee
th
at
th
e
m
o
d
el
p
er
f
o
r
m
s
ef
f
ec
tiv
ely
wh
en
ap
p
lied
o
n
f
r
esh
,
u
n
test
ed
d
ata.
T
h
e
Mo
b
ileNetV2
'
s
s
u
p
er
io
r
p
e
r
f
o
r
m
an
ce
i
n
ac
cu
r
atel
y
class
if
y
in
g
co
co
n
u
t
leaf
d
is
ea
s
es
i
s
attr
ib
u
ted
to
its
o
p
ti
m
ized
d
esig
n
,
a
b
ilit
y
to
h
an
d
le
a
v
a
r
iety
o
f
leaf
c
h
ar
ac
ter
is
tics
,
as
well
as
th
e
r
eg
u
lar
izatio
n
m
eth
o
d
s
.
T
h
is
m
ak
es
it
a
v
alu
a
b
le
r
eso
u
r
ce
f
o
r
ag
r
ic
u
ltu
r
al
an
d
d
is
ea
s
e
m
an
ag
em
en
t p
r
o
p
o
s
als
.
T
ab
le
1
d
is
p
lay
s
th
e
p
er
f
o
r
m
an
ce
o
f
Mo
b
ileNetV2
o
n
th
e
co
co
n
u
t
im
ag
e
d
ataset
with
d
if
f
er
en
t
class
if
ier
s
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
R
esNet5
0
,
C
NN,
an
d
VGG1
9
ar
e
m
ea
s
u
r
ed
an
d
m
atc
h
ed
with
th
at
o
f
th
e
im
p
lem
en
ted
Mo
b
ileNetV2
m
o
d
el.
T
h
e
o
b
tain
e
d
o
u
tco
m
es e
x
h
ib
it
th
at
th
e
im
p
lem
en
ted
Mo
b
ileNetV2
m
o
d
el
o
u
tp
er
f
o
r
m
s
o
t
h
er
m
o
d
els
with
r
ec
all,
p
r
ec
is
io
n
,
F1
-
s
co
r
e,
an
d
ac
cu
r
ac
y
v
alu
es
o
f
ab
o
u
t
9
9
.
9
9
%,
9
9
.
9
9
%,
9
9
.
9
9
%,
a
n
d
9
9
.
9
9
%,
r
esp
ec
tiv
ely
.
T
ab
le
1
.
Per
f
o
r
m
an
ce
an
aly
s
is
o
f
Mo
b
ileNetV2
with
d
i
f
f
er
e
n
t c
lass
if
ier
s
o
n
C
o
co
n
u
t im
ag
e
d
ataset
D
a
t
a
s
e
t
s
C
l
a
s
si
f
i
e
r
s
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
C
o
c
o
n
u
t
i
ma
g
e
R
e
sN
e
t
5
0
6
0
.
0
0
3
0
.
0
0
3
7
.
0
0
6
0
.
0
0
C
N
N
8
7
.
0
0
9
1
.
0
0
8
3
.
0
0
8
5
.
0
0
V
G
G
1
9
9
4
.
0
0
9
5
.
0
0
9
3
.
0
0
9
4
.
0
0
M
o
b
i
l
e
N
e
t
V
2
9
9
.
9
9
9
9
.
9
9
9
9
.
9
9
9
9
.
9
9
T
ab
le
2
d
em
o
n
s
tr
ates
th
e
p
er
f
o
r
m
an
ce
o
f
R
esNet5
0
o
n
t
h
e
co
co
n
u
t
im
ag
e
d
ataset
with
d
i
f
f
er
en
t
FE
m
eth
o
d
s
.
T
h
e
p
e
r
f
o
r
m
an
ce
m
e
tr
ics
o
f
I
n
ce
p
tio
n
v
3
,
C
NN,
an
d
VGG1
6
ar
e
co
m
p
a
r
ed
with
t
h
e
R
esNet5
0
m
o
d
el.
T
h
e
ac
co
m
p
lis
h
ed
o
u
tco
m
es
r
ev
ea
l
th
at
th
e
R
esNet5
0
m
o
d
e
l
ac
h
iev
es
s
u
p
er
io
r
f
ea
tu
r
e
ex
t
r
ac
tio
n
r
esu
lts
with
r
ec
all,
p
r
ec
is
io
n
,
F1
-
s
co
r
e,
an
d
ac
cu
r
ac
y
v
alu
es o
f
ab
o
u
t 9
9
.
9
9
%,
9
9
.
9
9
%,
9
9
.
9
9
%,
a
n
d
9
9
.
9
9
%,
r
esp
ec
tiv
ely
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
3
,
J
u
n
e
20
25
:
2
8
3
4
-
2
8
4
4
2840
T
ab
le
2
.
Per
f
o
r
m
an
ce
an
aly
s
is
o
f
R
esNet5
0
with
d
if
f
er
e
n
t FE
m
eth
o
d
s
o
n
co
c
o
n
u
t
im
a
g
e
d
ataset
D
a
t
a
s
e
t
s
M
e
t
h
o
d
s
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
C
o
c
o
n
u
t
i
ma
g
e
I
n
c
e
p
t
i
o
n
v
3
6
2
.
0
0
3
5
.
0
0
3
6
.
0
0
6
1
.
0
0
C
N
N
8
8
.
0
0
9
1
.
0
0
8
3
.
0
0
8
5
.
0
0
V
G
G
1
6
9
4
.
0
0
9
5
.
0
0
9
3
.
0
0
9
4
.
0
0
R
e
sN
e
t
5
0
9
9
.
9
9
9
9
.
9
9
9
9
.
9
9
9
9
.
9
9
T
ab
le
3
d
is
p
lay
s
th
e
ef
f
ec
tiv
e
n
ess
o
f
Mo
b
ileNetV2
o
n
th
e
c
o
co
n
u
t
leaf
im
a
g
e
d
ataset
with
d
if
f
er
e
n
t
class
if
ier
s
.
T
h
e
p
er
f
o
r
m
an
ce
m
etr
ics
o
f
R
esNet5
0
,
C
NN,
an
d
VGG1
9
ar
e
m
ea
s
u
r
ed
an
d
co
n
tr
asted
ag
ain
s
t
th
at
o
f
th
e
im
p
lem
en
ted
Mo
b
ileNetV2
m
o
d
el.
T
h
e
o
b
tain
ed
o
u
tc
o
m
es
s
u
r
p
ass
th
e
ex
is
tin
g
m
o
d
el
s
in
ter
m
s
o
f
r
ec
all,
p
r
ec
is
io
n
,
F1
-
s
co
r
e,
an
d
ac
c
u
r
ac
y
,
with
co
r
r
esp
o
n
d
i
n
g
v
al
u
e
s
o
f
ab
o
u
t
9
9
.
5
6
%,
9
9
.
4
%,
9
9
.
4
8
%,
an
d
9
9
.
2
%.
T
ab
le
3
.
Per
f
o
r
m
an
ce
an
aly
s
is
o
f
Mo
b
ileNetV2
with
d
i
f
f
er
e
n
t c
lass
if
ier
s
o
n
co
co
n
u
t
lea
f
i
m
ag
e
d
ataset
D
a
t
a
s
e
t
s
C
l
a
s
si
f
i
e
r
s
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
C
o
c
o
n
u
t
l
e
a
f
i
ma
g
e
R
e
sN
e
t
5
0
4
5
.
7
8
4
6
.
0
0
4
5
.
0
0
4
5
.
5
0
C
N
N
3
0
.
5
4
3
0
.
0
0
2
9
.
0
0
2
9
.
5
0
V
G
G
1
9
8
9
.
8
6
9
0
.
0
0
8
8
.
0
0
8
9
.
0
0
M
o
b
i
l
e
N
e
t
9
9
.
2
9
9
.
4
9
9
.
5
6
9
9
.
4
8
T
ab
le
4
p
r
esen
ts
th
e
o
u
tco
m
es
o
f
R
esNet5
0
o
n
t
h
e
co
c
o
n
u
t
leaf
im
a
g
e
d
ataset
with
d
i
f
f
er
en
t
FE
m
eth
o
d
s
.
T
h
e
m
etr
ics
o
f
th
e
s
tate
-
of
-
th
e
-
ar
t
m
eth
o
d
s
n
am
el
y
,
I
n
ce
p
tio
n
v
3
,
C
NN,
an
d
VGG1
6
ar
e
co
m
p
ar
e
d
with
th
at
o
f
t
h
e
R
esNet5
0
m
o
d
el.
T
h
e
attain
ed
o
u
tco
m
e
s
v
alid
ate
th
e
R
esNet5
0
m
o
d
el
b
ased
o
n
r
ec
all
,
p
r
ec
is
io
n
,
F1
-
s
co
r
e,
a
n
d
ac
c
u
r
ac
y
o
f
ab
o
u
t
9
9
.
5
6
%,
9
9
.
4
%,
9
9
.
4
8
%,
an
d
9
9
.
2
%,
th
e
r
eb
y
p
r
o
v
in
g
s
u
p
e
r
io
r
to
o
th
er
FE
m
eth
o
d
s
.
T
ab
le
4
.
Per
f
o
r
m
an
ce
an
aly
s
is
o
f
R
esNet5
0
with
d
if
f
er
e
n
t FE
m
eth
o
d
s
o
n
co
c
o
n
u
t
leaf
im
ag
e
d
ataset
D
a
t
a
s
e
t
s
M
e
t
h
o
d
s
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
C
o
c
o
n
u
t
l
e
a
f
i
ma
g
e
I
n
c
e
p
t
i
o
n
v
3
4
5
.
7
8
4
6
.
7
4
5
.
5
6
4
6
.
0
0
C
N
N
3
1
.
5
4
3
0
.
5
4
2
8
.
9
3
0
.
0
0
V
G
G
1
6
8
8
.
9
9
1
.
0
0
8
9
.
0
0
9
0
.
0
0
R
e
sN
e
t
5
0
9
9
.
2
9
9
.
4
9
9
.
5
6
9
9
.
4
8
Fig
u
r
e
4
illu
s
tr
ates
th
e
co
n
f
u
s
io
n
m
atr
ix
f
o
r
Mo
b
ileNetV2
o
n
th
e
co
co
n
u
t
im
ag
e
d
ataset,
wh
er
e
th
e
R
OC
cu
r
v
e
is
d
ep
lo
y
ed
f
o
r
c
o
m
p
u
tin
g
th
e
class
if
ier
’
s
p
er
f
o
r
m
an
ce
.
I
n
th
e
co
n
f
u
s
io
n
m
atr
ix
,
0
d
en
o
tes
th
e
h
ea
lth
y
co
co
n
u
t
an
d
1
d
en
o
te
s
th
e
in
f
ec
ted
co
co
n
u
t.
T
h
e
i
m
p
lem
en
ted
m
o
d
el’
s
r
esu
lts
a
r
e
p
r
esen
ted
in
th
is
s
ec
tio
n
,
an
d
th
e
c
o
co
n
u
t
d
is
ea
s
e
p
er
f
o
r
m
an
ce
p
r
ed
ictio
n
is
ex
ec
u
ted
b
ased
o
n
th
e
Mo
b
ile
NetV2
m
o
d
el.
T
h
e
m
o
d
el’
s
class
if
icatio
n
ac
cu
r
ac
y
d
u
r
in
g
tr
ain
in
g
a
n
d
v
alid
atio
n
o
n
th
e
co
c
o
n
u
t
im
ag
e
d
ataset
is
s
h
o
w
n
in
Fig
u
r
e
5
,
wh
er
e
th
e
ac
cu
r
ac
y
is
s
ee
n
to
d
if
f
e
r
b
etwe
en
0
.
4
to
1
.
T
h
e
t
r
ain
in
g
an
d
v
alid
atio
n
lo
s
s
o
f
Mo
b
ileNetV2
o
n
th
e
c
o
co
n
u
t
im
ag
e
d
ataset
is
illu
s
tr
ated
in
Fig
u
r
e
6
.
T
h
e
tr
ain
in
g
ac
cu
r
ac
y
attain
ed
is
o
b
s
er
v
ed
to
b
e
0
.
9
9
,
as sh
o
wn
in
Fig
u
r
e
5
.
Fig
u
r
e
4
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
Mo
b
ileNetV2
u
s
in
g
co
co
n
u
t
im
a
g
e
d
ataset
Fig
u
r
e
5
.
Acc
u
r
ac
y
tr
ain
i
n
g
an
d
v
alid
atio
n
’
s
g
r
ap
h
ical
r
e
p
r
esen
tatio
n
o
n
co
co
n
u
t
im
ag
e
d
ataset
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
p
r
ed
ictio
n
o
f c
o
co
n
u
t a
n
d
c
o
co
n
u
t le
a
f d
is
ea
s
e
u
s
in
g
Mo
b
ileN
etV
2
…
(
K
a
vith
a
Ma
g
a
d
i
Go
p
a
la
kris
h
n
a
)
2841
Fig
u
r
e
7
illu
s
tr
ates
th
e
co
n
f
u
s
io
n
m
atr
ix
f
o
r
Mo
b
ileNetV2
o
n
th
e
co
c
o
n
u
t
leaf
im
a
g
e
d
at
aset.
I
n
th
e
co
n
f
u
s
io
n
m
atr
i
x
,
0
r
ep
r
esen
ts
C
ater
p
illar
s
,
1
r
ep
r
esen
ts
L
e
af
lets
,
2
in
d
icate
s
Dr
y
in
g
o
f
L
ea
f
lets
,
3
in
d
icate
s
Flaccid
ity
,
an
d
4
i
n
d
icate
s
Ye
llo
win
g
.
T
h
e
m
o
d
el’
s
r
esu
lts
an
d
p
r
ed
ictio
n
ac
cu
r
ac
y
o
f
c
o
co
n
u
t
d
is
ea
s
es
ar
e
p
r
esen
ted
in
th
is
s
ec
tio
n
.
T
h
is
in
v
o
lv
es
a
b
in
ar
y
class
if
icatio
n
task
to
ca
teg
o
r
ize
co
co
n
u
t
leav
es
as
eith
er
in
f
ec
ted
o
r
h
ea
lth
y
.
Fig
u
r
e
8
d
is
p
lay
s
th
e
tr
ain
in
g
an
d
v
ali
d
atio
n
ac
cu
r
ac
y
o
f
th
e
Mo
b
ile
NetV2
m
o
d
el
u
s
in
g
th
e
co
co
n
u
t
leaf
im
ag
e
d
ataset,
with
ac
cu
r
ac
y
v
al
u
es
r
a
n
g
in
g
f
r
o
m
0
.
9
5
5
t
o
0
.
9
9
5
.
T
h
e
tr
ain
in
g
ac
c
u
r
ac
y
is
r
ep
r
esen
ted
in
b
lu
e,
wh
ile
th
e
v
alid
atio
n
ac
cu
r
ac
y
is
s
h
o
wn
i
n
r
ed
.
Fig
u
r
e
9
p
r
esen
ts
th
e
tr
a
in
in
g
an
d
v
alid
atio
n
lo
s
s
o
f
th
e
Mo
b
ileNetV2
m
o
d
el
o
v
er
1
0
ep
o
ch
s
u
s
in
g
th
e
co
co
n
u
t
leaf
im
ag
e
d
ataset
.
T
h
e
tr
a
in
in
g
an
d
v
alid
atio
n
lo
s
s
es r
an
g
e
b
etwe
e
n
0
.
0
2
an
d
0
.
1
6
,
wh
ile
th
e
ac
c
u
r
ac
y
is
0
.
9
9
,
as sh
o
wn
i
n
Fig
u
r
e
8
.
Fig
u
r
e
6
.
L
o
s
s
tr
ain
in
g
an
d
v
al
id
atio
n
’
s
g
r
ap
h
ical
r
ep
r
esen
tatio
n
o
n
co
c
o
n
u
t
im
a
g
e
d
ataset
Fig
u
r
e
7
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
Mo
b
ileNetV2
u
s
in
g
co
co
n
u
t
leaf
im
ag
e
d
ataset
Fig
u
r
e
8
.
Gr
a
p
h
ical
d
e
p
ictio
n
o
f
tr
ain
in
g
an
d
v
alid
atio
n
ac
cu
r
ac
y
o
n
co
c
o
n
u
t
leaf
im
ag
e
d
ataset
Fig
u
r
e
9
.
Gr
a
p
h
ical
r
ep
r
esen
tatio
n
tr
ain
in
g
a
n
d
v
alid
atio
n
lo
s
s
o
n
c
o
co
n
u
t
leaf
im
ag
e
d
ataset
4
.
2
.
Co
m
pa
ra
t
iv
e
a
na
ly
s
is
T
h
e
ef
f
ec
tiv
e
n
ess
o
f
t
h
e
class
if
ier
is
ev
alu
ated
b
y
co
m
p
ar
in
g
i
t
with
th
e
e
x
is
tin
g
m
eth
o
d
s
in
[
1
9
]
–
[
2
1
]
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
is
ass
ess
ed
u
s
in
g
b
o
th
th
e
c
o
co
n
u
t
im
ag
e
an
d
c
o
co
n
u
t
leaf
im
ag
e
d
atasets
.
Fo
r
th
e
co
co
n
u
t
im
ag
e
d
ataset,
th
e
ac
cu
r
ac
y
is
9
9
.
9
9
%,
F1
-
s
co
r
e
is
9
9
.
9
9
%,
r
ec
all
is
9
9
.
9
9
%,
an
d
p
r
ec
is
io
n
is
9
9
.
9
9
%.
Similar
ly
,
f
o
r
th
e
c
o
co
n
u
t
leaf
im
ag
e
d
ataset
,
th
e
ac
cu
r
ac
y
is
9
9
.
2
%,
F1
-
s
co
r
e
is
9
9
.
4
8
%,
r
e
ca
ll
is
9
9
.
5
6
% a
n
d
p
r
ec
is
io
n
is
9
9
.
4
%.
T
a
b
le
5
r
e
p
r
esen
ts
th
e
co
m
p
a
r
ativ
e
an
al
y
s
is
o
f
p
r
o
p
o
s
ed
m
eth
o
d
with
ex
is
tin
g
alg
o
r
ith
m
s
.
T
ab
le
5
.
C
o
m
p
a
r
ativ
e
an
aly
s
is
o
f
p
r
o
p
o
s
ed
an
d
e
x
is
tin
g
m
et
h
o
d
s
M
e
t
h
o
d
s
D
a
t
a
s
e
t
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
Ef
f
i
c
i
e
n
t
D
e
t
-
D
2
[
1
9
]
P
l
a
n
t
v
i
l
l
a
g
e
d
a
t
a
s
e
t
&
l
e
a
v
e
s
i
ma
g
e
d
a
t
a
s
e
t
-
8
8
.
1
3
9
4
.
0
6
9
0
.
9
9
DL
-
W
D
M
[
2
0
]
C
o
c
o
n
u
t
t
r
e
e
i
m
a
g
e
d
a
t
a
s
e
t
9
5
.
7
1
N
/
A
N
/
A
N
/
A
M
I
N
-
S
V
M
[
2
1
]
C
o
c
o
n
u
t
t
r
e
e
i
m
a
g
e
d
a
t
a
s
e
t
9
5
.
3
5
N
/
A
N
/
A
N
/
A
M
o
b
i
l
e
N
e
t
V
2
C
o
c
o
n
u
t
i
ma
g
e
9
9
.
9
9
9
9
.
9
9
9
9
.
9
9
9
9
.
9
9
C
o
c
o
n
u
t
l
e
a
f
i
m
a
g
e
9
9
.
2
9
9
.
4
9
9
.
5
6
9
9
.
4
8
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
3
,
J
u
n
e
20
25
:
2
8
3
4
-
2
8
4
4
2842
4
.
3
.
Dis
cus
s
io
n
I
n
th
is
r
esear
ch
,
co
co
n
u
t
an
d
c
o
co
n
u
t
lea
f
im
ag
e
d
atasets
ar
e
ex
p
lo
ited
to
an
aly
ze
t
h
e
ef
f
ec
ti
v
en
ess
o
f
th
e
p
r
o
p
o
s
ed
Mo
b
ileNetV2
b
y
co
m
p
ar
in
g
with
th
e
ex
is
tin
g
m
o
d
els
:
E
f
f
icien
tDet
-
D2
[
1
9
]
,
DL
-
W
DM
[
2
0
]
,
MI
N
-
SVM
[
2
1
]
,
PC
R
[
2
2
]
,
an
d
co
r
o
n
atin
e
-
in
s
en
s
itiv
e
1
b
g
e
n
e
(
C
OI
1
b
)
[
2
3
]
.
T
h
e
f
ea
tu
r
e
v
ec
to
r
s
ar
e
ex
tr
ac
ted
u
s
in
g
th
e
R
esNet5
0
tech
n
iq
u
e
f
o
r
h
ig
h
ca
teg
o
r
izatio
n
.
T
h
e
ac
cu
r
ac
y
o
f
th
e
co
c
o
n
u
t
le
af
im
ag
e
d
ataset
is
im
p
r
o
v
e
d
to
9
9
.
2
%,
with
an
F1
-
s
co
r
e
o
f
9
9
.
4
8
%,
r
ec
all
o
f
9
9
.
5
6
%
,
an
d
p
r
ec
is
io
n
o
f
9
9
.
4
%,
as
o
p
p
o
s
ed
to
th
e
ex
is
tin
g
E
f
f
icien
tDet
-
D2
[
1
9
]
,
DL
-
W
DM
[
2
0
]
an
d
MI
N
-
SVM
[
2
1
]
.
Desp
ite
th
is
,
th
e
ex
is
tin
g
E
f
f
i
cien
tDet
-
D2
[
1
9
]
m
eth
o
d
r
e
q
u
ir
es
m
in
im
al
s
to
r
ag
e
s
p
ac
e
,
m
ak
in
g
it
s
u
itab
le
f
o
r
in
teg
r
atio
n
in
t
o
s
m
ar
tp
h
o
n
es
o
r
d
r
o
n
es
with
li
m
ited
r
eso
u
r
ce
s
.
Similar
ly
,
th
e
PC
R
[
2
2
]
m
eth
o
d
d
o
es
n
o
t
id
en
tify
p
ath
o
g
en
s
in
an
y
s
am
p
le
d
ata,
th
u
s
m
ak
in
g
it
in
ef
f
ec
tiv
e
f
o
r
p
r
ed
ictin
g
d
i
s
ea
s
e
-
f
r
ee
r
eg
io
n
s
.
Ad
d
itio
n
ally
,
th
e
C
OI
1
b
[
2
3
]
m
o
d
el
d
o
es
n
o
t
id
en
tify
p
ath
o
g
en
s
in
an
y
s
am
p
le
d
ata,
th
u
s
m
ak
in
g
it
in
e
f
f
ec
tiv
e
f
o
r
p
r
ed
ictin
g
d
is
ea
s
e
-
f
r
ee
r
e
g
io
n
s
.
Ad
d
itio
n
ally
,
th
e
Ad
d
i
tio
n
ally
,
th
e
C
OI
1
b
m
o
d
el
[
2
3
]
d
o
es
n
o
t
an
aly
ze
p
o
ten
tial
b
r
ee
d
in
g
ap
p
r
o
ac
h
es
an
d
f
o
c
u
s
es
o
n
a
tim
e
-
co
u
r
s
e
in
v
esti
g
atio
n
o
f
g
en
e
ex
p
r
e
s
s
io
n
in
a
s
ig
n
alin
g
co
n
tex
t.
I
n
co
m
p
ar
is
o
n
to
th
e
s
e
ex
is
tin
g
m
o
d
els
[
2
2
]
,
[
2
3
]
,
th
e
p
r
o
p
o
s
ed
Mo
b
ileNetV2
o
f
f
er
s
a
lig
h
tweig
h
t
ar
ch
itectu
r
e
s
p
ec
if
ically
d
esig
n
ed
f
o
r
m
o
b
ile
d
ev
ices.
T
h
is
en
ab
les
r
ea
l
-
tim
e
im
ag
e
p
r
o
ce
s
s
in
g
with
lo
w
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
,
p
r
o
v
id
in
g
co
m
p
etitiv
e
ac
cu
r
ac
y
lev
els
an
d
m
ak
in
g
d
is
ea
s
e
m
o
n
ito
r
in
g
m
o
r
e
ef
f
ec
t
iv
e
an
d
ac
ce
s
s
ib
le,
p
ar
ticu
lar
ly
i
n
en
v
ir
o
n
m
en
ts
with
lim
ited
r
eso
u
r
ce
s
.
Fro
m
th
e
o
v
er
all
r
esu
lt
an
aly
s
is
,
it
is
o
b
s
er
v
ed
th
at
th
e
ex
is
tin
g
E
f
f
icien
tDet
-
D2
[
1
9
]
o
b
tain
s
a
p
r
ec
is
io
n
o
f
8
8
.
1
3
%,
r
ec
all
o
f
9
4
.
0
6
%
an
d
F1
-
s
co
r
e
o
f
9
9
.
4
8
%,
o
n
th
e
Plan
t
Villag
e
Data
s
e
t
an
d
L
ea
v
es
I
m
ag
e
d
ataset.
W
h
ile
th
e
ex
is
tin
g
DL
-
W
DM
[
2
0
]
an
d
MI
N
-
SVM
[
2
1
]
m
eth
o
d
s
attain
an
ac
c
u
r
ac
y
o
f
9
5
.
7
1
%
an
d
9
5
.
3
5
%,
r
esp
ec
tiv
ely
o
n
t
h
e
co
co
n
u
t
tr
ee
im
a
g
e
d
ataset.
I
t
is
h
en
ce
clea
r
th
at
th
e
p
r
o
p
o
s
ed
Mo
b
ileNetV2
o
u
tp
er
f
o
r
m
s
th
e
ex
is
tin
g
m
eth
o
d
s
with
all
p
er
f
o
r
m
a
n
ce
m
etr
ics.
T
h
e
Mo
b
ileNetV2
ap
p
r
o
a
ch
f
o
r
c
o
co
n
u
t
lea
f
d
is
ea
s
e
p
r
ed
ictio
n
ac
h
iev
es
a
n
in
cr
ea
s
ed
ac
cu
r
ac
y
o
f
9
9
.
9
9
%,
F1
-
s
co
r
e
o
f
9
9
.
9
9
%,
r
ec
all
o
f
9
9
.
9
9
%
an
d
p
r
ec
is
io
n
o
f
9
9
.
9
9
%
on
th
e
c
o
co
n
u
t
im
ag
e
d
ataset
.
Desp
ite
b
ein
g
lig
h
tweig
h
t,
ce
r
tain
d
o
m
ain
-
s
p
ec
if
ic
task
s
r
eq
u
ir
e
m
o
r
e
co
m
p
lex
r
ep
r
ese
n
tatio
n
s
with
g
r
ea
ter
c
o
m
p
u
tat
io
n
al
r
eso
u
r
ce
s
.
Ad
d
itio
n
ally
,
d
u
e
to
th
e
in
h
er
e
n
t
ac
cu
r
ac
y
tr
a
d
e
-
o
f
f
s
,
Mo
b
ileNe
tV2
is
n
o
t v
er
y
s
u
itab
le
f
o
r
a
p
p
licatio
n
s
th
at
d
em
an
d
h
ig
h
er
p
r
ec
is
io
n
.
5.
CO
NCLU
SI
O
N
I
n
th
is
r
esear
c
h
,
th
e
p
r
o
p
o
s
e
d
Mo
b
ileNetV2
tec
h
n
iq
u
e
is
im
p
lem
en
ted
to
en
h
a
n
ce
th
e
ef
f
ec
tiv
e
p
r
ed
ictio
n
o
f
co
c
o
n
u
t
an
d
co
co
n
u
t
leaf
d
is
ea
s
es.
T
h
e
co
co
n
u
t
an
d
co
co
n
u
t
leaf
im
ag
es
ar
e
co
llected
f
r
o
m
d
atasets
,
alo
n
g
with
ad
d
itio
n
al
f
ea
tu
r
es.
Featu
r
e
v
ec
to
r
s
a
r
e
ex
tr
ac
ted
f
r
o
m
th
e
p
r
e
-
p
r
o
ce
s
s
ed
d
ata
u
s
in
g
th
e
R
esNet5
0
tech
n
iq
u
e.
T
h
ese
ex
tr
ac
ted
o
p
tim
al
f
ea
tu
r
es
ar
e
t
h
en
f
ed
in
to
th
e
class
if
icatio
n
p
r
o
ce
s
s
,
wh
er
e
th
e
im
p
lem
en
ted
M
o
b
ileNetV2
m
eth
o
d
is
u
s
ed
f
o
r
class
if
icatio
n
.
W
h
en
c
o
m
p
ar
e
d
to
th
e
e
x
i
s
tin
g
m
eth
o
d
s
,
t
h
e
co
co
n
u
t
im
ag
e
a
n
d
C
o
co
n
u
t
leaf
im
ag
e
d
atasets
p
r
o
v
id
e
th
e
b
est
r
esu
lts
f
r
o
m
th
e
im
p
lem
en
ted
m
eth
o
d
.
T
h
e
im
p
lem
en
ted
m
et
h
o
d
ac
h
iev
e
s
ac
cu
r
ac
y
v
alu
es
o
f
9
9
.
9
9
%
an
d
9
9
.
2
%
r
esp
ec
tiv
ely
,
o
n
co
co
n
u
t
im
ag
e
a
n
d
co
co
n
u
t
leaf
im
ag
e
d
atasets
,
as
co
m
p
ar
ed
to
th
e
ex
is
tin
g
m
eth
o
d
s
f
o
r
th
e
p
r
e
d
ictio
n
o
f
c
o
co
n
u
t
an
d
co
co
n
u
t
leaf
d
is
ea
s
e.
T
h
ese
f
in
d
in
g
s
h
i
g
h
lig
h
t n
o
v
el
p
atter
n
s
in
th
e
a
p
p
licatio
n
o
f
DL
m
o
d
els f
o
r
d
i
s
ea
s
e
id
en
tific
atio
n
an
d
r
e
v
ea
l
r
esear
ch
g
a
p
s
th
a
t
r
eq
u
ir
e
f
u
r
th
er
i
n
v
esti
g
atio
n
.
T
h
is
d
ataset
is
wid
ely
u
s
e
d
b
y
th
e
s
cien
tific
co
m
m
u
n
ity
t
o
cr
ea
te
an
d
ass
ess
DL
m
o
d
els
f
o
r
th
e
id
en
ti
f
icatio
n
o
f
p
la
n
t
d
is
ea
s
es.
Fa
r
m
er
s
will
f
in
d
th
is
ap
p
r
o
ac
h
v
er
y
h
elp
f
u
l in
lear
n
in
g
ab
o
u
t
th
eir
c
r
o
p
y
ield
.
I
n
t
h
e
f
u
tu
r
e,
th
is
r
esear
ch
ca
n
b
e
f
u
r
th
er
ex
ten
d
ed
b
y
u
s
in
g
a
h
y
b
r
id
m
et
h
o
d
t
o
im
p
r
o
v
e
th
e
ac
c
u
r
ac
y
o
f
co
c
o
n
u
t
leaf
d
is
ea
s
e
d
etec
tio
n
,
b
y
m
ea
n
s
o
f
co
co
n
u
t
leaf
im
ag
e
d
ataset.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Kav
ith
a
Ma
g
ad
i
Go
p
alak
r
is
h
n
a
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
R
av
ip
r
ak
ash
Ma
d
en
u
r
L
in
g
ar
aju
✓
✓
✓
✓
✓
✓
✓
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
p
r
ed
ictio
n
o
f c
o
co
n
u
t a
n
d
c
o
co
n
u
t le
a
f d
is
ea
s
e
u
s
in
g
Mo
b
ileN
etV
2
…
(
K
a
vith
a
Ma
g
a
d
i
Go
p
a
la
kris
h
n
a
)
2843
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
th
e
f
in
d
i
n
g
s
o
f
th
is
s
tu
d
y
ar
e
o
p
en
ly
a
v
ailab
le
in
C
o
co
n
u
t
leaf
d
ata
r
e
p
o
s
ito
r
y
at
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
s
a
mith
a
9
6
/co
co
n
u
td
is
ea
s
e
s
.
RE
F
E
R
E
NC
E
S
[
1
]
S
.
K
.
M
a
n
o
h
a
r
a
n
,
R
.
K
.
M
e
g
a
l
i
n
g
a
m,
A
.
H
.
K
o
t
a
,
V
.
K
.
T.
P
.
,
a
n
d
K
.
S
.
S
a
n
k
a
r
d
a
s,
“
H
y
b
r
i
d
f
u
z
z
y
su
p
p
o
r
t
v
e
c
t
o
r
mac
h
i
n
e
a
p
p
r
o
a
c
h
f
o
r
C
o
c
o
n
u
t
t
r
e
e
c
l
a
ssi
f
i
c
a
t
i
o
n
u
s
i
n
g
i
mag
e
m
e
a
s
u
r
e
m
e
n
t
,
”
En
g
i
n
e
e
ri
n
g
Ap
p
l
i
c
a
t
i
o
n
s
o
f
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
1
2
6
,
N
o
v
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
n
g
a
p
p
a
i
.
2
0
2
3
.
1
0
6
8
0
6
.
[
2
]
E.
E
l
-
S
h
a
f
e
i
y
,
A
.
A
.
A
b
o
h
a
n
y
,
W
.
M
.
El
mess
e
r
y
,
a
n
d
A
.
A
.
A
.
El
-
M
a
g
e
e
d
,
“
Est
i
ma
t
i
o
n
o
f
c
o
c
o
n
u
t
mat
u
r
i
t
y
b
a
se
d
o
n
f
u
z
z
y
n
e
u
r
a
l
n
e
t
w
o
r
k
a
n
d
s
p
e
r
m
w
h
a
l
e
o
p
t
i
mi
z
a
t
i
o
n
,
”
N
e
u
r
a
l
C
o
m
p
u
t
i
n
g
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
3
5
,
n
o
.
2
6
,
p
p
.
1
9
5
4
1
–
1
9
5
6
4
,
S
e
p
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
5
2
1
-
0
2
3
-
0
8
7
6
1
-
0.
[
3
]
J.
A
.
F
e
r
r
e
i
r
a
,
A
.
C
.
F
a
ss
o
n
i
,
J.
M
.
S
.
F
e
r
r
e
i
r
a
,
P
.
M
.
P
.
L
i
n
s
,
a
n
d
C
.
B
.
G
.
B
o
t
t
o
l
i
,
“
C
y
p
r
o
c
o
n
a
z
o
l
e
t
r
a
n
s
l
o
c
a
t
i
o
n
i
n
c
o
c
o
n
u
t
p
a
l
m
t
r
e
e
u
si
n
g
v
e
g
e
t
a
t
i
v
e
e
n
d
o
t
h
e
r
a
p
y
:
E
v
a
l
u
a
t
i
o
n
b
y
LC
-
M
S
/
M
S
a
n
d
m
a
t
h
e
mat
i
c
a
l
m
o
d
e
l
i
n
g
,
”
H
o
rt
i
c
u
l
t
u
r
a
e
,
v
o
l
.
8
,
n
o
.
1
2
,
N
o
v
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
h
o
r
t
i
c
u
l
t
u
r
a
e
8
1
2
1
0
9
9
.
[
4
]
N
.
D
.
R
o
sl
a
n
,
S
.
S
u
n
d
r
a
m,
L.
W
.
H
o
n
g
,
K
.
L
.
L
i
n
g
,
a
n
d
G
.
V
a
d
a
m
a
l
a
i
,
“
A
n
a
l
y
si
s
o
f
C
o
c
o
n
u
t
c
a
d
a
n
g
-
c
a
d
a
n
g
v
i
r
o
i
d
v
a
r
i
a
n
t
s
o
n
f
i
e
l
d
samp
l
e
s
e
x
h
i
b
i
t
i
n
g
v
a
r
i
a
t
i
o
n
i
n
o
r
a
n
g
e
sp
o
t
t
i
n
g
s
y
mp
t
o
m e
x
p
r
e
ss
i
o
n
a
n
d
s
e
v
e
r
i
t
y
,
”
M
o
l
e
c
u
l
a
r
Bi
o
l
o
g
y
Re
p
o
rt
s
,
v
o
l
.
5
0
,
n
o
.
1
1
,
p
p
.
9
6
9
9
–
9
7
0
5
,
N
o
v
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
0
3
3
-
0
2
3
-
0
8
7
7
1
-
0.
[
5
]
M
.
M
a
r
a
y
e
t
a
l
.
,
“
A
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
-
e
n
a
b
l
e
d
c
o
c
o
n
u
t
t
r
e
e
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
mo
d
e
l
f
o
r
s
mart
a
g
r
i
c
u
l
t
u
r
e
,
”
C
o
m
p
u
t
e
rs
a
n
d
E
l
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
v
o
l
.
1
0
4
,
D
e
c
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
p
e
l
e
c
e
n
g
.
2
0
2
2
.
1
0
8
3
9
9
.
[
6
]
L.
Y
u
e
t
a
l
.
,
“
A
n
o
n
-
d
e
st
r
u
c
t
i
v
e
c
o
c
o
n
u
t
f
r
u
i
t
a
n
d
see
d
t
r
a
i
t
s
e
x
t
r
a
c
t
i
o
n
met
h
o
d
b
a
se
d
o
n
M
i
c
r
o
-
C
T
a
n
d
d
e
e
p
l
a
b
V
3
+
m
o
d
e
l
,
”
Fro
n
t
i
e
rs
i
n
P
l
a
n
t
S
c
i
e
n
c
e
,
v
o
l
.
1
3
,
D
e
c
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
8
9
/
f
p
l
s
.
2
0
2
2
.
1
0
6
9
8
4
9
.
[
7
]
D
.
G
o
me
z
e
t
a
l
.
,
“
A
d
v
a
n
c
i
n
g
c
o
mm
o
n
b
e
a
n
(
P
h
a
se
o
l
u
s
v
u
l
g
a
r
i
s
L.
)
d
i
s
e
a
s
e
d
e
t
e
c
t
i
o
n
w
i
t
h
Y
O
LO
d
r
i
v
e
n
d
e
e
p
l
e
a
r
n
i
n
g
t
o
e
n
h
a
n
c
e
a
g
r
i
c
u
l
t
u
r
a
l
A
I
,
”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
r
t
s
,
v
o
l
.
1
4
,
n
o
.
1
,
J
u
l
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
0
2
4
-
6
6
2
8
1
-
w.
[
8
]
S
.
V
i
d
h
a
n
a
a
r
a
c
h
c
h
i
,
J
.
L.
W
i
j
e
k
o
o
n
,
W
.
A
.
S
.
P
.
A
b
e
y
s
i
r
i
w
a
r
d
h
a
n
a
,
a
n
d
M
.
W
i
j
e
s
u
n
d
a
r
a
,
“
Ea
r
l
y
d
i
a
g
n
o
si
s
a
n
d
s
e
v
e
r
i
t
y
a
s
sessm
e
n
t
o
f
w
e
l
i
g
a
m
a
c
o
c
o
n
u
t
l
e
a
f
w
i
l
t
d
i
se
a
se
a
n
d
c
o
c
o
n
u
t
c
a
t
e
r
p
i
l
l
a
r
i
n
f
e
s
t
a
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
i
ma
g
e
p
r
o
c
e
s
si
n
g
t
e
c
h
n
i
q
u
e
s
,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
1
3
,
n
o
.
2
,
p
p
.
2
4
4
6
3
–
2
4
4
7
7
,
A
u
g
.
2
0
2
5
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
5
.
3
5
3
7
6
6
4
.
[
9
]
S
.
F
.
I
sl
a
m,
N
.
C
h
a
k
r
a
b
a
r
t
y
,
a
n
d
M
.
S
.
U
d
d
i
n
,
“
C
i
t
r
u
s
l
e
a
f
d
i
s
e
a
s
e
d
e
t
e
c
t
i
o
n
t
h
r
o
u
g
h
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
B
u
l
l
e
t
i
n
o
f
E
l
e
c
t
r
i
c
a
l
En
g
i
n
e
e
ri
n
g
a
n
d
I
n
f
o
rm
a
t
i
c
s (BE
EI
)
,
v
o
l
.
1
3
,
n
o
.
4
,
p
p
.
2
6
4
3
–
2
6
4
9
,
A
u
g
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
5
9
1
/
e
e
i
.
v
1
3
i
4
.
4
5
2
1
.
[
1
0
]
M
.
S
y
a
r
i
e
f
a
n
d
W
.
S
e
t
i
a
w
a
n
,
“
C
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
f
o
r
mai
z
e
l
e
a
f
d
i
s
e
a
s
e
i
m
a
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
T
ELKO
MN
I
K
A
(
T
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
C
o
m
p
u
t
i
n
g
El
e
c
t
r
o
n
i
c
s
a
n
d
C
o
n
t
r
o
l
)
,
v
o
l
.
1
8
,
n
o
.
3
,
p
p
.
1
3
7
6
–
1
3
8
1
,
Ju
n
.
2
0
2
0
,
d
o
i
:
1
0
.
1
2
9
2
8
/
t
e
l
k
o
m
n
i
k
a
.
v
1
8
i
3
.
1
4
8
4
0
.
[
1
1
]
X
.
H
u
a
n
g
,
M
.
M
.
A
l
o
b
a
e
d
y
,
Y
.
F
a
z
e
a
,
S
.
B
.
G
o
y
a
l
,
a
n
d
Z.
D
e
n
g
,
“
D
i
se
a
s
e
i
n
f
e
c
t
i
o
n
c
l
a
ssi
f
i
c
a
t
i
o
n
i
n
c
o
c
o
n
u
t
t
r
e
e
b
a
se
d
o
n
a
n
e
n
h
a
n
c
e
d
v
i
su
a
l
g
e
o
met
r
y
g
r
o
u
p
m
o
d
e
l
,
”
Pr
o
c
e
sses
,
v
o
l
.
1
3
,
n
o
.
3
,
p
.
6
8
9
,
F
e
b
.
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
p
r
1
3
0
3
0
6
8
9
.
[
1
2
]
Y
.
M
.
A
b
d
A
l
g
a
n
i
,
O
.
J.
M
a
r
q
u
e
z
C
a
r
o
,
L
.
M
.
R
o
b
l
a
d
i
l
l
o
B
r
a
v
o
,
C
.
K
a
u
r
,
M
.
S
.
A
l
A
n
s
a
r
i
,
a
n
d
B
.
K
i
r
a
n
B
a
l
a
,
“
L
e
a
f
d
i
s
e
a
s
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
a
n
d
c
l
a
ssi
f
i
c
a
t
i
o
n
u
si
n
g
o
p
t
i
m
i
z
e
d
d
e
e
p
l
e
a
r
n
i
n
g
,
”
Me
a
su
r
e
m
e
n
t
:
S
e
n
so
r
s
,
v
o
l
.
2
5
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
me
a
se
n
.
2
0
2
2
.
1
0
0
6
4
3
.
[
1
3
]
R
.
K
.
M
e
g
a
l
i
n
g
a
m,
S
.
K
.
M
a
n
o
h
a
r
a
n
,
a
n
d
R
.
B
.
M
a
r
u
t
h
a
b
a
b
u
,
“
I
n
t
e
g
r
a
t
e
d
f
u
z
z
y
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
f
o
r
i
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
c
o
c
o
n
u
t
ma
t
u
r
i
t
y
w
i
t
h
o
u
t
h
u
ma
n
i
n
t
e
r
v
e
n
t
i
o
n
,
”
N
e
u
r
a
l
C
o
m
p
u
t
i
n
g
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
3
6
,
n
o
.
1
1
,
p
p
.
6
1
3
3
–
6
1
4
5
,
A
p
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
5
2
1
-
023
-
0
9
4
0
2
-
2.
[
1
4
]
J.
W
.
R
a
n
d
l
e
s
,
C
.
A
.
C
u
e
t
o
,
G
.
V
a
d
a
mal
a
i
,
a
n
d
D
.
H
a
n
o
l
d
,
“
F
i
n
d
i
n
g
t
h
e
c
o
c
o
n
u
t
c
a
d
a
n
g
-
c
a
d
a
n
g
a
n
d
t
i
n
a
n
g
a
j
a
v
i
r
o
i
d
s,
n
a
t
u
r
a
l
l
y
o
c
c
u
r
r
i
n
g
p
a
t
h
o
g
e
n
s
o
f
t
r
o
p
i
c
a
l
m
o
n
o
c
o
t
y
l
e
d
o
n
s
o
f
O
c
e
a
n
i
a
,
”
i
n
F
u
n
d
a
m
e
n
t
a
l
s
o
f
V
i
r
o
i
d
B
i
o
l
o
g
y
,
E
l
se
v
i
e
r
,
2
0
2
4
,
p
p
.
1
5
1
–
1
6
8
.
d
o
i
:
1
0
.
1
0
1
6
/
B
9
7
8
-
0
-
323
-
9
9
6
8
8
-
4
.
0
0
0
0
5
-
5.
[
1
5
]
S
.
T
h
i
t
e
,
Y
.
S
u
r
y
a
w
a
n
s
h
i
,
K
.
P
a
t
i
l
,
a
n
d
P
.
C
h
u
mc
h
u
,
“
C
o
c
o
n
u
t
(
C
o
c
o
s
n
u
c
i
f
e
r
a
)
t
r
e
e
d
i
s
e
a
s
e
d
a
t
a
s
e
t
:
A
d
a
t
a
s
e
t
f
o
r
d
i
s
e
a
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
f
o
r
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
l
i
c
a
t
i
o
n
s,
”
D
a
t
a
i
n
Bri
e
f
,
v
o
l
.
5
1
,
D
e
c
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
d
i
b
.
2
0
2
3
.
1
0
9
6
9
0
.
[
1
6
]
S
.
V
i
g
n
e
sh
w
a
r
a
n
a
n
d
V
.
N
.
Ta
m
b
u
r
i
,
“
I
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
c
o
c
o
n
u
t
p
a
l
m
t
r
e
e
s
u
si
n
g
si
n
g
l
e
s
h
o
t
d
e
t
e
c
t
o
r
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
,
”
S
p
a
t
i
a
l
I
n
f
o
rm
a
t
i
o
n
Re
se
a
rc
h
,
v
o
l
.
3
1
,
n
o
.
6
,
p
p
.
6
9
5
–
7
0
7
,
D
e
c
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s4
1
3
2
4
-
023
-
0
0
5
4
2
-
0.
[
1
7
]
L.
G
.
D
i
v
y
a
n
t
h
,
P
.
S
o
n
i
,
C
.
M
.
P
a
r
e
e
k
,
R
.
M
a
c
h
a
v
a
r
a
m
,
M
.
N
a
d
i
m
i
,
a
n
d
J.
P
a
l
i
w
a
l
,
“
D
e
t
e
c
t
i
o
n
o
f
c
o
c
o
n
u
t
c
l
u
st
e
r
s
b
a
sed
o
n
o
c
c
l
u
s
i
o
n
c
o
n
d
i
t
i
o
n
u
s
i
n
g
a
t
t
e
n
t
i
o
n
-
g
u
i
d
e
d
f
a
st
e
r
R
-
C
N
N
f
o
r
r
o
b
o
t
i
c
h
a
r
v
e
st
i
n
g
,
”
Fo
o
d
s
,
v
o
l
.
1
1
,
n
o
.
2
3
,
D
e
c
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
f
o
o
d
s1
1
2
3
3
9
0
3
.
[
1
8
]
T.
K
a
l
a
i
s
e
l
v
i
a
n
d
V
.
N
a
r
mat
h
a
,
“
C
o
t
t
o
n
c
r
o
p
d
i
se
a
se
d
e
t
e
c
t
i
o
n
u
s
i
n
g
F
R
C
M
se
g
me
n
t
a
t
i
o
n
a
n
d
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
c
l
a
ss
i
f
i
e
r
,
”
i
n
C
o
m
p
u
t
a
t
i
o
n
a
l
V
i
si
o
n
a
n
d
Bi
o
-
I
n
s
p
i
re
d
C
o
m
p
u
t
i
n
g
,
2
0
2
3
,
p
p
.
5
5
7
–
5
7
7
.
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
19
-
9
8
1
9
-
5
_
4
0
.
[
1
9
]
R
.
K
.
La
k
sh
mi
a
n
d
N
.
S
a
v
a
r
i
mu
t
h
u
,
“
P
LD
D
—
a
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
p
l
a
n
t
l
e
a
f
d
i
sea
s
e
d
e
t
e
c
t
i
o
n
,
”
I
EEE
C
o
n
su
m
e
r
E
l
e
c
t
r
o
n
i
c
s
Ma
g
a
zi
n
e
,
v
o
l
.
1
1
,
n
o
.
3
,
p
p
.
4
4
–
4
9
,
M
a
y
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
M
C
E
.
2
0
2
1
.
3
0
8
3
9
7
6
.
[
2
0
]
V
.
K
a
v
i
t
h
a
ma
n
i
a
n
d
S
.
U
ma
M
a
h
e
sw
a
r
i
,
“
I
n
v
e
s
t
i
g
a
t
i
o
n
o
f
d
e
e
p
l
e
a
r
n
i
n
g
f
o
r
w
h
i
t
e
f
l
y
i
d
e
n
t
i
f
i
c
a
t
i
o
n
i
n
c
o
c
o
n
u
t
t
r
e
e
l
e
a
v
e
s,”
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
0
,
N
o
v
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
sw
a
.
2
0
2
3
.
2
0
0
2
9
0
.
[
2
1
]
R
.
K
.
M
e
g
a
l
i
n
g
a
m,
S
.
K
u
t
t
a
n
k
u
l
a
n
g
a
r
a
M
a
n
o
h
a
r
a
n
,
D
.
H
.
T.
A
.
B
a
b
u
,
G
.
S
r
i
r
a
m
,
K
.
L
o
k
e
sh
,
a
n
d
S
.
K
a
r
i
p
a
r
a
mb
i
l
S
u
d
h
e
e
sh
,
“
C
o
c
o
n
u
t
t
r
e
e
s
c
l
a
ss
i
f
i
c
a
t
i
o
n
b
a
s
e
d
o
n
h
e
i
g
h
t
,
i
n
c
l
i
n
a
t
i
o
n
,
a
n
d
o
r
i
e
n
t
a
t
i
o
n
u
si
n
g
M
I
N
-
S
V
M
a
l
g
o
r
i
t
h
m,
”
N
e
u
ra
l
C
o
m
p
u
t
i
n
g
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
3
5
,
n
o
.
1
6
,
p
p
.
1
2
0
5
5
–
1
2
0
7
1
,
J
u
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
5
2
1
-
023
-
0
8
3
3
9
-
w.
[
2
2
]
P
.
R
.
D
e
S
i
l
v
a
,
C
.
N
.
P
e
r
e
r
a
,
B
.
W
.
B
a
h
d
e
r
,
a
n
d
R
.
N
.
A
t
t
a
n
a
y
a
k
e
,
“
N
e
s
t
e
d
P
C
R
-
b
a
s
e
d
r
a
p
i
d
d
e
t
e
c
t
i
o
n
o
f
p
h
y
t
o
p
l
a
sm
a
l
e
a
f
w
i
l
t
Evaluation Warning : The document was created with Spire.PDF for Python.