I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
,
p
p
.
824
~
830
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
1
.
p
p
8
2
4
-
8
3
0
824
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Ada
ptive t
ra
nsfo
r
mer archit
ectu
re
for sca
la
ble earth
o
bserv
a
tion v
ia
hypers
pectral ima
g
ing
Dev
endra
K
um
a
r
Sa
ra
g
o
o
r
M
a
da
na
y
a
k
a
1
,
2
,
Dev
a
na
t
ha
n
M
uthuk
ris
hn
a
n
1
1
S
c
h
o
o
l
o
f
E
l
e
c
t
r
o
n
i
c
s
a
n
d
C
o
mm
u
n
i
c
a
t
i
o
n
E
n
g
i
n
e
e
r
i
n
g
,
R
EV
A
U
n
i
v
e
r
si
t
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
o
n
i
c
s a
n
d
C
o
m
mu
n
i
c
a
t
i
o
n
En
g
i
n
e
e
r
i
n
g
,
G
o
v
e
r
n
m
e
n
t
En
g
i
n
e
e
r
i
n
g
C
o
l
l
e
g
e
,
C
h
a
mara
j
a
n
a
g
a
r
,
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
Oct
9
,
2
0
2
5
R
ev
is
ed
Dec
2
6
,
2
0
2
5
Acc
ep
ted
J
an
1
0
,
2
0
2
6
Hy
p
e
rsp
e
c
tral
ima
g
e
(HSI)
c
las
sifica
ti
o
n
is
o
n
e
o
f
t
h
e
c
rit
ica
l
p
ro
c
e
ss
e
s
in
v
o
lv
e
d
in
re
m
o
te
se
n
sin
g
a
p
p
li
c
a
ti
o
n
t
h
a
t
p
lay
s
a
c
ru
c
ial
r
o
le
t
o
wa
rd
s
e
a
rth
o
b
se
rv
a
ti
o
n
.
Ow
i
n
g
to
c
o
m
p
lex
sp
a
ti
a
l
-
sp
e
c
tral
re
latio
n
sh
i
p
a
n
d
h
i
g
h
d
ime
n
sio
n
a
l
it
y
,
it
is
q
u
it
e
a
c
h
a
ll
e
n
g
in
g
tas
k
t
o
su
b
jec
t
HSI
c
o
n
ten
t
t
o
c
o
n
v
e
n
ti
o
n
a
l
d
a
ta
a
n
a
ly
t
ics
o
r
e
x
isti
n
g
m
e
th
o
d
s.
He
n
c
e
,
t
h
e
p
ro
p
o
se
d
stu
d
y
in
tro
d
u
c
e
s
a
n
o
v
e
l
c
o
m
p
u
tati
o
n
a
l
m
o
d
e
l
k
n
o
wn
a
s
a
d
a
p
ti
v
e
sp
e
c
tra
-
sp
a
ti
a
l
tran
sfo
rm
e
r
(ASS
T)
t
o
a
d
d
re
ss
th
e
se
o
n
g
o
in
g
c
h
a
ll
e
n
g
e
s
a
n
d
sh
o
rt
c
o
m
in
g
o
f
e
x
isti
n
g
a
rti
f
icia
l
in
telli
g
e
n
c
e
(AI)
b
a
se
d
m
o
d
e
ll
in
g
.
T
h
e
p
ro
p
o
s
e
d
m
o
d
e
l
c
o
n
tri
b
u
tes
to
wa
rd
s
a
n
o
v
e
l
tran
s
fo
rm
e
r
-
b
a
se
d
a
rc
h
it
e
c
tu
re
wh
e
re
a
d
isti
n
c
t
sp
e
c
tral
-
sp
a
ti
a
l
a
tt
e
n
ti
o
n
m
e
th
o
d
h
a
s
b
e
e
n
u
se
d
with
tra
n
sfo
rm
e
r
e
n
c
o
d
e
r
.
Th
is
n
o
v
e
l
c
o
m
b
in
a
ti
o
n
fa
c
il
i
tate
s
h
ig
h
ly
a
d
a
p
ti
v
e
a
n
d
c
o
n
tex
tu
a
ll
y
e
n
rich
e
d
fe
a
tu
re
e
x
trac
ti
o
n
.
Tes
ted
o
n
u
n
iv
e
rsa
ll
y
sta
n
d
a
rd
HSI
d
a
tas
e
t
o
f
P
a
v
ia
Un
iv
e
rsity
,
t
h
e
p
ro
p
o
se
d
AS
S
T
m
o
d
e
l
h
a
s
b
e
e
n
b
e
n
c
h
m
a
rk
e
d
w
it
h
n
o
ti
c
e
9
7
.
2
6
%
o
f
o
v
e
ra
ll
a
c
c
u
ra
c
y
(OA
)
a
n
d
fa
ste
r
p
ro
c
e
ss
in
g
d
u
ra
ti
o
n
c
o
m
p
u
te
d
v
ia
train
i
n
g
ti
m
e
(TT
)
a
n
d
re
sp
o
n
se
ti
m
e
(RT)
i
n
c
o
n
tras
t
to
f
re
q
u
e
n
tl
y
a
d
o
p
te
d
m
a
c
h
in
e
lea
rn
in
g
(M
L
)
a
n
d
d
e
e
p
lea
rn
in
g
(DL)
m
o
d
e
ls.
T
h
e
a
c
c
o
m
p
li
sh
e
d
st
u
d
y
o
u
tc
o
m
e
s
tru
ly
e
x
h
ib
i
ted
h
i
g
h
l
y
im
p
ro
v
e
d
fe
a
tu
r
e
re
p
re
se
n
tatio
n
a
s
we
ll
a
s
ro
b
u
st
p
e
rfo
rm
a
n
c
e
a
g
a
in
st
c
las
s
i
m
b
a
lan
c
e
p
ro
b
lem
s
to
wa
rd
s
sc
a
lab
le
d
a
t
a
a
n
a
ly
sis
o
f
HSI
c
o
n
ten
ts
f
o
r
e
a
rth
o
b
se
rv
a
ti
o
n
.
K
ey
w
o
r
d
s
:
Ar
tific
ial
in
tellig
en
ce
C
las
s
if
icatio
n
Dee
p
lear
n
in
g
E
ar
th
o
b
s
er
v
atio
n
Hy
p
er
s
p
ec
tr
al
im
ag
e
Ma
ch
in
e
lear
n
in
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Dev
en
d
r
a
Ku
m
ar
Sar
ag
o
o
r
M
ad
an
ay
ak
a
Sch
o
o
l o
f
E
lectr
o
n
ics an
d
C
o
m
m
u
n
icatio
n
E
n
g
in
ee
r
in
g
,
R
E
VA
Un
iv
er
s
ity
R
u
k
m
in
i K
n
o
wled
g
e
Par
k
,
Ye
lah
an
k
a,
Kattig
en
a
h
alli,
B
en
g
alu
r
u
,
Kar
n
atak
a
5
6
0
0
6
4
,
I
n
d
i
a
E
m
ail:
s
m
d
ev
en
d
r
a
k
u
m
ar
1
5
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
A
h
y
p
er
s
p
ec
tr
al
im
a
g
e
(
HSI
)
b
elo
n
g
s
to
a
s
p
ec
if
ic
ty
p
e
o
f
i
m
ag
e
d
ata
th
at
c
o
n
s
is
ts
o
f
lar
g
e
r
an
g
e
o
f
elec
tr
o
m
ag
n
etic
s
p
ec
tr
u
m
b
a
n
d
s
f
ac
ilit
atin
g
d
etailed
an
al
y
s
is
o
f
s
u
r
f
ac
es
d
e
p
en
d
i
n
g
u
p
o
n
th
eir
s
p
ec
tr
al
s
ig
n
atu
r
es
[
1
]
.
HSI
co
n
ten
t
s
f
ac
ilit
ates
lan
d
co
v
er
m
o
n
ito
r
in
g
a
n
d
p
r
ec
is
e
id
en
tific
atio
n
o
f
m
i
n
er
al
co
m
p
o
s
itio
n
an
d
v
eg
etatio
n
h
ea
lth
[
2
]
.
T
h
e
e
n
r
ich
e
d
q
u
alit
y
o
f
HSI
is
b
en
ef
icial
as
well
as
it
is
al
s
o
a
m
ajo
r
im
p
ed
im
en
t
to
wa
r
d
s
th
e
task
o
f
class
if
icatio
n
[
3
]
.
Du
e
to
i
n
cr
ea
s
ed
n
u
m
b
er
o
f
s
p
ec
tr
al
b
an
d
s
,
HSI
s
u
f
f
er
s
f
r
o
m
cu
r
s
e
o
f
d
im
en
s
io
n
ality
th
at
ev
en
t
u
ally
lead
s
to
eith
e
r
d
eg
r
ad
ed
m
o
d
el
p
er
f
o
r
m
an
ce
o
r
o
v
er
f
itti
n
g
.
T
h
e
class
if
icatio
n
ac
cu
r
ac
y
is
also
co
m
p
licated
b
y
s
p
atial
h
et
er
o
g
en
eity
,
atm
o
s
p
h
er
ic
i
n
ter
f
er
en
ce
,
n
o
is
e,
an
d
s
p
ec
tr
al
v
ar
iab
ilit
y
.
At
p
r
esen
t,
th
er
e
wer
e
v
ar
io
u
s
r
esea
r
ch
attem
p
ts
b
ein
g
ca
r
r
ie
d
o
u
t
to
s
o
lv
e
th
ese
ch
allen
g
es
[
4
]
,
[
5
]
wh
ile
a
r
tific
ial
in
tellig
en
ce
(
A
I
)
,
in
t
h
e
f
o
r
m
o
f
m
ac
h
in
e
lear
n
i
n
g
(
ML
)
an
d
d
ee
p
lea
r
n
in
g
(
DL
)
,
h
as
s
h
o
wn
p
r
o
m
is
in
g
s
o
lu
tio
n
to
war
d
s
ad
d
r
ess
in
g
s
u
ch
f
o
r
m
o
f
class
if
icatio
n
p
r
o
b
lem
.
B
o
th
ML
an
d
DL
ap
p
r
o
ac
h
es
ar
e
q
u
ite
ca
p
ab
le
o
f
lear
n
i
n
g
c
o
m
p
lex
p
atter
n
s
ass
o
ciate
d
with
HSI
co
n
ten
ts
.
T
h
e
r
e
ar
e
v
ar
io
u
s
f
r
eq
u
e
n
tly
a
d
o
p
te
d
m
eth
o
d
s
in
ML
e.
g
.
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
e
(
SVM)
,
r
a
n
d
o
m
f
o
r
est
(
R
F)
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
d
a
p
tive
tr
a
n
s
fo
r
mer a
r
ch
itectu
r
e
fo
r
s
ca
la
b
le
ea
r
th
…
(
D
ev
en
d
r
a
K
u
ma
r
S
a
r
a
g
o
o
r
Ma
d
a
n
a
y
a
ka
)
825
co
n
v
o
l
u
tio
n
al
n
e
u
r
al
n
etwo
r
k
(
C
NN)
,
an
d
m
an
y
m
o
r
e
ap
p
r
o
ac
h
es
f
o
r
th
is
p
u
r
p
o
s
e.
Ho
wev
er
,
ML
m
o
d
els
h
av
e
d
ep
en
d
en
cies
o
f
m
an
u
al
f
ea
tu
r
e
ex
tr
ac
tio
n
wh
ile
DL
m
o
d
el
r
eq
u
ir
es
lar
g
e
s
ca
le
an
n
o
tated
d
ataset
with
p
o
ten
tial
av
ailab
ilit
y
o
f
co
m
p
u
tatio
n
al
p
o
wer
an
d
r
eso
u
r
ce
s
.
Hen
ce
,
ir
r
esp
ec
tiv
e
o
f
p
o
ten
t
ial
ad
v
an
tag
es,
ML
an
d
DL
m
o
d
els h
av
e
c
h
ar
ac
ter
is
tic
ch
allen
g
es to
war
d
s
ea
r
th
o
b
s
er
v
atio
n
.
F
r
o
m
t
h
e
c
o
n
t
e
x
t
o
f
r
e
l
a
t
e
d
w
o
r
k
t
o
w
a
r
d
s
M
L
a
p
p
r
o
a
c
h
e
s
,
t
h
e
r
e
a
r
e
d
i
f
f
e
r
e
n
t
v
a
r
i
a
n
t
s
o
f
l
e
a
r
n
i
n
g
m
o
d
e
l
s
i
m
p
l
e
m
e
n
t
e
d
t
i
l
l
d
a
t
e
.
S
V
M
i
s
o
n
e
s
u
c
h
s
u
p
e
r
v
i
s
e
d
M
L
m
o
d
e
l
w
h
i
c
h
i
s
q
u
i
t
e
e
f
f
e
c
t
i
v
e
f
o
r
H
S
I
c
l
a
s
s
i
f
i
c
a
t
i
o
n
a
s
i
t
i
s
q
u
i
t
e
r
o
b
u
s
t
t
o
d
i
m
e
n
s
i
o
n
a
l
i
t
y
p
r
o
b
l
e
m
s
,
c
o
u
l
d
b
e
u
s
e
d
f
o
r
r
e
a
l
-
t
i
m
e
a
n
a
l
y
s
i
s
,
l
i
g
h
t
w
e
i
g
h
t
f
o
r
i
n
f
e
r
e
n
c
e
[
6
]
,
[
7
]
.
H
o
w
e
v
e
r
,
t
h
e
y
a
r
e
v
e
r
y
s
h
a
l
l
o
w
m
o
d
e
l
w
h
i
l
e
n
o
s
p
a
t
i
a
l
c
o
n
t
e
x
t
i
s
m
o
d
e
l
w
h
i
c
h
i
s
e
s
s
e
n
t
i
a
l
i
n
e
a
r
t
h
’
s
o
b
s
e
r
v
a
t
i
o
n
.
RF
h
a
s
a
l
s
o
b
e
e
n
e
x
p
e
r
i
m
e
n
t
e
d
t
o
w
a
r
d
s
H
S
I
c
l
a
s
s
i
f
i
c
a
t
i
o
n
a
s
i
t
c
a
n
h
a
n
d
l
e
t
h
e
h
i
g
h
-
d
i
m
e
n
s
i
o
n
a
l
d
a
t
a
q
u
i
t
e
w
e
l
l
e
v
e
n
w
i
t
h
s
m
a
l
l
l
a
b
e
l
e
d
d
a
t
a
s
e
t
o
f
f
e
r
i
n
g
s
c
a
l
a
b
l
e
o
u
t
c
o
m
e
s
[
8
]
–
[
1
0
]
.
H
o
w
e
v
e
r
,
i
t
h
a
s
s
i
m
i
l
a
r
p
r
o
b
l
e
m
l
i
k
e
S
V
M
i
.
e
.
,
n
o
n
-
i
n
c
l
u
s
i
o
n
o
f
s
p
a
t
i
a
l
c
o
n
t
e
x
t
c
o
n
s
i
d
e
r
a
t
i
o
n
w
h
i
l
e
t
h
e
y
a
r
e
l
e
s
s
e
f
f
e
c
t
i
v
e
f
o
r
o
v
e
r
l
a
p
p
i
n
g
c
l
a
s
s
e
s
o
f
s
p
e
c
t
r
a
l
d
a
t
a
.
F
r
o
m
t
h
e
p
e
r
s
p
e
c
t
i
v
e
o
f
t
h
e
D
L
m
e
t
h
o
d
,
i
t
h
a
s
b
e
e
n
n
o
t
e
d
t
h
a
t
C
N
N
h
a
s
b
e
e
n
q
u
i
t
e
d
o
m
i
n
a
n
t
l
y
u
s
e
d
a
d
o
p
t
i
n
g
t
w
o
d
i
f
f
e
r
e
n
t
a
p
p
r
o
a
c
h
e
s
v
i
z
.
s
t
a
n
d
a
l
o
n
e
u
s
a
g
e
o
f
C
N
N
[
1
1
]
–
[
1
3
]
,
h
y
b
r
i
d
u
s
a
g
e
o
f
C
N
N
[
1
4
]
–
[
1
6
]
,
a
n
d
m
i
x
e
d
d
i
m
e
n
s
i
o
n
a
l
-
i
n
c
l
u
s
i
o
n
i
n
C
N
N
[
1
7
]
–
[
1
9
]
.
T
h
e
m
i
x
e
d
d
i
m
e
n
s
i
o
n
a
l
a
p
p
r
o
a
c
h
e
s
a
r
e
o
f
t
w
o
t
y
p
e
s
f
u
r
t
h
e
r
v
i
z
.
i
)
s
t
a
n
d
a
r
d
3
D
-
C
N
N
[
2
0
]
,
[
2
1
]
a
n
d
i
i
)
m
u
l
t
i
-
s
c
a
l
e
3
D
-
C
N
N
a
l
s
o
k
n
o
w
n
a
s
M
3
D
-
C
N
N
[
2
2
]
,
[
2
3
]
.
S
t
u
d
i
e
s
h
a
v
e
b
e
e
n
a
l
s
o
c
a
r
r
i
e
d
o
u
t
u
s
i
n
g
e
x
t
r
e
m
e
g
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
(
X
G
B
o
o
s
t
)
t
o
f
i
n
d
t
h
a
t
i
t
s
p
r
e
d
i
c
t
i
v
e
p
e
r
f
o
r
m
a
n
c
e
i
s
q
u
i
t
e
h
i
g
h
a
n
d
c
a
n
p
e
r
f
o
r
m
b
e
t
t
e
r
t
h
a
n
RF
a
n
d
S
V
M
t
o
o
t
o
w
a
r
d
s
H
S
I
c
l
a
s
s
i
f
i
c
a
t
i
o
n
[
2
4
]
–
[
2
6
]
.
T
h
e
id
en
tifie
d
r
esear
ch
p
r
o
b
le
m
s
f
r
o
m
ex
is
tin
g
ap
p
r
o
ac
h
es
ar
e
as
f
o
llo
ws:
i)
o
win
g
to
in
clu
s
io
n
o
f
d
ee
p
cu
b
e
p
r
o
ce
s
s
in
g
,
3
D
-
C
NN
is
co
m
p
u
tatio
n
ally
ex
p
e
n
s
iv
e,
ii)
th
e
p
r
im
e
s
h
o
r
tco
m
in
g
o
f
3
D
-
C
NN
is
ass
o
ciate
d
with
h
ig
h
er
i
n
clin
a
tio
n
to
war
d
s
o
v
er
f
itti
n
g
with
l
ess
f
lex
ib
ilit
y
in
h
an
d
lin
g
s
p
e
ctr
al
d
ep
en
d
en
cies
o
f
lo
n
g
er
r
an
g
es,
iii)
a
d
o
p
tio
n
o
f
M3
D
-
C
NN
co
u
ld
r
esu
lt
in
m
ax
im
ized
ar
c
h
itectu
r
al
co
m
p
lex
ity
wh
ich
co
u
ld
b
e
m
ain
ly
d
u
e
to
i
n
clu
s
io
n
o
f
m
u
lti
-
k
er
n
el
an
d
m
u
lti
-
b
r
a
n
c
h
m
o
d
u
les
f
o
r
h
an
d
li
n
g
lar
g
e
s
ca
le
o
f
s
p
atial
an
d
s
p
ec
tr
al
d
ata
in
h
y
p
er
s
p
ec
tr
al
co
n
ten
t,
an
d
iv
)
XGBo
o
s
t
h
as
a
s
h
o
r
tco
m
in
g
o
f
h
y
p
er
p
a
r
am
eter
s
en
s
itiv
ity
wh
ile
its
tu
n
in
g
is
q
u
ite
co
m
p
u
tatio
n
ally
e
x
p
en
s
iv
e
an
d
o
f
te
n
tim
e
c
o
n
s
u
m
in
g
.
T
h
e
i
d
en
tif
ied
lim
itatio
n
o
f
all
th
e
ab
o
v
e
ap
p
r
o
ac
h
es
is
ad
d
r
e
s
s
ed
b
y
p
r
o
p
o
s
ed
s
y
s
tem
b
y
in
co
r
p
o
r
atin
g
m
u
lti
-
h
ea
d
s
elf
-
att
en
tio
n
th
at
m
o
d
els
r
elatio
n
s
h
ip
o
f
g
lo
b
al
c
o
n
tex
t
o
v
er
s
p
ec
tr
al
a
n
d
s
p
atial
d
i
m
en
s
io
n
.
Ap
ar
t
f
r
o
m
th
is
,
p
r
o
p
o
s
ed
s
o
lu
tio
n
also
o
f
f
er
s
ad
a
p
tiv
e
f
ea
tu
r
e
ex
tr
a
ctio
n
,
im
p
r
o
v
e
d
g
e
n
er
aliza
tio
n
an
d
b
etter
in
ter
p
r
eta
b
ilit
y
,
to
w
ar
d
s
ad
d
r
ess
in
g
t
h
e
id
en
tifie
d
r
esear
ch
p
r
o
b
lem
s
.
T
h
e
n
ex
t
s
ec
tio
n
d
is
cu
s
s
es
a
b
o
u
t
th
e
a
d
o
p
ted
r
esear
ch
m
e
th
o
d
o
lo
g
y
to
war
d
s
class
if
y
in
g
HSI
ex
h
ib
itin
g
a
d
d
r
ess
in
g
th
e
s
h
o
r
tco
m
in
g
s
ass
o
ciate
d
with
ex
is
tin
g
m
o
d
els.
T
h
e
aim
o
f
th
e
p
r
o
p
o
s
ed
s
tu
d
y
is
to
in
tr
o
d
u
ce
a
n
o
v
el
ad
a
p
t
iv
e
lear
n
in
g
s
tr
ateg
y
wh
ich
b
a
lan
ce
s
th
e
h
ig
h
er
ac
c
u
r
ac
y
alo
n
g
with
c
o
s
t
-
ef
f
ec
tiv
e
co
m
p
u
tatio
n
al
p
er
f
o
r
m
a
n
ce
wh
ile
class
if
y
in
g
HSI
co
n
ten
ts
.
T
h
e
ac
co
m
p
lis
h
m
en
t
o
f
th
is
s
tu
d
y
is
ca
r
r
ied
o
u
t
b
y
h
ar
n
ess
in
g
th
e
p
o
ten
tial
o
f
DL
m
o
d
els.
T
h
e
co
n
tr
ib
u
tio
n
o
f
th
e
p
ap
er
is
as f
o
llo
ws:
i)
T
h
e
p
r
o
p
o
s
ed
s
tu
d
y
co
m
b
in
es
b
o
th
s
p
ec
tr
al
a
n
d
s
p
atial
atten
tio
n
m
o
d
u
les
f
o
r
f
ac
ilit
atin
g
ad
ap
tiv
e
f
o
c
u
s
o
n
cr
itical
s
p
atial
r
eg
io
n
s
an
d
s
p
ec
tr
al
b
an
d
s
to
in
cr
ea
s
e
d
is
cr
im
in
ativ
e
ca
p
ab
ilit
y
to
war
d
s
lear
n
ed
f
ea
tu
r
e.
ii)
T
h
e
s
tu
d
y
m
o
d
el
in
tr
o
d
u
ce
s
a
tr
an
s
f
o
r
m
er
-
b
ased
g
lo
b
al
co
n
tex
t
m
o
d
ellin
g
wh
ich
u
s
es
m
u
lti
-
h
ea
d
s
elf
-
atten
tio
n
f
o
r
m
o
d
ellin
g
d
ep
en
d
e
n
cies
o
f
lo
n
g
e
r
r
a
n
g
es
o
v
er
s
p
ec
tr
al
-
s
p
atial
p
atch
es
u
n
lik
e
co
n
v
en
tio
n
al
C
NN
m
o
d
el
u
s
in
g
o
n
ly
l
o
ca
l p
atter
n
s
.
iii)
T
h
e
p
r
esen
ted
m
o
d
el
co
m
p
le
tely
elim
in
ates
an
y
r
ea
s
o
n
to
s
elec
t
th
e
b
an
d
m
an
u
ally
o
r
an
y
f
o
r
m
o
f
d
ep
en
d
e
n
cies
to
war
d
s
h
an
d
cr
af
ted
f
ea
tu
r
es
as
it
ca
n
p
er
f
o
r
m
en
d
-
to
-
en
d
f
ea
tu
r
e
lear
n
in
g
d
ir
ec
tly
f
r
o
m
r
aw
p
atch
es o
f
HSI
co
n
ten
ts
.
iv
)
T
h
e
p
r
o
p
o
s
ed
d
esig
n
in
teg
r
at
es
tr
an
s
f
o
r
m
er
b
lo
c
k
with
s
h
a
llo
w
co
n
v
e
n
tio
n
al
lay
e
r
s
f
o
r
ac
co
m
p
lis
h
in
g
an
o
p
tim
al
b
ala
n
ce
b
etwe
en
m
o
d
el
p
er
f
o
r
m
an
ce
an
d
c
o
m
p
u
tatio
n
al
e
f
f
icien
cy
th
er
e
b
y
in
cr
ea
s
in
g
its
s
co
p
e
to
war
d
s
p
r
ac
tical
ea
r
th
o
b
s
er
v
atio
n
s
ce
n
a
r
io
s
.
2.
M
E
T
H
O
D
T
h
e
p
r
o
p
o
s
ed
s
tu
d
y
aim
s
to
w
ar
d
s
d
esig
n
in
g
a
c
o
s
t
-
ef
f
ec
tiv
e
as
well
as
ad
ap
tiv
e
lear
n
in
g
s
tr
ateg
ies
ess
en
tially
m
ea
n
t
f
o
r
an
aly
z
in
g
em
er
g
in
g
d
em
a
n
d
s
to
w
ar
d
s
ea
r
th
o
b
s
er
v
atio
n
.
Fo
r
th
is
p
u
r
p
o
s
e,
th
e
im
p
lem
en
tatio
n
s
ch
e
m
e
u
s
es
HSI
wh
ich
is
f
r
e
q
u
en
tl
y
u
s
ed
f
o
r
i
n
v
esti
g
atin
g
th
e
to
p
ic.
H
SI
im
ag
e
th
at
f
o
r
m
s
th
e
b
asis
o
f
an
ea
r
th
o
b
s
er
v
atio
n
d
ata
is
q
u
ite
s
o
p
h
is
ticated
an
d
attr
ib
u
ted
b
y
en
r
ich
e
d
s
p
atial
an
d
s
p
ec
tr
a
l
in
f
o
r
m
atio
n
wh
ich
ca
n
f
lu
ctu
ate
o
v
er
o
n
e
g
eo
g
r
ap
h
ic
r
eg
i
o
n
to
an
o
th
e
r
alo
n
g
with
tem
p
o
r
al
s
ca
les.
T
h
e
d
esig
n
o
f
p
r
o
p
o
s
ed
ad
a
p
tiv
e
s
p
ec
tr
a
-
s
p
atial
tr
an
s
f
o
r
m
er
(
ASST)
f
r
am
ewo
r
k
ad
ap
ts
to
s
u
ch
f
o
r
m
o
f
f
lu
ctu
atio
n
b
y
u
s
in
g
b
o
th
s
p
atial
atten
tio
n
an
d
s
p
ec
tr
al
a
tten
tio
n
.
Fig
u
r
e
1
o
f
f
er
s
a
f
o
r
m
aliza
tio
n
o
f
t
h
e
ad
o
p
ted
d
esig
n
o
f
a
r
ch
itectu
r
e
with
v
ar
io
u
s
o
p
e
r
atio
n
al
co
m
p
o
n
e
n
ts
f
o
r
p
u
r
p
o
s
e
o
f
class
if
y
in
g
HSI
co
n
ten
t
to
war
d
s
ea
r
th
o
b
s
er
v
atio
n
.
Fo
l
lo
win
g
ar
e
d
is
cu
s
s
io
n
o
f
th
e
c
o
r
e
m
o
d
u
les in
v
o
l
v
ed
in
a
r
ch
it
ec
tu
r
e
d
esig
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
824
-
8
3
0
826
Fig
u
r
e
1
.
Pro
p
o
s
ed
ar
c
h
itectu
r
e
o
f
ASST
T
h
e
p
r
im
e
aim
o
f
th
is
m
o
d
u
l
e
is
to
d
esig
n
a
s
o
p
h
is
ticated
r
elatio
n
s
h
ip
with
in
th
e
HSI
c
o
n
ten
ts
b
y
in
teg
r
atin
g
tr
a
n
s
f
o
r
m
er
-
b
ased
g
lo
b
al
r
ea
s
o
n
i
n
g
,
s
p
atial
atten
tio
n
,
an
d
s
p
ec
tr
al
atten
tio
n
f
o
r
y
ield
in
g
a
u
n
if
ie
d
ar
ch
itectu
r
e.
T
h
is
m
o
d
u
le
in
itially
ap
p
lies
s
p
ec
tr
al
atten
tio
n
u
s
in
g
1
×
1
co
n
v
o
lu
tio
n
th
at
h
ig
h
lig
h
ts
ess
en
tial
s
p
ec
tr
al
b
an
d
s
to
war
d
s
ea
ch
p
ix
el
.
Fu
r
th
er
,
s
p
atial
atten
tio
n
is
ap
p
lied
u
s
in
g
3
×
3
co
n
v
o
lu
tio
n
f
o
r
ac
q
u
ir
in
g
b
o
u
n
d
ar
y
in
f
o
r
m
atio
n
a
n
d
s
p
atial
tex
tu
r
e.
T
h
e
jo
in
t
f
ea
tu
r
e
m
ap
is
th
en
s
u
b
jecte
d
to
r
e
s
h
ap
in
g
wh
ile
it
is
p
ass
ed
th
r
o
u
g
h
a
tr
a
n
s
f
o
r
m
er
en
co
d
er
with
a
d
e
p
lo
y
m
e
n
t
o
f
s
elf
-
atten
tio
n
with
m
u
lti
-
h
ea
d
s
f
o
r
lea
r
n
in
g
d
ep
en
d
e
n
cies
o
v
e
r
t
h
e
c
o
m
p
l
ete
s
et
o
f
p
atch
es.
T
h
e
o
u
tco
m
e
o
f
th
e
m
o
d
u
le
o
p
e
r
atio
n
is
th
en
p
o
o
led
f
o
r
d
esig
n
in
g
a
f
ea
t
u
r
e
v
ec
to
r
o
f
c
o
n
s
tan
t
s
ize
th
at
is
p
ass
ed
th
r
o
u
g
h
f
u
lly
co
n
n
ec
ted
lay
er
s
.
At
th
e
en
d
,
So
f
tMa
x
ac
tiv
atio
n
is
u
s
ed
f
o
r
g
en
er
ati
n
g
p
r
o
b
ab
ilit
ies o
f
class
es.
T
h
e
im
p
lem
en
tatio
n
s
tep
s
ar
e
as
f
o
llo
ws:
i)
I
n
p
u
t:
th
e
r
ep
r
esen
tatio
n
o
f
ea
ch
in
p
u
t
p
atch
is
g
iv
en
as
∈
×
×
.
ii)
Sp
ec
tr
al
atten
tio
n
b
l
o
ck
:
th
e
s
tu
d
y
a
p
p
lies
a
1
×
1
co
n
v
o
l
u
tio
n
in
o
r
d
er
to
m
o
d
el
s
p
e
ctr
al
atten
tio
n
:
1
=
(
(
1
×
1
(
)
)
)
,
1
∈
×
×
wh
er
e
t
h
e
v
ar
iab
le
σ
r
e
p
r
esen
ts
r
ec
tifie
d
lin
ea
r
u
n
it
(
R
eL
U
)
ac
tiv
atio
n
,
wh
ile
BN
d
ep
icts
b
atch
n
o
r
m
aliza
tio
n
,
a
n
d
d
1
r
ep
r
esen
ts
th
e
q
u
an
tity
o
f
th
e
o
u
t
p
u
t
ch
an
n
els
(
e.
g
.
,
6
4
)
.
iii)
Sp
atial
atten
tio
n
b
lo
ck
: th
e
s
y
s
tem
ap
p
lies
a
3
×
3
c
o
n
v
o
lu
tio
n
:
2
=
(
(
3
×
3
(
1
)
)
)
,
2
∈
×
×
2
.
iv
)
T
r
an
s
f
o
r
m
e
r
en
co
d
er
b
lo
ck
:
t
h
is
m
o
d
u
le
is
u
s
ed
f
o
r
f
latten
in
g
th
e
s
p
atial
d
im
e
n
s
io
n
s
to
w
ar
d
s
th
e
m
u
lti
-
h
ea
d
atten
tio
n
=
ℎ
(
2
)
∈
2
×
2
wh
ile
th
is
f
o
llo
wed
b
y
a
p
p
ly
in
g
m
u
lti
-
h
ea
d
s
elf
-
atten
tio
n
=
(
,
)
∈
2
×
2
.
Fu
r
th
er
,
n
o
r
m
aliza
tio
n
an
d
r
esid
u
al
co
n
n
ec
tio
n
ar
e
ap
p
lied
,
=
(
+
)
wh
ile
g
lo
b
al
av
er
a
g
e
p
o
o
lin
g
is
ap
p
lied
to
g
en
e
r
ate
th
e
em
b
ed
d
in
g
v
ec
to
r
=
(
)
∈
2
.
v)
C
las
s
if
icatio
n
lay
er
:
th
e
em
b
ed
d
in
g
is
m
ad
e
to
p
a
s
s
th
r
o
u
g
h
f
u
lly
c
o
n
n
ec
te
d
lay
er
v
iz.
ℎ
1
=
(
.
+
1
)
,
ℎ
1
∈
2
5
6
,
ℎ
1
́
=
(
ℎ
1
,
=
0
.
5
)
,
an
d
̂
=
(
2
.
ℎ
1
′
+
2
)
,
̂
∈
,
wh
er
e
th
e
v
ar
ia
b
le
̂
r
ep
r
esen
ts
p
r
o
b
ab
ilit
ies
o
f
p
r
e
d
icted
cla
s
s
es,
wh
ile
weig
h
t
m
atr
ices
is
r
ep
r
esen
te
d
as
W
1
an
d
W
2
,
an
d
(
b
1
b
2
)
r
e
p
r
esen
ts
b
ias v
ec
to
r
s
.
T
h
e
co
r
e
aim
o
f
th
e
n
ex
t
m
o
d
u
le
o
f
tr
ain
i
n
g
p
r
o
ce
s
s
an
d
lo
s
s
f
u
n
ctio
n
is
to
o
p
tim
ize
th
e
ASST
m
o
d
u
le
b
y
r
ed
u
cin
g
th
e
f
r
e
q
u
en
cies
o
f
p
r
ed
ictio
n
er
r
o
r
s
v
i
a
s
u
p
er
v
is
ed
tr
ain
in
g
ad
o
p
tin
g
lab
eled
p
atch
in
g
.
T
h
e
tr
ain
in
g
is
ca
r
r
ie
d
o
u
t
f
o
r
th
e
m
o
d
el
u
s
in
g
class
if
icatio
n
lo
s
s
f
u
n
ctio
n
t
h
at
co
m
p
ar
es
tr
u
e
o
n
e
h
o
t
en
c
o
d
e
d
lab
el
with
th
e
p
r
ed
icted
class
p
r
o
b
ab
ilit
ies.
T
h
e
tr
ain
in
g
o
f
th
e
m
o
d
el
is
ca
r
r
ied
o
u
t
u
s
in
g
ca
teg
o
r
ical
cr
o
s
s
-
en
tr
o
p
y
lo
s
s
as (
1
)
.
=
−
1
∑
∑
,
.
l
og
(
̂
,
)
=
1
=
1
(
1
)
I
n
(
1
)
,
th
e
v
ar
iab
le
y
k
,
c
an
d
̂
,
r
ep
r
esen
ts
g
r
o
u
n
d
tr
u
t
h
o
n
e
-
h
o
t
lab
el
ass
o
ciate
d
f
o
r
c
cla
s
s
an
d
p
r
ed
icted
p
r
o
b
a
b
ilit
y
to
war
d
s
c
class
r
esp
ec
tiv
ely
.
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
also
u
s
es
ea
r
ly
s
t
o
p
p
in
g
c
r
iter
io
n
f
o
r
en
h
an
cin
g
th
e
r
eliab
ilit
y
wh
en
th
e
p
er
f
o
r
m
an
ce
ass
o
ciate
d
with
th
e
v
alid
atio
n
s
et
r
esis
t
s
to
im
p
r
o
v
is
e
wh
ile
b
est
m
o
d
els
ar
e
s
av
ed
u
s
in
g
ch
ec
k
p
o
i
n
ts
an
d
o
v
er
f
itti
n
g
is
r
esis
ted
b
y
em
p
lo
y
in
g
d
r
o
p
o
u
t.
I
t
is
ess
en
tial
to
u
n
d
er
s
tan
d
t
h
e
s
ig
n
if
ican
ce
o
f
th
is
m
o
d
u
le
wh
er
e
th
e
DL
m
o
d
els
will
ten
d
to
o
v
er
f
it
if
t
h
ey
ar
e
u
s
ed
with
o
u
t
r
eg
u
lar
izatio
n
o
r
ea
r
ly
s
to
p
p
i
n
g
cr
iter
io
n
is
u
s
ed
.
T
h
is
f
ac
t
is
ap
p
licab
le
f
o
r
r
estricte
d
lab
eled
d
ata
p
r
esen
t
with
in
HSI
co
n
ten
t.
Ap
ar
t
f
r
o
m
th
is
,
th
er
e
ar
e
also
ch
a
n
ce
s
o
f
p
o
o
r
g
e
n
er
aliza
b
ilit
y
o
f
m
o
d
els
with
o
u
t
p
r
o
p
er
v
alid
atio
n
.
Hen
ce
,
th
is
m
o
d
u
le
en
s
u
r
es
o
f
o
p
tim
al
p
er
f
o
r
m
an
ce
b
y
tr
ain
in
g
e
n
h
an
ce
m
en
t
ca
r
r
ied
o
u
t
h
er
e
with
o
u
t
an
y
n
ee
d
o
f
o
v
e
r
tr
ain
in
g
.
T
h
e
c
o
n
tr
ib
u
tio
n
o
f
th
is
m
o
d
u
le
is
th
at
it
en
h
a
n
ce
s
th
e
co
n
v
er
g
en
ce
to
an
o
p
tim
al
s
o
lu
tio
n
an
d
en
h
a
n
ce
s
th
e
m
o
d
el
r
o
b
u
s
tn
ess
wh
ile
o
v
er
f
itti
n
g
is
p
r
ev
e
n
ted
in
th
e
v
ase
o
f
class
if
y
in
g
HSI
co
n
ten
ts
o
n
d
ata
-
s
ca
r
ce
e
co
s
y
s
tem
.
T
h
e
n
e
x
t
s
ec
tio
n
p
r
esen
ts
d
is
cu
s
s
io
n
o
f
th
e
r
esu
lt
ac
co
m
p
lis
h
ed
f
r
o
m
im
p
lem
en
tin
g
th
e
p
r
o
p
o
s
ed
A
SS
T
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
d
a
p
tive
tr
a
n
s
fo
r
mer a
r
ch
itectu
r
e
fo
r
s
ca
la
b
le
ea
r
th
…
(
D
ev
en
d
r
a
K
u
ma
r
S
a
r
a
g
o
o
r
Ma
d
a
n
a
y
a
ka
)
827
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
h
as
u
s
e
d
Pav
ia
Un
iv
er
s
ity
h
y
p
er
s
p
ec
t
r
al
d
ataset
th
at
h
as
b
ee
n
ca
p
tu
r
ed
u
s
in
g
r
ef
lectiv
e
o
p
tics
s
y
s
tem
im
ag
in
g
s
p
ec
tr
o
m
eter
(
R
OSI
S)
s
en
s
o
r
o
v
er
th
e
Pav
ia
city
,
I
tal
y
.
T
h
is
co
n
s
is
t
o
f
1
0
3
s
p
ec
tr
al
b
a
n
d
s
af
ter
o
p
tin
g
o
u
t
f
o
r
wate
r
ab
s
o
r
p
tio
n
an
d
n
o
is
y
b
an
d
f
r
o
m
o
r
ig
in
a
l
1
1
5
b
a
n
d
s
.
T
h
e
d
im
en
s
io
n
o
f
im
ag
e
is
6
1
0
×
3
4
0
p
i
x
els
wh
ile
th
e
s
p
atial
r
eso
lu
tio
n
o
f
a
n
im
ag
e
is
1
.
3
m
eter
s
th
at
co
v
er
s
9
lab
eled
class
es
s
u
ch
as
b
u
i
ld
in
g
s
,
tr
ee
s
,
m
ea
d
o
ws,
an
d
asp
h
alt
th
at
ar
e
u
s
u
ally
s
ee
n
o
v
er
t
h
e
p
r
ac
tical
s
ce
n
ar
io
s
o
f
u
r
b
a
n
ec
o
s
y
s
tem
.
T
h
e
ASST
m
o
d
el
is
s
u
b
jecte
d
to
co
m
p
a
r
ativ
e
a
n
aly
s
is
with
th
e
two
co
n
v
en
tio
n
al
s
ch
em
e
i.e
.
,
M3
D
-
C
NN
an
d
3
D
-
C
NN
m
eth
o
d
alo
n
g
with
SVM.
Alth
o
u
g
h
,
th
er
e
ar
e
v
ar
io
u
s
ex
is
tin
g
AI
s
ch
em
es
d
ep
lo
y
i
n
g
ML
an
d
DL
m
eth
o
d
s
to
war
d
s
an
aly
zin
g
HSI
co
n
ten
ts
lead
in
g
to
war
d
s
class
if
icatio
n
,
b
o
th
3
D
C
NN
an
d
M3
D
-
C
NN
h
as
b
ee
n
p
ar
ticu
lar
ly
s
elec
ted
m
ain
ly
d
u
e
to
th
eir
p
o
ten
tial
p
er
f
o
r
m
an
ce
as
well
as q
u
ite
f
r
eq
u
en
tly
ad
o
p
ted
m
eth
o
d
s
t
o
war
d
s
ac
q
u
ir
in
g
s
p
atial
an
d
s
p
ec
tr
al
f
ea
tu
r
es
f
r
o
m
th
e
HSI
co
n
ten
ts
.
T
h
e
s
tu
d
y
h
as
s
elec
ted
5
×
5
as
th
e
p
atch
s
i
ze
f
o
r
ac
q
u
ir
in
g
th
e
lo
g
ical
s
p
atial
co
n
tex
t
wh
ile
r
esis
tin
g
p
o
s
s
ib
le
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
.
A
b
atch
s
ize
o
f
3
2
h
as
b
ee
n
co
n
s
id
er
ed
to
p
r
o
v
id
e
an
eq
u
ilib
r
iu
m
b
etwe
en
g
r
ad
ie
n
t stab
ilit
y
an
d
tr
ain
in
g
tim
e
(
T
T
)
.
T
h
e
ac
co
m
p
lis
h
ed
o
u
tco
m
e
is
s
h
o
w
n
in
T
ab
le
1
.
T
h
e
ass
ess
m
en
t
also
co
n
s
id
er
s
0
.
5
d
r
o
p
o
u
t
r
ate
f
o
r
m
in
im
izin
g
th
e
o
v
e
r
f
itti
n
g
.
Ad
am
o
p
tim
izer
is
u
s
ed
f
o
r
its
ad
ap
tiv
e
ca
p
ab
ilit
y
to
war
d
s
m
in
im
izin
g
n
o
is
y
g
r
ad
ien
ts
.
T
h
e
r
ate
o
f
lear
n
i
n
g
is
k
ep
t
at
0
.
0
0
1
th
at
o
f
f
er
s
s
t
ab
le
co
n
v
er
g
en
ce
in
p
r
o
p
o
s
ed
m
o
d
ellin
g
.
All
th
e
ab
o
v
e
-
m
e
n
tio
n
ed
p
er
f
o
r
m
an
c
e
m
etr
ic
h
as
b
ee
n
u
s
ed
f
o
r
a
s
s
es
s
in
g
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
p
r
o
p
o
s
ed
ASST
m
o
d
el
in
co
n
tr
ast
to
ex
is
tin
g
m
o
d
els
o
f
SVM,
M3
D
-
C
NN,
an
d
3
D
-
C
NN.
I
t
s
h
o
u
ld
b
e
n
o
ted
th
at
t
h
e
s
co
r
e
s
m
en
tio
n
ed
in
T
ab
le
1
h
as
ar
r
iv
ed
af
ter
r
ig
o
r
o
u
s
ass
ess
m
en
t
o
n
m
u
ltip
le
r
o
u
n
d
s
to
co
n
f
i
r
m
th
at
th
er
e
is
a
p
o
ten
tial
co
n
s
is
ten
cy
in
all
t
h
e
r
o
u
n
d
s
.
Ov
er
all
ac
cu
r
ac
y
(
OA)
is
d
ef
in
ed
as
th
e
r
atio
o
f
th
e
cu
m
u
lativ
e
n
u
m
b
er
o
f
p
r
ec
is
ely
class
if
ied
s
am
p
les
to
th
e
cu
m
u
lativ
e
n
u
m
b
er
o
f
s
am
p
les.
Kap
p
a
c
o
ef
f
icien
t
(
KC
)
is
a
s
tatis
t
ical
esti
m
ate
a
s
s
o
ciate
d
with
th
e
ac
cu
r
ac
y
co
r
r
ec
ted
o
r
in
ter
-
r
ater
ag
r
ee
m
en
t
f
o
r
ch
an
ce
.
Av
e
r
ag
e
ac
cu
r
ac
y
(
AA)
m
etr
ic
is
ca
lcu
lated
as
m
ea
n
o
f
th
e
s
tan
d
alo
n
e
class
o
f
ac
cu
r
ac
y
o
v
er
all
th
e
class
e
s
.
T
T
r
ef
er
s
to
to
tal
d
u
r
atio
n
co
n
s
u
m
ed
b
y
th
e
f
r
am
ewo
r
k
t
o
lear
n
f
r
o
m
th
e
tr
ai
n
in
g
d
ata
o
v
er
all
t
h
e
ep
o
ch
s
.
R
esp
o
n
s
e
tim
e
(
R
T
)
m
etr
ic
is
also
k
n
o
wn
as
in
f
er
en
ce
tim
e
an
d
is
s
tat
ed
as
d
u
r
atio
n
c
o
n
s
u
m
ed
b
y
th
e
tr
ain
ed
m
o
d
el
f
o
r
class
if
y
in
g
a
b
atch
o
f
s
am
p
le
o
r
a
n
o
v
el
in
p
u
t sam
p
le.
T
ab
le
1
.
Nu
m
e
r
ical
s
co
r
e
ac
co
m
p
lis
h
ed
M
o
d
e
l
O
A
(
%)
KC
A
A
(
%)
TT
(
mi
n
)
R
T
(
ms)
A
S
S
T
9
7
.
2
6
0
.
9
6
0
9
9
8
.
4
5
25
15
S
V
M
9
4
.
3
4
0
.
9
2
5
0
9
2
.
9
8
60
45
M3D
-
C
N
N
9
5
.
7
6
0
.
9
4
5
0
9
5
.
0
8
75
38
3D
-
C
N
N
9
6
.
5
3
0
.
9
5
1
9
7
.
5
7
55
40
Fig
u
r
e
2
r
ep
r
esen
ts
th
e
g
r
ap
h
ical
o
u
tco
m
e
o
f
th
e
co
m
p
ar
ati
v
e
an
aly
s
is
b
ein
g
ca
r
r
ie
d
o
u
t
t
o
ev
alu
ate
th
e
ef
f
ec
tiv
en
ess
o
f
all
th
e
co
n
s
id
er
ed
m
o
d
els.
T
h
e
f
in
al
tr
ain
in
g
ac
cu
r
ac
y
is
n
o
ted
as
9
9
.
3
%
wh
ile
f
in
al
v
alid
atio
n
ac
cu
r
ac
y
is
n
o
ted
as
9
8
.
7
%.
T
h
e
o
p
tim
al
ep
o
ch
b
ein
g
s
elec
ted
is
3
2
d
ep
e
n
d
i
n
g
u
p
o
n
m
ax
im
u
m
v
alid
atio
n
ac
cu
r
ac
y
.
T
h
e
ac
co
m
p
lis
h
ed
o
u
tco
m
e
s
h
o
wca
s
es
th
at
p
r
o
p
o
s
ed
ASST
m
o
d
el
ev
en
tu
ally
ex
ce
lled
s
u
p
er
io
r
p
e
r
f
o
r
m
an
ce
o
n
all
a
s
p
ec
ts
o
f
p
er
f
o
r
m
an
ce
m
etr
ic
in
co
n
tr
ast
to
ex
is
tin
g
s
y
s
tem
.
T
h
e
OA
,
KC
,
AA
,
TT
,
an
d
RT
o
u
tco
m
es
ar
e
s
h
o
wn
in
Fig
u
r
e
s
2
(
a
)
to
2
(
e)
r
esp
ec
tiv
ely
.
T
h
e
d
is
cu
s
s
io
n
is
ca
r
r
ied
o
u
t
with
r
esp
ec
t to
all
th
e
p
er
f
o
r
m
a
n
ce
m
etr
ic
as f
o
llo
ws:
i)
I
n
ter
p
r
etatio
n
o
f
OA
s
co
r
es
:
th
e
o
u
tco
m
e
s
h
o
wca
s
es
s
u
p
er
i
o
r
ab
ilit
y
o
f
th
e
m
o
d
el
to
wa
r
d
s
ap
p
r
o
p
r
iately
class
if
y
in
g
a
m
ajo
r
p
r
o
p
o
r
tio
n
o
f
th
e
h
y
p
e
r
s
p
ec
tr
al
p
ix
els
o
v
er
all
th
e
class
es.
T
h
e
lo
we
r
p
er
f
o
r
m
an
ce
d
em
o
n
s
tr
atio
n
o
f
SVM
is
attr
ib
u
ted
to
th
eir
r
estricte
d
ca
p
a
city
in
m
o
d
ellin
g
th
e
c
o
m
p
le
x
s
p
atial
an
d
s
p
ec
tr
al
d
ep
en
d
en
cies
to
war
d
s
th
e
h
y
p
e
r
s
p
ec
tr
al
d
ata.
On
th
e
o
th
er
h
a
n
d
,
th
ese
ch
allen
g
es
ar
e
a
d
d
r
ess
ed
b
y
DL
f
r
am
ewo
r
k
e.
g
.
,
M3
D
-
C
NN
an
d
3
D
-
C
NN
in
m
u
ch
b
etter
way
in
co
n
tr
ast
to
cla
s
s
ical
lear
n
in
g
m
o
d
els.
Ho
wev
er
,
DL
m
o
d
el
s
also
en
co
u
n
ter
lim
itatio
n
s
e
.
g
.
,
o
v
er
f
itti
n
g
wh
ich
is
m
ain
ly
d
u
e
to
t
h
e
in
ad
eq
u
ate
m
o
d
ellin
g
o
f
s
p
ec
tr
al
co
r
r
elatio
n
o
r
i
n
ad
eq
u
ate
s
p
atial
co
n
te
x
t.
T
h
e
p
r
o
p
o
s
ed
ASST
f
r
am
ewo
r
k
h
as
ex
h
ib
ited
en
h
a
n
ce
d
p
er
f
o
r
m
a
n
ce
m
ain
ly
d
u
e
to
its
in
clu
s
io
n
o
f
th
e
tr
an
s
f
o
r
m
er
-
o
r
ien
ted
d
esig
n
s
tr
u
ctu
r
e
th
at
is
k
n
o
wn
to
ad
ap
tiv
ely
i
n
teg
r
ate
b
o
th
s
p
atial
an
d
s
p
ec
tr
al
atten
tio
n
m
et
h
o
d
s
.
ii)
I
n
ter
p
r
etatio
n
o
f
KC
s
co
r
es
:
th
e
r
elativ
ely
lo
wer
v
alu
e
o
f
KC
f
o
r
SVM
(
=
0
.
92
)
also
d
em
o
n
s
tr
ates
th
e
ch
allen
g
es
ass
o
ciate
d
in
ac
co
m
p
lis
h
in
g
t
h
e
co
n
s
is
ten
t
class
if
icatio
n
ag
r
ee
m
en
t
with
r
esp
ec
t
to
th
e
class
ical
m
eth
o
d
s
.
Fo
r
th
e
in
cr
ea
s
ed
v
alu
e
o
f
OA
ass
o
ciate
d
with
th
e
p
r
o
p
o
s
ed
ASST
m
o
d
el
an
d
b
etter
s
tr
ateg
y
t
o
war
d
s
f
ea
tu
r
e
ex
tr
ac
tio
n
,
it
is
a
n
ticip
ated
th
at
KC
s
co
r
e
will
b
e
ev
en
t
u
ally
b
etter
th
an
th
e
e
x
is
tin
g
m
o
d
el
ex
h
ib
itin
g
r
eliab
ilit
y
f
o
r
h
ig
h
-
o
r
d
er
class
if
icatio
n
.
T
h
e
f
ea
tu
r
es
ar
e
ad
ap
tiv
ely
weig
h
ted
b
y
th
e
tr
an
s
f
o
r
m
er
r
esu
ltin
g
in
m
in
im
ized
m
is
class
if
icatio
n
an
d
m
a
x
im
izatio
n
o
f
co
n
s
is
ten
cy
to
war
d
s
class
ag
r
ee
m
en
t.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
824
-
8
3
0
828
iii)
I
n
ter
p
r
etatio
n
o
f
AA
s
co
r
es:
th
e
p
r
im
e
ca
u
s
e
o
f
s
h
o
r
tco
m
in
g
s
ass
o
ciate
d
with
th
e
e
x
is
tin
g
s
y
s
tem
o
r
ig
in
ates
f
r
o
m
u
n
ev
en
p
er
f
o
r
m
an
ce
th
at
is
n
o
ted
o
v
er
v
ar
i
ed
class
es
esp
ec
ially
f
o
r
th
o
s
e
with
less
o
r
s
o
m
etim
es
n
o
n
e
n
u
m
b
er
o
f
l
ab
eled
s
am
p
les.
T
h
e
s
tr
u
g
g
le
to
war
d
s
g
en
er
aliza
tio
n
is
h
ig
h
er
in
C
NN
m
o
d
els
f
o
r
m
in
o
r
ity
class
es
wh
ile
s
o
p
h
is
ticated
f
ea
tu
r
e
le
ar
n
in
g
is
s
er
io
u
s
ly
lack
i
n
g
i
n
SVM
m
o
d
el
wh
ich
lead
s
to
s
k
ewe
d
m
ea
n
ac
cu
r
ac
y
.
T
h
e
d
is
cr
im
in
atin
g
ca
p
ab
ilit
y
to
war
d
s
f
ea
tu
r
es
o
v
er
all
th
e
class
es
ar
e
im
p
r
o
v
ed
u
p
o
n
b
y
th
e
atten
tio
n
-
b
ased
m
eth
o
d
o
f
ASST
m
o
d
el.
T
h
is
is
d
o
n
e
b
y
ad
ap
tiv
ely
em
p
h
asizin
g
o
n
in
f
o
r
m
ativ
e
c
u
es
o
f
s
p
ec
tr
al
an
d
s
p
atial
attr
ib
u
te.
T
h
e
is
s
u
es
o
f
class
im
b
alan
ce
ef
f
ec
t
ar
e
p
o
te
n
tially
m
itig
ated
b
y
th
e
atten
tio
n
-
g
u
i
d
ed
lear
n
in
g
an
d
y
ield
s
to
m
o
r
e
co
n
s
is
ten
t
ac
cu
r
ac
y
p
er
f
o
r
m
an
ce
o
v
er
all
th
e
class
es.
iv
)
I
n
ter
p
r
etatio
n
o
f
T
T
/R
T
s
co
r
es:
ASST
d
em
an
d
s
o
n
ly
2
5
m
in
u
tes
an
d
1
5
s
ec
o
n
d
s
f
o
r
co
m
p
letin
g
th
e
tr
ain
in
g
cy
cle
an
d
g
en
e
r
atin
g
f
in
al
r
esp
o
n
s
e
r
esp
ec
tiv
ely
wh
ich
ar
e
m
u
ch
l
o
wer
d
u
r
atio
n
in
v
o
l
v
ed
in
co
n
tr
ast
to
ex
is
tin
g
m
o
d
els.
Fro
m
th
e
co
n
tex
t
o
f
ex
is
tin
g
m
o
d
els,
3
D
-
C
NN
h
as
ex
h
ib
ited
5
5
m
in
u
tes
wh
ich
is
b
etter
in
co
n
tr
ast
to
its
co
u
n
ter
p
ar
t
3
D
-
C
NN
an
d
SVM
m
o
d
els.
Similar
ly
,
th
er
e
is
n
o
s
ig
n
if
ican
t
d
if
f
er
en
ce
f
o
r
p
e
r
f
o
r
m
an
ce
o
f
R
T
f
o
r
M3
D
-
C
NN
an
d
3
D
-
C
NN
m
u
ch
wh
ich
is
ag
ain
s
lig
h
tly
h
ig
h
er
th
a
n
SVM
m
o
d
el.
T
h
e
p
r
im
e
r
ea
s
o
n
f
o
r
th
is
is
f
o
u
n
d
to
b
e
in
v
o
lv
ed
h
ig
h
er
co
s
t
o
f
co
m
p
u
tatio
n
ass
o
ciate
d
with
th
e
DL
ap
p
r
o
ac
h
es
m
ain
ly
d
u
e
to
th
e
co
m
p
lex
ar
c
h
itectu
r
es
o
f
d
e
ep
lay
er
s
an
d
co
n
v
o
l
u
tio
n
th
at
ca
n
r
estrict
th
eir
p
r
ac
tical
d
e
p
lo
y
m
e
n
t
wh
e
n
it
co
m
es
to
r
eso
u
r
ce
-
co
n
s
tr
ain
ed
s
y
s
tem
s
.
T
h
e
d
ee
p
n
etwo
r
k
s
a
r
e
p
o
te
n
tially
av
o
id
ed
b
y
th
is
ar
ch
itectu
r
e
wh
ile
it
s
till
m
ain
tain
s
its
e
x
p
r
ess
iv
en
ess
y
ield
in
g
to
m
o
r
e
c
o
s
t
-
ef
f
ec
tiv
e
co
m
p
u
tatio
n
al
s
o
lu
tio
n
to
wa
r
d
s
class
if
icatio
n
o
f
HSI
co
n
te
n
ts
.
T
h
e
n
o
v
elty
o
f
an
o
u
tco
m
e
is
its
m
in
im
ized
co
m
p
u
tatio
n
al
b
u
r
d
e
n
,
th
at
is
ev
i
d
en
tly
witn
ess
ed
f
r
o
m
r
ed
u
ce
d
d
u
r
atio
n
s
co
r
es,
with
o
u
t
co
m
p
r
o
m
is
in
g
with
th
e
ac
cu
r
ac
y
s
co
r
es.
T
h
is
g
r
o
u
n
d
o
f
o
u
tco
m
e
s
tr
u
ctu
r
e
o
f
f
e
r
s
ev
id
en
ce
t
o
war
d
s
h
ig
h
er
a
p
p
licab
ilit
y
.
I
t
r
elate
s
to
r
eso
u
r
ce
-
co
n
s
tr
ain
t
ap
p
licatio
n
s
in
ea
r
th
'
s
o
b
s
er
v
atio
n
.
(
a)
(
b
)
(
c)
(
d
)
(
e)
Fig
u
r
e
2
.
B
en
ch
m
ar
k
ed
o
u
tco
m
e
of
(
a
)
OA
,
(
b
)
KC
,
(
c)
AA
,
(
d
)
TT
,
an
d
(
e)
RT
4.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
h
as
p
r
esen
ted
a
n
o
v
el
m
o
d
el
n
am
e
d
as
ASST
th
at
is
m
ea
n
t
f
o
r
class
if
y
in
g
HSI
co
n
ten
t
in
o
r
d
er
to
f
u
lf
il
th
e
co
r
e
p
u
r
p
o
s
e
o
f
r
em
o
te
s
en
s
in
g
ap
p
licatio
n
th
at
p
lay
s
a
cr
itical
r
o
le
in
ea
r
th
o
b
s
er
v
atio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
d
a
p
tive
tr
a
n
s
fo
r
mer a
r
ch
itectu
r
e
fo
r
s
ca
la
b
le
ea
r
th
…
(
D
ev
en
d
r
a
K
u
ma
r
S
a
r
a
g
o
o
r
Ma
d
a
n
a
y
a
ka
)
829
T
h
e
s
tu
d
y
is
m
ea
n
t
f
o
r
m
itig
atin
g
th
e
o
n
g
o
in
g
c
h
allen
g
es
as
s
o
ciate
d
with
co
m
p
u
tatio
n
al
i
n
ef
f
icien
cy
,
lim
ited
lab
eled
d
ata,
an
d
h
ig
h
d
im
e
n
s
io
n
ality
.
T
h
e
atten
tio
n
m
eth
o
d
an
d
tr
an
s
f
o
r
m
er
en
c
o
d
er
s
wer
e
u
tili
ze
d
to
co
m
b
in
e
s
p
atial
a
n
d
s
p
ec
tr
al
i
n
f
o
r
m
atio
n
b
y
th
e
m
o
d
el.
B
en
ch
m
ar
k
e
d
d
ataset
o
f
Pav
ia
Un
iv
er
s
ity
h
as
b
ee
n
u
s
ed
to
f
in
d
th
at
ASST
ex
h
ib
its
b
etter
p
er
f
o
r
m
a
n
ce
o
n
ac
c
u
r
ac
y
-
b
ased
m
etr
ic
an
d
d
u
r
atio
n
-
b
ased
m
etr
ic
in
co
n
tr
ast
to
f
r
eq
u
e
n
tly
ad
o
p
ted
class
if
icatio
n
m
o
d
els
u
s
in
g
ML
/DL
m
eth
o
d
s
.
T
h
e
OA
r
e
co
r
d
ed
f
o
r
p
r
o
p
o
s
e
d
m
o
d
el
is
9
7
.
2
6
%
wh
ile
it
h
as
b
ee
n
p
r
o
v
en
t
o
s
ig
n
if
ican
tly
m
in
im
ize
b
o
th
tr
ai
n
in
g
a
n
d
r
e
s
p
o
n
s
e
d
u
r
atio
n
.
T
h
e
ar
ch
itectu
r
al
d
esig
n
p
r
esen
te
d
b
y
ASST
n
o
t
o
n
ly
e
n
h
an
c
es
th
e
class
if
icatio
n
p
er
f
o
r
m
a
n
ce
b
u
t
also
o
f
f
e
r
s
f
aster
co
n
v
er
g
e
n
ce
p
er
f
o
r
m
a
n
ce
to
o
.
W
ith
a
r
ig
o
r
o
u
s
an
al
y
s
is
,
it
i
s
n
o
ted
th
at
ASST
m
o
d
el
ac
co
m
p
lis
h
es
h
ig
h
er
ex
ten
t
o
f
r
o
b
u
s
tn
ess
to
class
im
b
alan
ce
th
e
r
eb
y
m
ak
in
g
it
i
n
cr
ea
s
in
g
ly
a
p
p
lica
b
le
f
o
r
r
ea
l
-
wo
r
ld
d
ep
lo
y
m
e
n
t
to
o
.
A
n
in
te
r
esti
n
g
lear
n
i
n
g
o
u
tco
m
e
o
f
t
h
is
m
o
d
el
is
to
k
n
o
w
th
at
tr
an
s
f
o
r
m
e
r
-
b
ased
a
r
ch
itectu
r
e
is
ca
p
ab
le
o
f
ad
ap
tin
g
to
war
d
s
s
p
atial
-
s
p
ec
tr
al
m
o
d
el
v
er
y
ef
f
ec
tiv
ely
.
T
h
e
f
u
tu
r
e
w
o
r
k
o
f
th
is
s
tu
d
y
will
b
e
f
u
r
th
er
t
o
o
p
tim
ize
t
h
e
lear
n
in
g
m
o
d
el
i
n
p
r
esen
ce
o
f
m
o
r
e
c
o
m
p
lex
s
et
o
f
in
f
o
r
m
atio
n
a
n
d
test
if
y
in
g
th
em
o
n
m
o
r
e
co
m
p
lex
f
o
r
m
o
f
o
b
jectiv
e
f
u
n
ctio
n
.
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
Dev
en
d
r
a
Ku
m
ar
Sar
ag
o
o
r
Ma
d
an
ay
a
k
a
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Dev
an
ath
an
Mu
th
u
k
r
is
h
n
a
n
✓
✓
✓
✓
✓
✓
✓
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
t
h
at
s
u
p
p
o
r
t
th
e
f
i
n
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
av
a
ilab
le
f
r
o
m
th
e
c
o
r
r
esp
o
n
d
in
g
au
th
o
r
,
[
DKSM]
,
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
A
.
B
h
a
r
g
a
v
a
,
A
.
S
a
c
h
d
e
v
a
,
K
.
S
h
a
r
ma,
M
.
H
.
A
l
s
h
a
r
i
f
,
P
.
U
t
h
a
n
sa
k
u
l
,
a
n
d
M
.
U
t
h
a
n
s
a
k
u
l
,
“
H
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
i
n
g
a
n
d
i
t
s
a
p
p
l
i
c
a
t
i
o
n
s:
a
r
e
v
i
e
w
,
”
H
e
l
i
y
o
n
,
v
o
l
.
1
0
,
n
o
.
1
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
h
e
l
i
y
o
n
.
2
0
2
4
.
e
3
3
2
0
8
.
[
2
]
M.
-
F
.
C
h
e
n
g
,
A
.
M
u
k
u
n
d
a
n
,
R
.
K
a
r
mak
a
r
,
M
.
A
.
E
.
V
a
l
a
p
p
i
l
,
J.
J
o
u
h
a
r
,
a
n
d
H
.
-
C
.
W
a
n
g
,
“
M
o
d
e
r
n
t
r
e
n
d
s a
n
d
r
e
c
e
n
t
a
p
p
l
i
c
a
t
i
o
n
s
o
f
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
i
n
g
:
a
r
e
v
i
e
w
,
”
T
e
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
1
3
,
n
o
.
5
,
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
t
e
c
h
n
o
l
o
g
i
e
s
1
3
0
5
0
1
7
0
.
[
3
]
X
.
W
a
n
g
,
J
.
Li
u
,
W
.
C
h
i
,
W
.
W
a
n
g
,
a
n
d
Y
.
N
i
,
“
A
d
v
a
n
c
e
s
i
n
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
me
t
h
o
d
s
w
i
t
h
smal
l
s
a
mp
l
e
s
:
a
r
e
v
i
e
w
,
”
Re
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
5
,
n
o
.
1
5
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
r
s1
5
1
5
3
7
9
5
.
[
4
]
F
.
A
l
i
e
t
a
l
.
,
“
S
p
e
c
t
r
a
l
i
n
t
e
l
l
i
g
e
n
c
e
:
A
I
-
d
r
i
v
e
n
h
y
p
e
r
sp
e
c
t
r
a
l
i
m
a
g
i
n
g
f
o
r
a
g
r
i
c
u
l
t
u
r
a
l
a
n
d
e
c
o
s
y
st
e
m
a
p
p
l
i
c
a
t
i
o
n
s
,
”
Ag
ro
n
o
m
y
,
v
o
l
.
1
4
,
n
o
.
1
0
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
a
g
r
o
n
o
m
y
1
4
1
0
2
2
6
0
.
[
5
]
I
.
P
o
d
l
e
s
n
y
k
h
,
M
.
K
o
v
a
l
e
v
,
a
n
d
P
.
P
l
a
t
o
n
o
v
,
“
To
w
a
r
d
s
t
h
e
f
u
t
u
r
e
o
f
u
b
i
q
u
i
t
o
u
s
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
i
n
g
:
I
n
n
o
v
a
t
i
o
n
s
i
n
se
n
s
o
r
c
o
n
f
i
g
u
r
a
t
i
o
n
s
a
n
d
c
o
st
r
e
d
u
c
t
i
o
n
f
o
r
w
i
d
e
s
p
r
e
a
d
a
p
p
l
i
c
a
b
i
l
i
t
y
,
”
T
e
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
1
2
,
n
o
.
1
1
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
t
e
c
h
n
o
l
o
g
i
e
s
1
2
1
1
0
2
2
1
.
[
6
]
X
.
W
a
n
g
,
“
K
h
a
t
r
i
-
R
a
o
f
a
c
t
o
r
i
z
a
t
i
o
n
b
a
se
d
b
i
-
l
e
v
e
l
s
u
p
p
o
r
t
v
e
c
t
o
r
mac
h
i
n
e
f
o
r
h
y
p
e
r
sp
e
c
t
r
a
l
i
ma
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
I
EEE
J
o
u
r
n
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
A
p
p
l
i
e
d
Ea
r
t
h
O
b
ser
v
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
s
i
n
g
,
v
o
l
.
1
8
,
p
p
.
9
6
3
6
–
9
6
4
9
,
2
0
2
5
,
d
o
i
:
1
0
.
1
1
0
9
/
j
st
a
r
s.
2
0
2
5
.
3
5
5
6
3
5
1
.
[
7
]
M
.
A
.
S
h
a
f
a
e
y
e
t
a
l
.
,
“
P
i
x
e
l
-
w
i
s
e
c
l
a
ssi
f
i
c
a
t
i
o
n
o
f
h
y
p
e
r
sp
e
c
t
r
a
l
i
ma
g
e
s
w
i
t
h
1
D
c
o
n
v
o
l
u
t
i
o
n
a
l
S
V
M
n
e
t
w
o
r
k
s
,
”
I
EEE
A
c
c
e
s
s
,
v
o
l
.
1
0
,
p
p
.
1
3
3
1
7
4
–
1
3
3
1
8
5
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
a
c
c
e
ss.
2
0
2
2
.
3
2
3
1
5
7
9
.
[
8
]
M
.
N
i
c
k
z
a
mi
r
a
n
d
S
.
M
.
S
.
A
.
G
a
n
d
a
b
,
“
A
h
y
b
r
i
d
r
a
n
d
o
m
f
o
r
e
st
a
n
d
C
N
N
f
r
a
mew
o
r
k
f
o
r
t
i
l
e
-
w
i
s
e
o
i
l
-
w
a
t
e
r
c
l
a
ss
i
f
i
c
a
t
i
o
n
i
n
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
e
s
,
”
a
rX
i
v
:
2
5
0
2
.
0
0
2
3
2
,
2
0
2
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
824
-
8
3
0
830
[
9
]
X
.
Li
,
X
.
F
a
n
,
Q
.
L
i
,
a
n
d
X
.
Zh
a
o
,
“
R
S
-
N
e
t
:
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
e
l
a
n
d
c
o
v
e
r
c
l
a
ssi
f
i
c
a
t
i
o
n
b
a
se
d
o
n
s
p
e
c
t
r
a
l
i
mag
e
r
c
o
m
b
i
n
e
d
w
i
t
h
r
a
n
d
o
m f
o
r
e
s
t
a
l
g
o
r
i
t
h
m,
”
E
l
e
c
t
r
o
n
i
c
s
,
v
o
l
.
1
3
,
n
o
.
2
0
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s1
3
2
0
4
0
4
6
.
[
1
0
]
J.
H
w
a
n
g
,
K
.
-
O
.
C
h
o
i
,
S
.
Je
o
n
g
,
a
n
d
S
.
Le
e
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
i
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
e
d
i
b
l
e
v
e
g
e
t
a
b
l
e
o
i
l
s
f
r
o
m
f
a
t
t
y
a
c
i
d
c
o
m
p
o
si
t
i
o
n
s
a
n
d
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
e
s,”
C
u
rre
n
t
Re
s
e
a
rc
h
i
n
Fo
o
d
S
c
i
e
n
c
e
,
v
o
l
.
8
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
r
f
s.2
0
2
4
.
1
0
0
7
4
2
.
[
1
1
]
H
.
C
h
e
n
,
F
.
M
i
a
o
,
a
n
d
X
.
S
h
e
n
,
“
H
y
p
e
r
sp
e
c
t
r
a
l
r
e
m
o
t
e
se
n
s
i
n
g
i
m
a
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
w
i
t
h
C
N
N
b
a
se
d
o
n
q
u
a
n
t
u
m
g
e
n
e
t
i
c
-
o
p
t
i
m
i
z
e
d
sp
a
r
se
r
e
p
r
e
se
n
t
a
t
i
o
n
,
”
I
EE
E
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
9
9
9
0
0
–
9
9
9
0
9
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
a
c
c
e
ss.
2
0
2
0
.
2
9
9
7
9
1
2
.
[
1
2
]
X
.
Lu
,
D
.
Y
a
n
g
,
F
.
Ji
a
,
Y
.
Y
a
n
g
,
a
n
d
L.
Zh
a
n
g
,
“
H
y
p
e
r
sp
e
c
t
r
a
l
i
m
a
g
e
c
l
a
s
si
f
i
c
a
t
i
o
n
b
a
s
e
d
o
n
mu
l
t
i
l
e
v
e
l
j
o
i
n
t
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
on
n
e
t
w
o
r
k
,
”
I
E
EE
J
o
u
rn
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
Ap
p
l
i
e
d
E
a
rt
h
O
b
serv
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
4
,
p
p
.
1
0
9
7
7
–
1
0
9
8
9
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
j
st
a
r
s.
2
0
2
1
.
3
1
2
3
3
7
1
.
[
1
3
]
T.
-
H
.
H
si
e
h
a
n
d
J.
-
F
.
K
i
a
n
g
,
“
C
o
m
p
a
r
i
s
o
n
o
f
C
N
N
a
l
g
o
r
i
t
h
ms
o
n
h
y
p
e
r
sp
e
c
t
r
a
l
i
m
a
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
i
n
a
g
r
i
c
u
l
t
u
r
a
l
l
a
n
d
s,
”
S
e
n
so
rs
,
v
o
l
.
2
0
,
n
o
.
6
,
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
0
0
6
1
7
3
4
.
[
1
4
]
Q
.
S
h
e
n
mi
n
g
,
L.
X
i
a
n
g
,
a
n
d
G
.
Z
h
i
h
u
a
,
“
A
n
e
w
h
y
p
e
r
s
p
e
c
t
r
a
l
i
m
a
g
e
c
l
a
s
si
f
i
c
a
t
i
o
n
me
t
h
o
d
b
a
se
d
o
n
sp
a
t
i
a
l
-
sp
e
c
t
r
a
l
f
e
a
t
u
r
e
s,
”
S
c
i
e
n
t
i
f
i
c
Re
p
o
r
t
s
,
v
o
l
.
1
2
,
n
o
.
1
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
5
9
8
-
0
2
2
-
0
5
4
2
2
-
5.
[
1
5
]
C
.
S
h
i
,
D
.
L
i
a
o
,
T
.
Z
h
a
n
g
,
a
n
d
L
.
W
a
n
g
,
“
H
y
p
e
r
s
p
e
c
t
r
a
l
i
m
a
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
b
a
se
d
o
n
e
x
p
a
n
si
o
n
c
o
n
v
o
l
u
t
i
o
n
n
e
t
w
o
r
k
,
”
I
E
EE
T
ra
n
s
a
c
t
i
o
n
s
o
n
G
e
o
sc
i
e
n
c
e
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
6
0
,
p
p
.
1
–
1
6
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
t
g
r
s.2
0
2
2
.
3
1
7
4
0
1
5
.
[
1
6
]
F
.
U
l
l
a
h
e
t
a
l
.
,
“
D
e
e
p
h
y
p
e
r
s
p
e
c
t
r
a
l
sh
o
t
s:
d
e
e
p
s
n
a
p
smo
o
t
h
w
a
v
e
l
e
t
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
h
o
t
s
e
n
sem
b
l
e
f
o
r
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
I
EEE
J
o
u
rn
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
A
p
p
l
i
e
d
E
a
rt
h
O
b
ser
v
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
7
,
p
p
.
1
4
–
3
4
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
j
s
t
a
r
s.
2
0
2
3
.
3
3
1
4
9
0
0
.
[
1
7
]
Z.
G
e
,
G
.
C
a
o
,
X
.
L
i
,
a
n
d
P
.
F
u
,
“
H
y
p
e
r
sp
e
c
t
r
a
l
i
ma
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
m
e
t
h
o
d
b
a
s
e
d
o
n
2
D
–
3
D
C
N
N
a
n
d
m
u
l
t
i
b
r
a
n
c
h
f
e
a
t
u
r
e
f
u
si
o
n
,
”
I
EE
E
J
o
u
r
n
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
Ap
p
l
i
e
d
E
a
rt
h
O
b
ser
v
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
3
,
p
p
.
5
7
7
6
–
5
7
8
8
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
j
st
a
r
s.
2
0
2
0
.
3
0
2
4
8
4
1
.
[
1
8
]
C
.
Y
u
,
R
.
H
a
n
,
M
.
S
o
n
g
,
C
.
Li
u
,
a
n
d
C
.
-
I
.
C
h
a
n
g
,
“
A
si
mp
l
i
f
i
e
d
2
D
-
3
D
C
N
N
a
r
c
h
i
t
e
c
t
u
r
e
f
o
r
h
y
p
e
r
sp
e
c
t
r
a
l
i
m
a
g
e
c
l
a
s
si
f
i
c
a
t
i
o
n
b
a
s
e
d
o
n
s
p
a
t
i
a
l
–
s
p
e
c
t
r
a
l
f
u
si
o
n
,
”
I
E
EE
J
o
u
r
n
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
A
p
p
l
i
e
d
E
a
rt
h
O
b
s
e
rv
a
t
i
o
n
s
a
n
d
Re
m
o
t
e
S
e
n
s
i
n
g
,
v
o
l
.
1
3
,
p
p
.
2
4
8
5
–
2
5
0
1
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
j
st
a
r
s.2
0
2
0
.
2
9
8
3
2
2
4
.
[
1
9
]
H
.
W
e
i
,
Y
.
W
a
n
g
,
Y
.
S
u
n
,
J.
Z
h
e
n
g
,
a
n
d
X
.
Y
u
,
“
A
j
o
i
n
t
n
e
t
w
o
r
k
o
f
3
D
-
2
D
C
N
N
f
e
a
t
u
r
e
h
i
e
r
a
r
c
h
y
a
n
d
p
y
r
a
mi
d
a
l
r
e
si
d
u
a
l
mo
d
e
l
f
o
r
h
y
p
e
r
s
p
e
c
t
r
a
l
i
m
a
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
3
,
p
p
.
1
7
2
7
3
–
1
7
2
8
4
,
2
0
2
5
,
d
o
i
:
1
0
.
1
1
0
9
/
a
c
c
e
ss.
2
0
2
5
.
3
5
3
2
0
1
6
.
[
2
0
]
D.
-
H
.
J
u
n
g
,
J.
D
.
K
i
m
,
H
.
-
Y
.
K
i
m,
T.
S
.
Le
e
,
H
.
S
.
K
i
m,
a
n
d
S
.
H
.
P
a
r
k
,
“
A
h
y
p
e
r
sp
e
c
t
r
a
l
d
a
t
a
3
D
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
c
l
a
ss
i
f
i
c
a
t
i
o
n
m
o
d
e
l
f
o
r
d
i
a
g
n
o
s
i
s
o
f
g
r
a
y
m
o
l
d
d
i
s
e
a
se
i
n
st
r
a
w
b
e
r
r
y
l
e
a
v
e
s
,
”
Fr
o
n
t
i
e
rs
i
n
P
l
a
n
t
S
c
i
e
n
c
e
,
v
o
l
.
1
3
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
8
9
/
f
p
l
s
.
2
0
2
2
.
8
3
7
0
2
0
.
[
2
1
]
J.
Z
h
o
u
,
S
.
Z
e
n
g
,
Z
.
X
i
a
o
,
J.
Z
h
o
u
,
H
.
L
i
,
a
n
d
Z.
K
a
n
g
,
“
A
n
e
n
h
a
n
c
e
d
s
p
e
c
t
r
a
l
f
u
si
o
n
3
D
C
N
N
m
o
d
e
l
f
o
r
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
Re
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
4
,
n
o
.
2
1
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
r
s1
4
2
1
5
3
3
4
.
[
2
2
]
Q
.
X
u
,
Y
.
X
i
a
o
,
D
.
W
a
n
g
,
a
n
d
B
.
L
u
o
,
“
C
S
A
-
M
S
O
3
D
C
N
N
:
m
u
l
t
i
s
c
a
l
e
o
c
t
a
v
e
3
D
C
N
N
w
i
t
h
c
h
a
n
n
e
l
a
n
d
s
p
a
t
i
a
l
a
t
t
e
n
t
i
o
n
f
o
r
h
y
p
e
r
s
p
e
c
t
r
a
l
i
ma
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
2
,
n
o
.
1
,
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
r
s1
2
0
1
0
1
8
8
.
[
2
3
]
X
.
W
a
n
,
Y
.
H
e
,
F
.
C
h
e
n
,
Z
.
S
u
n
,
a
n
d
D
.
M
o
,
“
H
LS
K
-
C
A
S
M
a
m
b
a
:
h
y
b
r
i
d
l
a
r
g
e
se
l
e
c
t
i
v
e
k
e
r
n
e
l
a
n
d
c
o
n
v
o
l
u
t
i
o
n
a
l
a
d
d
i
t
i
v
e
s
e
l
f
-
a
t
t
e
n
t
i
o
n
mam
b
a
f
o
r
h
y
p
e
r
s
p
e
c
t
r
a
l
i
m
a
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
J
o
u
rn
a
l
o
f
K
i
n
g
S
a
u
d
U
n
i
v
e
rsi
t
y
-
C
o
m
p
u
t
e
r
a
n
d
I
n
f
o
r
m
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
3
7
,
n
o
.
4
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
0
7
/
s4
4
4
4
3
-
025
-
0
0
0
6
0
-
z.
[
2
4
]
A
.
F
.
M
i
l
i
t
i
n
o
,
H
.
G
o
y
e
n
a
,
U
.
P
.
-
G
o
y
a
,
a
n
d
M
.
D
.
U
g
a
r
t
e
,
“
L
o
g
i
s
t
i
c
r
e
g
r
e
s
s
i
o
n
v
e
r
s
u
s
X
G
B
o
o
s
t
f
o
r
d
e
t
e
c
t
i
n
g
b
u
r
n
e
d
a
r
e
a
s
u
s
i
n
g
s
a
t
e
l
l
i
t
e
i
m
a
g
e
s
,
”
E
n
v
i
r
o
n
m
e
n
t
a
l
a
n
d
E
c
o
l
o
g
i
c
a
l
S
t
a
t
i
s
t
i
c
s
,
v
o
l
.
3
1
,
n
o
.
1
,
p
p
.
5
7
–
7
7
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
6
5
1
-
0
2
3
-
0
0
5
9
0
-
7.
[
2
5
]
A
.
S
a
m
a
t
,
E.
L
i
,
W
.
W
a
n
g
,
S
.
L
i
u
,
C
.
Li
n
,
a
n
d
J.
A
b
u
d
u
w
a
i
l
i
,
“
M
e
t
a
-
X
G
B
o
o
st
f
o
r
h
y
p
e
r
s
p
e
c
t
r
a
l
i
m
a
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
u
si
n
g
e
x
t
e
n
d
e
d
M
S
ER
-
g
u
i
d
e
d
m
o
r
p
h
o
l
o
g
i
c
a
l
p
r
o
f
i
l
e
s,”
Re
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
2
,
n
o
.
1
2
,
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
r
s1
2
1
2
1
9
7
3
.
[
2
6
]
N
.
Li
n
,
J.
F
u
,
R
.
Ji
a
n
g
,
G
.
L
i
,
a
n
d
Q
.
Y
a
n
g
,
“
L
i
t
h
o
l
o
g
i
c
a
l
c
l
a
ssi
f
i
c
a
t
i
o
n
b
y
h
y
p
e
r
s
p
e
c
t
r
a
l
i
m
a
g
e
s
b
a
se
d
o
n
a
t
w
o
-
l
a
y
e
r
X
G
B
o
o
s
t
mo
d
e
l
,
c
o
m
b
i
n
e
d
w
i
t
h
a
g
r
e
e
d
y
a
l
g
o
r
i
t
h
m
,
”
Re
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
5
,
n
o
.
1
5
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
r
s
1
5
1
5
3
7
6
4
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
De
v
e
n
d
r
a
K
u
m
a
r
S
a
r
a
g
o
o
r
M
a
d
a
n
a
y
a
k
a
is
c
u
rre
n
tl
y
wo
rk
in
g
a
s
a
ss
istan
t
p
ro
fe
ss
o
r
in
th
e
D
e
p
a
rtme
n
t
o
f
El
e
c
tro
n
ics
a
n
d
C
o
m
m
u
n
ica
ti
o
n
En
g
in
e
e
rin
g
,
G
o
v
e
rn
m
e
n
t
En
g
i
n
e
e
rin
g
C
o
ll
e
g
e
,
C
h
a
m
a
ra
ja
n
a
g
a
r,
Ka
rn
a
tak
a
.
He
h
a
s
o
v
e
ra
ll
tea
c
h
in
g
e
x
p
e
rie
n
c
e
o
f
1
8
y
e
a
rs
.
He
h
a
s
h
a
n
d
le
d
v
a
ri
o
u
s
c
o
u
rse
fo
r
UG
p
ro
g
ra
m
.
He
h
a
s
c
o
m
p
lete
d
B.
E
.
i
n
El
e
c
tr
o
n
ics
a
n
d
C
o
m
m
u
n
ica
ti
o
n
e
n
g
in
e
e
rin
g
a
t
Na
ti
o
n
a
l
in
st
it
u
te
o
f
E
n
g
i
n
e
e
rin
g
,
M
y
so
re
in
2
0
0
5
a
n
d
M
.
Tec
h
.
i
n
VL
S
I
De
sig
n
a
n
d
E
m
b
e
d
d
e
d
S
y
ste
m
De
sig
n
a
t
S
JC
E,
M
y
so
re
i
n
2
0
0
8
.
He
a
lso
wo
rk
e
d
h
a
s
a
p
ro
jec
t
a
ss
o
c
iate
a
ro
u
n
d
1
y
e
a
r
a
t
IIT
M
a
d
ra
s.
He
h
a
s
p
u
b
li
sh
e
d
r
e
se
a
rc
h
p
a
p
e
rs
o
n
re
p
u
ted
jo
u
rn
a
ls
i
n
v
a
rio
u
s
d
o
m
a
in
.
His
a
re
a
o
f
in
tere
st
in
c
lu
d
e
s
wire
les
s
c
o
m
m
u
n
ica
ti
o
n
,
n
e
two
rk
i
n
g
,
c
o
m
p
u
ter
v
isi
o
n
,
m
a
c
h
in
e
lea
rn
in
g
,
b
i
g
d
a
ta
,
a
n
d
d
e
e
p
lea
rn
i
n
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
sm
d
e
v
e
n
d
ra
k
u
m
a
r1
5
@g
m
a
il
.
c
o
m
.
Dr
.
De
v
a
n
a
t
h
a
n
Mu
th
u
k
r
ish
n
a
n
is
a
n
a
ss
o
c
iate
p
r
o
fe
ss
o
r
in
t
h
e
De
p
a
rtme
n
t
o
f
El
e
c
tro
n
ics
a
n
d
Co
m
m
u
n
ica
ti
o
n
En
g
i
n
e
e
rin
g
a
t
REVA
Un
i
v
e
rsit
y
,
Be
n
g
a
lu
r
u
,
wit
h
o
v
e
r
1
7
y
e
a
rs
o
f
e
x
p
e
rie
n
c
e
in
tea
c
h
i
n
g
,
re
se
a
rc
h
,
a
n
d
a
d
m
in
istrati
o
n
.
He
lea
d
s
th
e
S
a
telli
te
Tec
h
n
o
l
o
g
y
a
n
d
S
p
a
c
e
Ex
p
lo
ra
ti
o
n
(S
TS
E)
R&
D
a
c
ti
v
it
ies
a
t
RE
VA
Un
iv
e
rsity
,
f
o
c
u
si
n
g
o
n
d
e
v
e
lo
p
in
g
g
ro
u
n
d
sta
ti
o
n
s
a
n
d
CAN
S
ATs
f
o
r
sa
telli
te
sig
n
a
l
re
c
e
p
ti
o
n
a
n
d
a
tmo
sp
h
e
ric
a
n
a
ly
sis.
His
re
se
a
rc
h
in
tere
sts
i
n
c
lu
d
e
sa
telli
te
c
o
m
m
u
n
ica
ti
o
n
,
sp
a
c
e
e
x
p
lo
ra
ti
o
n
,
lo
w
-
p
o
we
r
VLS
I,
a
n
d
n
e
u
ra
l
n
e
tw
o
rk
s.
He
h
a
s
e
x
e
c
u
ted
fu
n
d
e
d
p
r
o
jec
ts
u
n
d
e
r
IS
RO
RE
S
P
ON
D,
F
AER,
a
n
d
KSC
S
T,
p
u
b
l
ish
e
d
o
v
e
r
4
0
re
se
a
rc
h
p
a
p
e
rs,
a
n
d
g
u
id
i
n
g
fi
v
e
P
h
.
D.
sc
h
o
lars
.
He
is
a
se
n
io
r
m
e
m
b
e
r
o
f
IEE
E,
fe
ll
o
w
o
f
IET
E,
a
n
d
c
u
rre
n
tl
y
se
rv
e
s
a
s
S
AC
Ch
a
ir
o
f
th
e
IEE
E
Ba
n
g
a
l
o
re
S
e
c
ti
o
n
(
2
0
2
5
)
.
He
c
a
n
b
e
c
o
n
tac
t
e
d
a
t
e
m
a
il
:
d
e
v
a
n
a
t
h
a
n
.
m
@re
v
a
.
e
d
u
.
i
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.