I
nte
rna
t
io
na
l J
o
urna
l o
f
Rec
o
nfig
ura
ble a
nd
E
m
be
dd
e
d Sy
s
t
e
m
s
(
I
J
R
E
S)
Vo
l.
14
,
No
.
2
,
J
u
ly
20
25
,
p
p
.
31
1
~
3
1
9
I
SS
N:
2089
-
4864
,
DOI
:
1
0
.
1
1
5
9
1
/i
j
r
es
.
v
14
.
i
2
.
pp
31
1
-
3
1
9
311
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
es.ia
esco
r
e.
co
m
Build
ing
a pho
ton
ic neural n
etw
o
rk
bas
ed on
m
ul
ti
-
o
pera
nd
m
ul
ti
m
o
de in
ter
f
erence
ring
reso
n
a
tors
T
ha
nh
T
ien D
o
,
H
a
i Y
en
P
h
a
m
,
T
r
un
g
T
ha
nh
L
e
F
a
c
u
l
t
y
o
f
En
g
i
n
e
e
r
i
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
I
n
t
e
r
n
a
t
i
o
n
a
l
S
c
h
o
o
l
(
V
N
U
-
I
S
)
,
V
i
e
t
n
a
m N
a
t
i
o
n
a
l
U
n
i
v
e
r
si
t
y
(
V
N
U
)
,
H
a
n
o
i
,
V
i
e
t
n
a
m
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
1
9
,
2
0
2
4
R
ev
i
s
ed
Dec
3
0
,
2
0
2
4
A
cc
ep
ted
Ma
r
2
9
,
2
0
2
5
P
h
o
t
o
n
ic
n
e
u
ra
l
n
e
tw
o
rk
s
(P
NN
s
)
o
f
f
e
r
si
g
n
if
ica
n
t
p
o
ten
ti
a
l
f
o
r
e
n
h
a
n
c
i
n
g
d
e
e
p
lea
rn
i
n
g
n
e
tw
o
rk
s,
p
ro
v
id
i
n
g
h
ig
h
-
sp
e
e
d
p
r
o
c
e
ss
in
g
a
n
d
l
o
w
e
n
e
rg
y
c
o
n
su
m
p
ti
o
n
.
I
n
th
is
p
a
p
e
r,
w
e
p
re
se
n
t
a
n
o
v
e
l
P
NN
a
rc
h
it
e
c
tu
re
t
h
a
t
e
m
p
lo
y
s
n
o
n
li
n
e
a
r
o
p
ti
c
a
l
n
e
u
r
o
n
s
u
si
n
g
m
u
lt
i
-
o
p
e
ra
n
d
4
×
4
m
u
lt
i
m
o
d
e
in
terf
e
re
n
c
e
(M
M
I)
m
u
lt
i
-
o
p
e
ra
n
d
rin
g
re
so
n
a
to
rs
(
M
ORRs
)
to
e
ff
icie
n
tl
y
p
e
rf
o
r
m
v
e
c
to
r
d
o
t
-
p
r
o
d
u
c
t
c
a
lcu
latio
n
s.
T
h
is
d
e
sig
n
is
in
teg
ra
ted
in
t
o
a
p
h
o
to
n
ic
c
o
n
v
o
lu
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
(P
CNN
)
w
it
h
tw
o
c
o
n
v
o
lu
ti
o
n
a
l
la
y
e
rs
a
n
d
o
n
e
f
u
ll
y
c
o
n
n
e
c
ted
la
y
e
r.
S
im
u
latio
n
e
x
p
e
rim
e
n
ts,
c
o
n
d
u
c
ted
u
sin
g
L
u
m
e
ri
c
a
l
a
n
d
A
n
s
y
s
to
o
ls,
d
e
m
o
n
stra
ted
th
a
t
th
e
m
o
d
e
l
a
c
h
iev
e
d
a
h
ig
h
tes
t
a
c
c
u
ra
c
y
o
f
9
8
.
2
6
%
o
n
t
h
e
M
NIST
d
a
tas
e
t,
w
it
h
tes
t
lo
ss
e
s
sta
b
il
izin
g
a
t
a
p
p
ro
x
im
a
tel
y
0
.
0
4
%
.
T
h
e
p
ro
p
o
se
d
m
o
d
e
l
w
a
s
e
v
a
lu
a
ted
,
d
e
m
o
n
stra
ti
n
g
h
ig
h
c
o
m
p
u
tati
o
n
sp
e
e
d
,
im
p
ro
v
e
d
a
c
c
u
ra
c
y
,
lo
w
sig
n
a
l
lo
ss
,
a
n
d
sc
a
lab
il
it
y
.
T
h
e
se
f
in
d
in
g
s
h
i
g
h
li
g
h
t
t
h
e
m
o
d
e
l’s
p
o
ten
ti
a
l
f
o
r
a
d
v
a
n
c
in
g
d
e
e
p
lea
rn
in
g
a
p
p
li
c
a
ti
o
n
s w
it
h
m
o
re
e
ff
icie
n
t
h
a
rd
w
a
re
i
m
p
le
m
e
n
t
a
ti
o
n
s.
K
ey
w
o
r
d
s
:
4
×4
m
u
lti
m
o
d
e
in
ter
f
er
en
ce
C
o
n
v
o
lu
tio
n
al
n
e
u
r
al
n
e
t
w
o
r
k
s
Hig
h
-
s
p
ee
d
co
m
p
u
ti
n
g
Mu
lti
-
o
p
er
an
d
r
in
g
r
eso
n
ato
r
s
P
h
o
to
n
ic
n
eu
r
al
n
et
w
o
r
k
s
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
:
T
r
u
n
g
T
h
an
h
L
e
Facu
lt
y
o
f
E
n
g
i
n
ee
r
i
n
g
a
n
d
T
e
ch
n
o
lo
g
y
,
I
n
ter
n
atio
n
a
l Sch
o
o
l (
VNU
-
I
S)
Vietn
a
m
Natio
n
al
Un
iv
er
s
it
y
(
VNU)
1
4
4
Xu
an
T
h
u
y
R
o
ad
,
C
au
Gi
a
y
,
Han
o
i
1
0
0
0
0
0
,
Vietn
am
E
m
ail:
t
h
a
n
h
.
le
@
v
n
u
.
ed
u
.
v
n
1.
I
NT
RO
D
UCT
I
O
N
C
o
n
v
o
lu
tio
n
al
n
e
u
r
al
n
et
w
o
r
k
s
(
C
NN
s
)
,
in
f
lu
e
n
ce
d
b
y
t
h
e
p
r
in
cip
les
o
f
b
io
lo
g
ica
l
n
er
v
o
u
s
s
y
s
te
m
s
,
h
av
e
d
ev
elo
p
ed
in
to
a
s
tr
o
n
g
f
o
r
m
o
f
ar
ti
f
icial
n
e
u
r
al
n
et
w
o
r
k
[
1
]
.
T
h
ey
ar
e
u
tili
ze
d
i
n
i
m
a
g
e
r
ec
o
g
n
itio
n
task
s
,
allo
w
i
n
g
f
o
r
a
n
o
tab
le
r
ed
u
ctio
n
in
t
h
e
co
m
p
lex
it
y
o
f
th
e
n
et
w
o
r
k
w
h
ile
s
t
ill
p
r
o
v
id
in
g
h
ig
h
l
y
ac
cu
r
at
e
p
r
ed
ictio
n
s
.
Fu
r
th
er
,
th
er
e
ar
e
m
a
n
y
ap
p
licatio
n
s
,
co
n
s
is
t
in
g
o
f
i
m
a
g
e
class
i
f
icatio
n
,
o
b
ject
d
etec
tio
n
,
f
ac
ial
r
ec
o
g
n
itio
n
,
r
ea
l
-
t
i
m
e
lan
g
u
a
g
e
tr
an
s
latio
n
,
an
d
o
t
h
er
r
elate
d
f
ield
s
[
2
]
,
[
3
]
.
T
h
e
g
r
o
w
in
g
co
m
p
le
x
it
y
o
f
ap
p
licatio
n
s
li
k
e
s
el
f
-
d
r
iv
i
n
g
v
eh
ic
les
a
n
d
clo
u
d
-
b
ased
ar
ti
f
icial
in
tel
lig
e
n
ce
(
A
I
)
s
er
v
ice
s
is
d
r
iv
i
n
g
t
h
e
n
ee
d
f
o
r
f
aster
,
m
o
r
e
en
er
g
y
-
e
f
f
icie
n
t n
e
u
r
o
m
o
r
p
h
ic
h
ar
d
w
ar
e.
Ho
w
e
v
er
,
cu
r
r
en
t
m
et
h
o
d
s
,
w
h
ich
lar
g
e
l
y
r
el
y
o
n
th
e
v
o
n
Neu
m
an
n
co
m
p
u
t
in
g
ar
ch
itect
u
r
e,
f
ac
e
a
f
u
n
d
a
m
en
ta
l
ch
alle
n
g
e
in
v
o
l
v
in
g
a
tr
ad
e
-
o
f
f
b
et
w
ee
n
t
h
e
s
p
ee
d
o
f
d
ata
tr
an
s
f
er
an
d
t
h
e
a
m
o
u
n
t
o
f
e
n
er
g
y
co
n
s
u
m
ed
.
T
h
is
li
m
itatio
n
ar
is
es
m
a
in
l
y
f
r
o
m
t
h
e
f
ac
t
t
h
at,
in
th
e
s
e
s
y
s
te
m
s
,
m
e
m
o
r
y
,
an
d
p
r
o
ce
s
s
in
g
u
n
it
s
ar
e
p
h
y
s
ica
ll
y
s
ep
ar
ated
,
lead
in
g
to
i
n
e
f
f
icien
c
ies
i
n
p
er
f
o
r
m
a
n
ce
a
n
d
en
er
g
y
co
n
s
u
m
p
tio
n
[
4
]
-
[
7
]
.
T
h
is
s
ep
ar
atio
n
cr
ea
tes
b
o
ttlen
ec
k
s
w
h
e
n
h
a
n
d
li
n
g
lar
g
e
v
o
lu
m
e
s
o
f
d
ata,
m
a
k
i
n
g
it
d
i
f
f
ic
u
lt
to
s
i
m
u
lta
n
eo
u
s
l
y
o
p
tim
ize
b
o
th
s
p
ee
d
an
d
en
er
g
y
ef
f
icie
n
c
y
.
Fu
r
t
h
er
m
o
r
e,
co
n
v
o
l
u
tio
n
o
p
er
atio
n
al
la
y
er
s
m
a
k
e
u
p
ab
o
u
t
9
0
%
o
f
a
C
NN
'
s
co
m
p
u
tatio
n
s
[
8
]
.
W
h
ile
th
er
e
is
s
tr
o
n
g
p
ar
alleli
s
m
w
it
h
i
n
la
y
er
s
,
d
ata
d
ep
en
d
en
cies
ac
r
o
s
s
la
y
er
s
ch
alle
n
g
e
an
y
ef
f
o
r
t
o
f
in
ter
-
la
y
er
p
ar
alleliza
tio
n
.
T
h
is
lead
s
to
s
ca
lab
ilit
y
is
s
u
e
s
w
it
h
p
o
w
er
an
d
p
er
f
o
r
m
a
n
ce
,
w
h
ic
h
ar
e
w
o
r
s
en
ed
b
y
t
h
e
in
h
er
e
n
t p
o
w
er
a
n
d
s
p
ee
d
li
m
i
ts
o
f
elec
tr
o
n
ic
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
14
,
No
.
2
,
J
u
l
y
20
25
:
311
-
3
1
9
312
P
h
o
to
n
ic
n
e
u
r
al
n
et
w
o
r
k
s
(
P
NNs)
ar
e
co
n
s
id
er
ed
h
i
g
h
l
y
p
r
o
m
i
s
i
n
g
f
o
r
t
h
e
n
e
x
t
g
en
er
atio
n
o
f
h
ar
d
w
ar
e
p
r
o
ce
s
s
o
r
s
d
esig
n
ed
f
o
r
n
eu
r
o
m
o
r
p
h
ic
co
m
p
u
t
in
g
.
P
h
o
to
n
ic
d
ev
ices
h
av
e
v
er
y
l
o
w
lo
s
s
,
h
elp
i
n
g
to
g
r
ea
tl
y
m
i
n
i
m
ize
s
ig
n
al
at
ten
u
atio
n
.
Fu
r
t
h
er
m
o
r
e,
th
e
y
ca
n
o
v
er
co
m
e
t
h
e
b
an
d
w
id
t
h
b
o
ttlen
ec
k
s
f
o
u
n
d
in
elec
tr
ical
s
y
s
te
m
s
,
ac
h
ie
v
in
g
h
i
g
h
co
m
p
u
ti
n
g
s
p
ee
d
s
[
9
]
-
[
1
2
]
.
Mo
r
eo
v
er
,
in
P
NNs,
lig
h
t
tr
an
s
m
is
s
io
n
d
ir
ec
tl
y
f
ac
ilit
ate
s
d
at
a
p
r
o
c
ess
i
n
g
,
ef
f
ec
ti
v
el
y
el
i
m
i
n
ati
n
g
th
e
n
ee
d
f
o
r
d
ata
s
h
u
ttli
n
g
,
w
h
ic
h
is
a
k
e
y
in
e
f
f
ic
ien
c
y
i
n
t
h
e
v
o
n
Ne
u
m
an
n
co
m
p
u
ti
n
g
m
o
d
el.
O
v
er
r
ec
en
t
y
ea
r
s
,
P
NN
s
h
a
v
e
g
ar
n
er
ed
s
i
g
n
if
ican
t
atten
tio
n
f
o
r
en
ab
lin
g
h
i
g
h
-
s
p
ee
d
,
lar
g
e
-
s
ca
le,
an
d
h
ig
h
l
y
p
ar
allel
o
p
tical
n
eu
r
o
m
o
r
p
h
ic
h
ar
d
w
ar
e.
Var
io
u
s
p
h
o
to
n
ic
C
NN
i
m
p
le
m
e
n
tati
o
n
s
h
a
v
e
b
ee
n
class
i
f
ied
in
to
f
o
u
r
g
r
o
u
p
s
:
o
p
tical
C
NNs
b
ased
o
n
lig
h
t
d
if
f
r
ac
tio
n
[
1
3
]
-
[
1
7
]
,
o
p
tical
C
NNs
u
s
i
n
g
li
g
h
t
i
n
ter
f
er
en
c
e
[
1
8
]
-
[
2
1
]
,
o
p
tical
C
NNs
u
t
ilizin
g
w
a
v
elen
g
t
h
d
iv
is
io
n
m
u
ltip
le
x
i
n
g
[
2
2
]
-
[
2
4
]
,
an
d
o
p
tical
C
NNs r
el
y
i
n
g
o
n
tu
n
ab
le
o
p
tical
atten
u
atio
n
[
2
5
]
,
[
2
6
]
.
Ho
w
e
v
er
,
p
r
ev
io
u
s
tr
ad
itio
n
al
P
NNs
ar
e
co
n
s
tr
ain
ed
b
y
h
i
g
h
ar
ea
co
s
t
s
an
d
t
h
e
li
m
ita
ti
o
n
o
f
o
n
e
m
u
ltip
l
y
-
ac
c
u
m
u
la
te
(
M
A
C
)
o
p
er
atio
n
p
er
p
h
o
to
n
ic
d
ev
ice.
Usi
n
g
m
atr
ix
s
i
n
g
u
lar
v
a
l
u
e
d
ec
o
m
p
o
s
itio
n
(
SVD)
an
d
u
n
itar
y
m
a
tr
ix
p
ar
am
etr
izat
io
n
as
d
escr
ib
ed
in
R
ec
k
et
a
l.
[
2
7
]
an
d
R
ib
eir
o
et
a
l.
[
2
8
]
.
Sh
e
n
et
a
l.
[
2
9
]
d
ev
elo
p
ed
an
d
i
m
p
le
m
e
n
ted
a
f
u
ll
y
P
NN
,
ac
h
iev
i
n
g
a
m
u
lti
la
y
e
r
p
er
ce
p
tr
o
n
(
ML
P
)
ar
ch
itect
u
r
e
th
r
o
u
g
h
ar
r
ay
s
o
f
Ma
ch
-
Z
e
h
n
d
er
in
ter
f
er
o
m
eter
s
(
MZ
I
s
)
.
Ho
w
ev
er
,
its
ar
ea
co
s
t
i
m
p
r
o
v
e
m
e
n
t
is
li
m
ited
.
T
h
ese
s
ca
lab
ilit
y
li
m
i
tatio
n
s
ar
e
a
k
e
y
ch
alle
n
g
e
th
at
P
NNs
s
ee
k
to
o
v
er
co
m
e.
A
h
ar
d
w
ar
e
-
s
o
f
t
w
ar
e
c
o
-
d
esig
n
o
f
s
li
m
m
ed
P
NN
b
ased
o
n
MZ
I
s
h
a
s
b
ee
n
p
r
o
p
o
s
ed
[
3
0
]
to
ac
h
iev
e
a
r
ed
u
ctio
n
o
f
1
5
%
to
3
8
%
in
th
e
n
u
m
b
er
o
f
MZ
I
s
r
eq
u
ir
ed
f
o
r
d
if
f
er
e
n
t
n
e
t
w
o
r
k
s
ize
s
.
M
o
r
eo
v
er
,
an
FF
T
-
b
ased
ar
ch
itectu
r
e
p
r
o
p
o
s
ed
b
y
Gu
et
a
l.
[
3
1
]
ac
h
iev
es 2
.
2
3
.
7
x
ar
ea
co
s
t i
m
p
r
o
v
e
m
en
t c
o
m
p
ar
ed
w
it
h
o
th
er
P
NNs.
Fu
r
t
h
er
m
o
r
e,
m
a
n
y
P
NNs
o
n
l
y
ac
h
iev
e
ac
c
u
r
ac
y
b
elo
w
9
5
%,
w
h
ic
h
i
s
s
ig
n
i
f
ica
n
tl
y
lo
wer
th
an
th
e
av
er
ag
e
9
9
%
o
f
elec
tr
ical
C
NNs
[
3
2
]
-
[
3
4
]
.
T
h
er
ef
o
r
e,
th
er
e
is
a
s
tr
o
n
g
n
ee
d
to
d
esig
n
a
n
e
w
P
NN
ar
ch
itect
u
r
e
th
at
ca
n
ac
h
ie
v
e
h
ig
h
ac
c
u
r
ac
y
a
n
d
lo
w
lo
s
s
co
m
p
ar
ab
le
to
th
a
t o
f
elec
tr
ical
C
NNs.
T
o
ad
d
r
ess
th
e
ar
ea
co
s
t
lim
i
t
atio
n
s
an
d
ac
h
ie
v
e
h
i
g
h
ac
cu
r
ac
y
w
ith
lo
w
lo
s
s
i
n
P
NNs,
w
e
p
r
o
p
o
s
e
a
n
o
v
el
P
NN
ar
ch
itect
u
r
e
u
s
in
g
a
n
e
w
m
u
lti
-
o
p
er
an
d
r
in
g
r
es
o
n
ato
r
(
MO
R
R
)
d
esig
n
,
w
h
ich
r
elies
o
n
a
s
i
n
g
l
e
4
×
4
m
u
lti
m
o
d
e
in
ter
f
er
en
ce
(
MMI
)
co
u
p
ler
o
n
s
ilico
n
w
a
v
eg
u
id
e
s
.
T
h
e
k
e
y
ad
v
a
n
ta
g
es
o
f
th
i
s
n
e
w
d
e
s
ig
n
in
cl
u
d
e
lo
w
lo
s
s
,
co
m
p
at
ib
ilit
y
w
it
h
co
m
p
le
m
e
n
tar
y
m
eta
l
o
x
id
e
s
e
m
ico
n
d
u
cto
r
(
C
MO
S
)
tech
n
o
lo
g
y
,
h
i
g
h
b
an
d
w
id
t
h
,
r
elax
ed
f
ab
r
icatio
n
to
ler
an
ce
s
,
a
n
d
r
ed
u
ce
d
s
e
n
s
it
iv
i
t
y
to
w
a
v
ele
n
g
th
o
r
p
o
lar
izatio
n
v
ar
iat
io
n
s
d
u
e
to
th
e
u
s
e
o
f
th
e
M
MI
co
u
p
ler
.
T
h
e
m
a
in
co
n
tr
ib
u
tio
n
s
a
r
e
as f
o
llo
w
s
:
-
Scalab
ilit
y
:
w
e
i
n
tr
o
d
u
ce
a
s
ca
lab
le
P
NN
a
r
ch
itectu
r
e
th
at
s
u
r
p
ass
es
p
r
ev
io
u
s
P
NNs
in
ter
m
s
o
f
f
o
o
tp
r
in
t
,
o
f
f
er
i
n
g
a
m
o
r
e
co
m
p
ac
t a
n
d
ef
f
icien
t d
esig
n
.
-
E
f
f
icien
c
y
:
t
h
e
p
r
o
p
o
s
ed
P
N
N
ar
ch
itectu
r
e
s
u
p
p
o
r
ts
h
ig
h
lev
els
o
f
p
ar
allel
co
m
p
u
tatio
n
,
ac
h
iev
in
g
b
o
t
h
h
ig
h
ac
cu
r
ac
y
an
d
lo
w
lo
s
s
.
T
h
is
ad
d
r
ess
es
k
e
y
c
h
alle
n
g
e
s
f
o
u
n
d
in
tr
ad
itio
n
al
elec
tr
ical
C
NNs,
o
f
f
er
i
n
g
i
m
p
r
o
v
ed
p
er
f
o
r
m
an
ce
i
n
P
NN
s
y
s
te
m
s
.
2.
M
E
T
H
O
D
I
n
th
is
s
ec
tio
n
,
w
e
f
ir
s
t
d
esi
g
n
a
n
o
v
el
MO
R
R
f
u
n
ctio
n
i
n
g
as
a
n
eu
r
o
n
in
a
P
NN.
I
t
is
c
o
n
s
tr
u
cted
u
s
i
n
g
a
4
×
4
MM
I
c
o
u
p
ler
w
i
th
s
ilico
n
-
b
ased
w
a
v
e
g
u
id
es.
Fo
llo
w
i
n
g
th
i
s
,
a
co
m
p
lete
P
NN
ar
ch
itect
u
r
e
is
p
r
o
p
o
s
ed
,
d
em
o
n
s
tr
ati
n
g
h
o
w
d
ata
ca
n
b
e
tr
an
s
f
o
r
m
ed
an
d
p
r
o
ce
s
s
ed
w
it
h
i
n
p
h
o
to
n
ic
co
m
p
o
n
e
n
ts
.
T
h
e
o
u
tp
u
t
i
s
g
e
n
er
ated
b
y
co
n
v
er
tin
g
o
p
tical
s
i
g
n
al
s
i
n
to
d
ig
ita
l
d
ata,
w
h
ich
i
s
t
h
en
u
s
ed
to
class
i
f
y
i
m
a
g
e
s
o
n
th
e
MN
I
ST
d
ataset.
T
h
e
ex
p
er
i
m
en
t
to
v
alid
ate
t
h
e
p
r
o
p
o
s
ed
P
NN
an
d
it
s
p
er
f
o
r
m
an
c
e
o
n
t
h
e
d
atase
t
is
co
n
d
u
cted
v
ia
s
i
m
u
la
tio
n
u
s
i
n
g
L
u
m
er
ical
a
n
d
I
n
ter
co
n
n
ec
t
(
An
s
y
s
)
to
o
ls
,
in
teg
r
ated
w
it
h
P
y
t
h
o
n
(
P
y
T
o
r
ch
)
to
ex
ec
u
te
a
f
u
l
l
y
o
p
er
atio
n
al
P
NN.
2
.
1
.
P
r
o
po
s
ed
m
ulti
-
o
pera
nd
ring
re
s
o
na
t
o
r
pro
f
ile
W
e
p
r
esen
t
a
n
o
v
el
MO
R
R
co
n
tr
o
lled
b
y
elec
tr
ical
s
ig
n
al
s
1
,
2
,
.
.
.
(
s
ee
Fig
u
r
e
1
)
.
T
h
e
MO
R
R
is
in
te
g
r
ated
w
it
h
a
4
×
4
MM
I
co
u
p
ler
o
n
s
ilico
n
w
a
v
eg
u
id
es.
T
h
e
w
a
v
e
g
u
id
e
ar
e
m
ad
e
o
f
a
s
ilico
n
on
in
s
u
lato
r
w
it
h
d
i
m
e
n
s
io
n
s
o
f
5
0
0
n
m
i
n
w
i
d
t
h
an
d
2
5
0
n
m
in
h
ei
g
h
t
f
o
r
b
o
th
in
p
u
t
an
d
o
u
tp
u
t
p
ath
s
.
T
h
e
s
elec
ted
len
g
t
h
(
)
an
d
w
id
th
(
)
o
f
th
e
4
×
4
MM
I
ar
e
=
3
2
=
214
an
d
=
6
,
r
esp
ec
tiv
el
y
[
3
5
]
.
W
h
er
e
=
0
−
1
is
t
h
e
b
ea
t
le
n
g
th
o
f
th
e
MM
I
;
0
an
d
1
ar
e
th
e
p
r
o
p
ag
atio
n
co
n
s
t
an
t
s
o
f
th
e
f
u
n
d
a
m
en
ta
l
an
d
f
ir
s
t
-
o
r
d
er
m
o
d
es
s
u
p
p
o
r
ted
b
y
t
h
e
m
u
lti
m
o
d
e
w
a
v
eg
u
id
e
w
it
h
t
h
e
w
id
th
o
f
.
E
ac
h
i
n
p
u
t
s
ig
n
al
in
d
u
ce
s
a
p
h
ase
s
h
i
f
t
(
)
,
w
ith
th
e
to
tal
ac
cu
m
u
lated
p
h
ase
s
h
i
f
t
g
i
v
e
n
b
y
=
∑
=
1
(
)
.
W
e
em
p
lo
y
th
e
m
as
n
e
u
r
o
n
s
in
P
NN
.
T
h
e
in
/te
n
s
it
y
b
u
il
d
u
p
f
u
n
ctio
n
o
f
d
escr
ib
ed
b
y
th
e
f
o
llo
w
in
g
[
3
6
]
:
=
(
)
=
|
2
−
2
+
2
1
−
2
+
2
2
|
,
=
∑
=
1
(
)
∝
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
B
u
ild
in
g
a
p
h
o
to
n
ic
n
e
u
r
a
l n
et
w
o
r
k
b
a
s
ed
o
n
mu
lti
-
o
p
era
n
d
mu
ltimo
d
e
in
terf
eren
ce
…
(
Th
a
n
h
Tien
Do
)
313
I
n
th
i
s
eq
u
at
io
n
,
an
d
r
ep
r
esen
t
t
h
e
o
u
tp
u
t
a
n
d
in
p
u
t
li
g
h
t
i
n
ten
s
it
y
at
t
h
e
i
n
p
u
t
a
n
d
t
h
r
o
u
g
h
p
o
r
ts
,
r
esp
ec
tiv
ely
.
B
o
th
o
f
th
e
m
ar
e
n
o
r
m
alize
d
w
i
th
i
n
a
r
an
g
e
b
et
w
ee
n
0
an
d
1
(
,
∈
[
0
,
1
]
)
,
is
s
elf
-
co
u
p
li
n
g
co
ef
f
ic
ien
t
an
d
is
th
e
s
in
g
le
-
p
as
s
a
m
p
lit
u
d
e
tr
an
s
m
is
s
io
n
f
ac
to
r
.
T
h
e
v
ar
ia
b
le
d
en
o
tes
th
e
in
p
u
t
v
o
lta
g
e
in
t
h
e
elec
tr
ic
d
o
m
a
in
,
a
n
d
is
t
h
e
w
eig
h
t
a
s
s
o
ciate
d
w
it
h
th
a
t
in
p
u
t.
T
h
e
w
ei
g
h
t
ca
n
b
e
ad
j
u
s
ted
th
r
o
u
g
h
v
ar
io
u
s
m
e
ch
an
i
s
m
s
s
u
c
h
as
a
lter
in
g
t
h
e
s
ize
o
f
ac
ti
v
e
r
eg
io
n
s
o
r
m
o
d
i
f
y
i
n
g
d
r
i
v
in
g
v
o
l
tag
e
s
.
T
h
e
(
)
tr
an
s
f
er
f
u
n
ctio
n
is
n
o
n
-
li
n
ea
r
an
d
ca
n
b
e
lev
e
r
ag
ed
in
n
e
u
r
al
n
et
w
o
r
k
co
m
p
u
tatio
n
.
Fig
u
r
e
1
.
A
m
u
l
ti
-
o
p
er
an
d
m
ic
r
o
-
r
in
g
r
eso
n
ato
r
Fo
r
th
e
p
r
o
p
o
s
ed
MO
R
R
s
tr
u
ctu
r
e,
th
e
f
r
ee
s
p
ec
tr
al
r
an
g
e
(
FS
R
)
is
v
er
y
h
i
g
h
.
As
a
r
es
u
lt,
a
h
i
g
h
b
an
d
w
id
t
h
an
d
h
i
g
h
b
it
r
ate
ca
n
b
e
ac
h
iev
ed
.
T
h
e
p
r
o
p
o
s
ed
d
esig
n
is
s
u
itab
le
f
o
r
h
ig
h
-
s
p
ee
d
s
y
s
te
m
s
.
T
h
e
f
i
n
ite
-
d
if
f
er
en
ce
ti
m
e
-
d
o
m
ain
(
FDT
D)
s
i
m
u
latio
n
f
o
r
ON
an
d
OFF
r
eso
n
a
n
ce
s
is
s
h
o
w
n
i
n
F
ig
u
r
e
2
.
T
h
e
Fig
u
r
e
2
(
a)
s
h
o
w
s
t
h
e
O
N
r
eso
n
an
ce
s
tate.
I
n
th
is
s
tate,
t
h
e
s
i
g
n
als
w
i
th
in
t
h
e
w
a
v
e
g
u
id
es
co
n
s
tr
u
cti
v
el
y
in
ter
f
er
e,
r
es
u
lti
n
g
i
n
t
h
e
p
r
o
p
ag
atio
n
o
f
t
h
e
s
i
g
n
al
th
r
o
u
g
h
th
e
w
a
v
e
g
u
id
e
s
y
s
te
m
.
T
h
i
s
s
tate
d
e
m
o
n
s
tr
a
tes
s
tr
o
n
g
s
i
g
n
al
tr
a
n
s
m
i
s
s
io
n
as
th
e
m
icr
o
r
in
g
is
r
eso
n
a
n
t
at
t
h
e
o
p
er
atin
g
w
a
v
ele
n
g
th
.
T
h
e
Fig
u
r
e
2
(
b
)
s
h
o
w
s
th
e
OF
F r
eso
n
a
n
ce
s
tate.
I
n
t
h
i
s
s
tate,
d
estru
c
tiv
e
i
n
ter
f
er
en
c
e
o
cc
u
r
s
,
r
esu
lti
n
g
in
w
ea
k
s
ig
n
al
tr
an
s
m
is
s
io
n
o
r
al
m
o
s
t
n
o
s
i
g
n
al
p
r
o
p
ag
atio
n
.
T
h
is
in
d
icate
s
t
h
at
th
e
m
icr
o
r
in
g
is
n
o
t
r
eso
n
an
t
at
th
e
o
p
er
atin
g
w
a
v
ele
n
g
t
h
,
ef
f
ec
tiv
e
l
y
b
lo
ck
i
n
g
th
e
s
ig
n
al
.
(
a)
(
b
)
Fig
u
r
e
2
.
Si
m
u
latio
n
s
o
f
t
h
e
s
i
g
n
al
p
r
o
p
ag
atio
n
v
ia
t
h
e
MO
R
R
u
s
in
g
a
4
×
4
MM
I
co
u
p
ler
o
n
s
ilico
n
w
a
v
e
g
u
id
es (
a)
ON
an
d
(
b
)
OFF r
eso
n
a
n
ce
s
2
.
2
.
P
r
o
po
s
ed
ph
o
t
o
nic c
o
nv
o
lutio
na
l neura
l net
w
o
rk
B
u
ild
in
g
o
n
th
e
p
r
o
p
o
s
ed
MO
R
R
,
w
e
in
tr
o
d
u
ce
a
n
o
v
el
P
NN,
w
h
er
e
th
e
in
te
n
s
it
y
o
f
in
p
u
t
p
o
w
er
i
s
cr
ea
ted
b
y
m
o
d
u
lati
n
g
m
an
y
d
i
f
f
er
e
n
t
lig
h
t
s
o
u
r
ce
s
,
s
u
ch
as
laser
d
io
d
es
th
at
o
p
er
ate
at
d
if
f
er
en
t
w
a
v
ele
n
g
t
h
s
.
I
n
itial
l
y
,
t
h
ese
la
s
er
d
io
d
es
ar
e
m
u
ltip
le
x
ed
at
f
ir
s
t
an
d
th
e
n
s
p
li
t
o
n
to
m
u
lt
ip
le
r
o
w
s
i
n
an
ar
r
a
y
o
f
m
icr
o
r
in
g
r
e
s
o
n
ato
r
s
(
s
ee
Fig
u
r
e
3
)
.
B
y
u
tili
zi
n
g
p
h
o
to
d
io
d
es
at
th
e
o
u
tp
u
t
to
ca
p
tu
r
e
th
e
p
h
o
to
n
ic
s
i
g
n
al
f
r
o
m
ea
ch
r
o
w
,
th
e
ac
c
u
m
u
lat
ed
p
o
w
er
o
f
all
w
a
v
ele
n
g
t
h
s
is
m
ea
s
u
r
ed
as th
e
o
u
tp
u
t p
o
w
er
o
f
lig
h
t.
Ass
u
m
in
g
w
e
h
av
e
a
n
×
m
atr
i
x
o
f
w
ei
g
h
ts
m
ap
p
in
g
o
n
to
a
n
ar
r
a
y
o
f
MO
R
R
,
i
t
is
s
p
lit
i
n
to
a
g
r
id
×
s
u
b
-
m
atr
ice
s
o
f
b
lo
ck
s
o
f
s
ize
×
(
-
th
e
n
u
m
b
er
o
f
o
p
er
an
d
s
in
t
h
e
r
in
g
r
eso
n
ato
r
s
,
an
d
=
÷
,
=
÷
)
.
E
ac
h
b
lo
ck
×
ap
p
lies
b
lo
ck
-
cir
cu
lan
t
w
eig
h
t
m
atr
i
x
tech
n
iq
u
e
[
3
7
]
.
T
h
is
s
tr
u
ct
u
r
e
allo
w
s
u
s
to
d
ep
lo
y
an
ar
r
a
y
o
f
MO
R
R
s
f
o
r
n
o
n
li
n
ea
r
co
m
p
u
tatio
n
s
i
n
P
NNs.
I
n
t
h
e
−
ℎ
r
o
w
,
t
h
e
MO
R
R
s
r
eso
n
ate
at
th
e
wav
elen
g
t
h
an
d
t
h
e
o
u
tp
u
t
(
)
=
(
∑
=
1
)
.
,
as
m
ea
s
u
r
ed
b
y
a
p
h
o
to
d
etec
to
r
.
Fin
all
y
,
t
h
e
lig
h
t in
ten
s
it
y
in
r
o
w
is
d
etec
te
d
an
d
m
ea
s
u
r
ed
as
(
)
=
∑
=
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
14
,
No
.
2
,
J
u
l
y
20
25
:
311
-
3
1
9
314
Fig
u
r
e
3
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
Si
m
ula
t
io
n
o
f
t
he
pro
po
s
ed
ph
o
t
o
nic neura
l net
w
o
rk
W
e
s
i
m
u
late
a
MO
R
R
n
e
u
r
o
n
w
i
th
t
h
r
ee
o
p
er
an
d
s
(
n
=3
)
at
a
r
eso
n
an
ce
w
a
v
ele
n
g
th
o
f
1
,
554.
28
n
m
.
A
p
h
o
to
n
ic
s
i
m
u
latio
n
is
co
n
d
u
cted
to
test
th
e
f
u
n
c
tio
n
ali
t
y
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el
w
it
h
i
n
a
C
NN.
T
h
e
C
NN
co
n
f
i
g
u
r
atio
n
i
n
cl
u
d
es t
h
e
f
ir
s
t t
w
o
co
n
v
o
lu
tio
n
al
la
y
er
s
,
f
o
l
lo
w
ed
b
y
a
f
latte
n
ed
la
y
er
f
o
r
class
i
f
icatio
n
.
E
ac
h
co
n
v
o
lu
tio
n
al
la
y
er
i
n
clu
d
e
s
3
2
k
er
n
els
w
it
h
a
s
ize
o
f
5
×
5
,
a
s
tr
id
e
o
f
2
,
an
d
a
p
ad
d
i
n
g
o
f
1
.
Af
ter
ea
c
h
co
n
v
o
lu
tio
n
al
la
y
er
,
w
e
ap
p
l
y
b
atch
n
o
r
m
al
izatio
n
(
B
atch
No
r
m
)
u
s
i
n
g
tr
an
s
-
i
m
p
ed
an
ce
a
m
p
li
f
ier
s
(
T
I
A
)
an
d
o
f
f
s
et
v
o
lta
g
e
s
ig
n
als
to
m
i
n
i
m
ize
laten
c
y
o
v
er
h
ea
d
.
B
y
ad
j
u
s
tin
g
th
e
g
ai
n
,
B
atch
No
r
m
ef
f
ec
tiv
e
l
y
s
m
o
o
th
s
th
e
a
m
p
lit
u
d
e
o
f
ac
tiv
atio
n
s
ig
n
a
ls
.
T
h
e
f
latten
ed
la
y
er
is
u
s
ed
f
o
r
th
e
f
in
al
clas
s
i
f
icat
io
n
,
p
r
o
d
u
cin
g
1
0
o
u
tp
u
ts
.
T
h
e
P
NN
is
tr
ain
ed
f
o
r
1
0
0
e
p
o
ch
s
u
s
i
n
g
th
e
A
d
a
m
o
p
ti
m
i
ze
r
.
T
h
e
d
ataset
is
MN
I
ST
.
I
n
p
u
t
d
ata,
k
er
n
el
s
,
an
d
ac
tiv
a
tio
n
f
u
n
cti
o
n
s
ar
e
r
ep
r
esen
ted
in
f
lo
at3
2
f
o
r
m
a
t.
T
h
e
p
r
o
p
o
s
ed
p
h
o
t
o
n
ic
C
NN
m
o
d
el
is
i
m
p
le
m
en
ted
u
s
in
g
an
o
p
en
-
s
o
u
r
ce
lib
r
ar
y
b
ased
o
n
P
y
to
r
ch
an
d
r
u
n
s
o
n
a
m
ac
h
i
n
e
eq
u
i
p
p
ed
w
it
h
an
I
n
tel
C
o
r
e
i5
-
9
7
0
0
C
P
U
an
d
NVI
DI
A
R
T
X
3
6
0
0
GP
U.
3
.
2
.
P
er
f
o
rm
a
nce
o
f
pro
po
s
e
d pho
t
o
nic neura
l net
w
o
rk
T
h
e
r
esu
lts
in
d
icate
th
at
th
e
h
ig
h
e
s
t
te
s
t
ac
c
u
r
ac
y
o
f
t
h
e
m
o
d
el
is
9
8
.
2
6
%
at
ep
o
ch
1
0
0
.
T
h
e
th
r
ee
h
ig
h
e
s
t
test
ac
cu
r
ac
ies
ar
e
9
8
.
7
2
%
(
ep
o
ch
7
9
)
,
9
8
.
7
%
(
ep
o
c
h
8
2
)
,
an
d
9
8
.
6
8
%
(
ep
o
ch
9
1
)
(
s
ee
Fig
u
r
e
4
)
.
T
h
e
test
lo
s
s
s
tab
ilizes
at
ap
p
r
o
x
i
m
atel
y
0
.
0
4
%
f
r
o
m
ep
o
ch
8
0
(
s
ee
Fig
u
r
e
5
)
.
T
h
e
te
s
t
ac
cu
r
ac
y
o
f
o
u
r
m
o
d
el
n
ea
r
l
y
r
ea
ch
es
t
h
e
9
9
%
lev
el,
s
i
m
ilar
to
tr
ad
itio
n
al
C
NN
m
o
d
els
lik
e
VG
G
-
1
6
,
R
esNet
-
5
0
,
an
d
A
le
x
Net
[
3
8
]
-
[
4
1
]
.
H
o
w
ev
er
,
th
e
s
p
ee
d
o
f
o
u
r
m
o
d
el
b
ased
o
n
lig
h
t
-
b
ased
co
m
p
u
t
atio
n
,
r
ep
r
esen
ts
a
s
ig
n
i
f
ica
n
t i
m
p
r
o
v
e
m
e
n
t o
v
er
th
e
d
ig
ita
l c
o
m
p
u
ta
tio
n
s
e
m
p
l
o
y
ed
b
y
t
h
ese
tr
ad
itio
n
a
l CNN
s
[
4
2
]
,
[
43]
.
Fig
u
r
e
4
.
T
r
ain
ac
cu
r
ac
y
a
n
d
test
ac
cu
r
ac
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
B
u
ild
in
g
a
p
h
o
to
n
ic
n
e
u
r
a
l n
et
w
o
r
k
b
a
s
ed
o
n
mu
lti
-
o
p
era
n
d
mu
ltimo
d
e
in
terf
eren
ce
…
(
Th
a
n
h
Tien
Do
)
315
Fig
u
r
e
5
.
T
r
ain
lo
s
s
an
d
te
s
t lo
s
s
W
e
ad
j
u
s
t
th
e
n
u
m
b
er
o
f
co
n
v
o
l
u
tio
n
a
l
la
y
er
s
to
o
b
s
er
v
e
th
e
c
h
an
g
es
in
ac
c
u
r
ac
y
a
n
d
lo
s
s
.
B
y
ev
alu
a
tin
g
d
if
f
er
en
t
co
n
f
i
g
u
r
a
tio
n
s
,
w
e
as
s
es
s
h
o
w
t
h
e
n
u
m
b
er
o
f
co
n
v
o
lu
tio
n
al
la
y
er
s
i
m
p
ac
ts
th
e
m
o
d
el
’
s
p
er
f
o
r
m
a
n
ce
i
n
ter
m
s
o
f
b
o
th
ac
cu
r
ac
y
a
n
d
lo
s
s
.
T
h
e
r
esu
l
t
s
s
h
o
w
t
h
at
as
t
h
e
n
u
m
b
er
o
f
c
o
n
v
o
l
u
tio
n
al
la
y
er
s
in
cr
ea
s
es,
ac
c
u
r
ac
y
i
m
p
r
o
v
e
s
w
h
ile
lo
s
s
d
ec
r
ea
s
e
s
(
s
ee
T
ab
le
1
)
.
T
h
e
o
p
tim
al
s
et
u
p
in
cl
u
d
es
t
h
r
ee
co
n
v
o
lu
tio
n
al
la
y
er
s
an
d
o
n
e
f
u
ll
y
co
n
n
ec
ted
la
y
er
,
ac
h
ie
v
i
n
g
th
e
h
i
g
h
est
ac
c
u
r
ac
y
(
9
9
.
0
7
%)
an
d
th
e
lo
w
est
lo
s
s
(
0
.
0
3
%).
Ho
w
e
v
er
,
ad
d
in
g
a
f
o
u
r
t
h
co
n
v
o
lu
tio
n
al
la
y
er
m
a
y
lead
to
o
v
er
f
it
tin
g
an
d
i
n
cr
ea
s
ed
co
m
p
u
tatio
n
al
co
m
p
le
x
it
y
.
Ad
d
itio
n
all
y
,
i
n
cr
ea
s
i
n
g
th
e
n
u
m
b
er
o
f
la
y
er
s
r
aises
th
e
co
m
p
u
tatio
n
al
lo
ad
an
d
r
eso
u
r
ce
r
eq
u
ir
em
e
n
t
s
,
w
h
ic
h
s
h
o
u
ld
b
e
ca
r
ef
u
ll
y
co
n
s
id
er
e
d
,
p
a
r
ticu
lar
l
y
w
h
e
n
d
ep
lo
y
i
n
g
m
o
d
els
o
n
d
ev
ices
w
it
h
li
m
ited
h
ar
d
w
ar
e
ca
p
ab
ili
ties
.
T
ab
le
1
.
Ou
r
m
o
d
els o
n
d
i
f
f
er
en
t n
u
m
b
er
o
f
la
y
er
s
O
u
r
mo
d
e
l
s
A
c
c
u
r
a
c
y
(
%)
L
o
ss
(
%)
1
.
C
o
n
v
+
1
F
C
9
7
.
3
9
0
.
0
8
2
.
C
o
n
v
+
1
F
C
9
8
.
7
0
.
0
6
3
.
C
o
n
+
1
F
C
9
9
.
0
7
0
.
0
3
3
.
3
.
Co
m
pa
ri
s
o
n w
it
h
o
t
her
ph
o
t
o
nic neura
l net
w
o
rk
W
e
r
u
n
o
u
r
m
o
d
el
u
s
i
n
g
a
p
r
ec
is
io
n
o
f
7
b
its
f
o
r
b
o
th
i
n
p
u
ts
a
n
d
w
ei
g
h
ts
.
T
h
e
m
o
d
el
a
ch
iev
e
s
a
n
ac
cu
r
ac
y
o
f
ap
p
r
o
x
i
m
atel
y
9
8
.
1
%
(
s
ee
Fi
g
u
r
e
6
)
,
w
it
h
a
lo
s
s
o
f
ar
o
u
n
d
0
.
0
4
%.
W
h
en
co
m
p
ar
ed
to
o
th
er
P
NN
(
s
ee
T
a
b
le
2
)
,
th
e
r
esu
lt
s
i
n
d
ic
ate
th
at
o
u
r
m
o
d
el
o
u
tp
er
f
o
r
m
s
th
e
m
i
n
ter
m
s
o
f
ac
cu
r
ac
y
o
n
t
h
e
s
a
m
e
MN
I
ST
d
ataset.
Fig
u
r
e
6
.
T
est ac
cu
r
ac
y
o
n
7
-
b
it f
o
r
p
ar
am
e
ter
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
14
,
No
.
2
,
J
u
l
y
20
25
:
311
-
3
1
9
316
T
ab
le
2
.
C
o
m
p
ar
is
o
n
b
et
w
ee
n
o
u
r
m
o
d
el
w
i
th
o
t
h
er
p
r
ev
io
u
s
P
NN
m
o
d
els
M
o
d
e
l
L
a
y
e
r
s
Q
u
a
n
t
i
z
a
t
i
o
n
A
c
c
u
r
a
c
y
(
%)
O
u
r
mo
d
e
l
2
C
o
n
v
+
1
F
C
7
b
i
t
9
8
.
1
I
D
N
N
-
V
M
M
[
1
6
]
2
F
C
/
9
2
.
6
M
R
R
-
V
M
M
[
1
7
]
3
F
C
4
b
i
t
9
7
.
4
1
M
ZI
-
V
M
M
[
1
8
]
2
F
C
5
b
i
t
7
6
.
7
0
P
C
M
-
V
M
M
[
1
9
]
1
C
o
n
v
+
1
F
C
7
b
i
t
9
5
.
3
0
T
O
P
S
-
C
A
[
2
0
]
C
o
n
v
+
1
F
C
7
b
i
t
88
3
.
4
.
Are
a
re
du
ct
io
n a
nd
s
y
s
t
e
m
ef
f
iciency
B
y
ta
k
i
n
g
ad
v
an
ta
g
e
o
f
u
s
in
g
t
h
e
b
lo
ck
-
cir
c
u
la
n
t
weig
h
t
m
atr
i
x
tec
h
n
iq
u
e
p
r
esen
ted
b
y
Din
g
et
a
l.
[3
7
]
,
o
u
r
m
o
d
el
ca
n
r
ed
u
ce
t
h
e
n
u
m
b
er
o
f
w
ei
g
h
t
m
atr
ice
s
f
o
r
b
etter
ef
f
icie
n
c
y
an
d
ar
ea
r
ed
u
ctio
n
.
Fig
u
r
e
7
illu
s
tr
ates
t
h
e
tr
a
n
s
f
o
r
m
a
tio
n
f
r
o
m
a
n
u
n
s
tr
u
ct
u
r
ed
w
ei
g
h
t
m
atr
ix
w
it
h
1
8
p
ar
am
eter
s
to
a
b
lo
ck
-
cir
cu
lan
t
w
eig
h
t
m
atr
i
x
w
it
h
o
n
l
y
6
p
ar
a
m
eter
s
in
a
n
e
u
r
al
n
et
w
o
r
k
.
I
n
t
h
e
b
lo
ck
-
cir
cu
lan
t
m
a
tr
ix
,
r
o
w
s
ar
e
g
en
er
ated
b
y
c
y
c
licall
y
s
h
if
ti
n
g
th
e
ele
m
e
n
ts
o
f
t
h
e
f
ir
s
t
r
o
w
,
s
i
g
n
i
f
ica
n
tl
y
r
ed
u
ci
n
g
t
h
e
n
u
m
b
er
o
f
s
to
r
ed
p
ar
am
eter
s
.
Fig
u
r
e
7
.
B
lo
ck
-
cir
cu
la
n
t
w
ei
g
h
t te
ch
n
iq
u
e
[3
7
]
B
y
u
tili
s
i
n
g
t
h
e
MO
R
R
d
ev
ic
e,
w
e
ca
n
co
m
p
r
es
s
a
v
ec
to
r
d
o
t
-
p
r
o
d
u
ct
o
p
er
atio
n
in
to
a
s
in
g
le
m
icr
o
-
r
in
g
.
Sp
ec
if
ical
l
y
,
a
×
m
a
tr
ix
in
o
u
r
m
o
d
el
co
n
ta
in
s
×
s
u
b
-
m
a
tr
ices
o
f
b
lo
c
k
o
f
s
ize
×
(
-
t
h
e
n
u
m
b
er
o
f
o
p
er
an
d
s
in
th
e
r
in
g
r
eso
n
ato
r
)
.
E
ac
h
s
u
b
-
m
a
tr
ix
o
f
s
ize
×
is
u
s
ed
in
th
e
f
o
r
m
o
f
a
cir
cu
lan
t
m
atr
i
x
.
T
h
is
m
ea
n
s
th
a
t
t
h
e
s
u
b
-
m
atr
i
x
h
as
a
s
p
ec
ial
s
tr
u
ct
u
r
e
in
w
h
ic
h
ea
ch
r
o
w
(
o
r
co
lu
m
n
)
i
s
a
c
y
clic
s
h
i
f
t
o
f
th
e
f
ir
s
t
r
o
w
(
o
r
co
lu
m
n
)
.
T
h
is
cir
cu
la
n
t
s
tr
u
ct
u
r
e
allo
w
s
f
o
r
r
ed
u
cin
g
t
h
e
n
u
m
b
er
o
f
weig
h
ts
r
e
q
u
ir
ed
a
n
d
m
ak
in
g
m
a
tr
ix
m
u
l
tip
licatio
n
m
o
r
e
e
f
f
icien
t.
F
u
r
t
h
er
,
th
e
cir
cu
lan
t
m
atr
i
x
s
tr
u
ct
u
r
e
allo
ws
f
o
r
ar
ea
r
ed
u
ctio
n
in
d
ev
ice
an
d
w
a
v
ele
n
g
th
u
s
a
g
e.
A
s
a
r
esu
lt,
it
lead
s
to
(
2
)
tim
es
ar
ea
r
ed
u
c
tio
n
a
n
d
ti
m
es
w
a
v
ele
n
g
t
h
u
s
a
g
e
s
a
v
i
n
g
s
.
I
n
co
m
p
ar
is
o
n
,
o
th
er
P
NNs,
s
u
ch
as
t
h
e
o
n
e
d
esi
g
n
ed
a
n
d
f
ab
r
icate
d
b
y
S
h
e
n
et
a
l.
[
2
9
]
,
r
eq
u
ir
e
O
(
2
)
MZ
I
s
f
o
r
i
m
p
lem
en
tatio
n
.
B
esid
es
th
i
s
,
w
a
v
elen
g
t
h
s
av
in
g
s
m
ak
e
o
p
tical
s
y
s
te
m
s
m
o
r
e
ef
f
icien
t
i
n
ter
m
s
o
f
co
s
t,
e
n
er
g
y
,
a
n
d
p
er
f
o
r
m
a
n
ce
w
h
ile
r
ed
u
cin
g
co
m
p
le
x
it
y
a
n
d
o
p
ti
m
izi
n
g
p
r
o
ce
s
s
i
n
g
ca
p
ab
ilit
ies.
4.
CO
NCLU
SI
O
N
I
n
th
i
s
s
tu
d
y
,
w
e
i
n
tr
o
d
u
ce
d
a
n
o
v
el
p
h
o
to
n
ic
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
et
w
o
r
k
(
P
C
N
N)
b
ased
o
n
m
u
lti
-
o
p
er
an
d
4
×
4
MM
I
m
icr
o
-
r
i
n
g
r
eso
n
ato
r
s
,
d
em
o
n
s
tr
ati
n
g
its
ef
f
ec
t
iv
e
n
es
s
in
d
ee
p
lear
n
in
g
tas
k
s
li
k
e
i
m
a
g
e
r
ec
o
g
n
itio
n
.
Ou
r
m
o
d
el
ac
h
ie
v
ed
h
ig
h
ac
cu
r
ac
y
,
lo
w
s
i
g
n
al
lo
s
s
,
r
ap
id
lig
h
t
-
b
ased
co
m
p
u
tat
io
n
,
s
h
o
w
i
n
g
g
r
ea
t
p
o
ten
tial
co
m
p
ar
ed
to
tr
ad
itio
n
al
d
ig
ita
l
C
N
N
ar
ch
itec
tu
r
es.
T
h
e
f
le
x
ib
ilit
y
o
f
th
is
m
o
d
el
allo
w
s
it
to
b
e
ad
ap
ted
to
v
ar
io
u
s
C
NN
s
tr
u
c
tu
r
es,
in
c
lu
d
i
n
g
Alex
Ne
t,
R
es
Net,
an
d
VGG1
6
,
o
f
f
er
in
g
p
r
o
m
is
i
n
g
r
esu
l
ts
f
o
r
f
u
tu
r
e
A
I
an
d
d
ee
p
lear
n
in
g
ad
v
an
ce
m
e
n
ts
.
T
h
is
s
tu
d
y
s
h
o
w
s
t
h
at
P
NN
ca
n
s
ig
n
i
f
ica
n
tl
y
i
m
p
r
o
v
e
s
p
ee
d
an
d
ef
f
ic
ien
c
y
i
n
lar
g
e
-
s
ca
le
d
ata
p
r
o
ce
s
s
in
g
an
d
a
u
to
n
o
m
o
u
s
s
y
s
te
m
s
.
F
u
t
u
r
e
r
esear
ch
s
h
o
u
ld
f
o
c
u
s
o
n
in
teg
r
ati
n
g
t
h
is
m
o
d
el
in
to
m
o
r
e
co
m
p
le
x
ar
c
h
itect
u
r
es
a
n
d
test
i
n
g
its
r
ea
l
-
ti
m
e
p
er
f
o
r
m
an
ce
.
Ov
er
co
m
in
g
c
h
alle
n
g
es
i
n
lar
g
e
-
s
ca
le
i
m
p
le
m
en
ta
tio
n
a
n
d
ad
ap
tin
g
it
t
o
v
ar
io
u
s
e
n
v
ir
o
n
m
en
ts
w
il
l
f
u
r
t
h
er
en
h
an
ce
i
ts
i
m
p
ac
t.
T
h
ese
f
in
d
i
n
g
s
s
u
g
g
e
s
t
th
at
P
NN
co
u
ld
p
lay
a
k
e
y
r
o
le
in
th
e
f
u
t
u
r
e
o
f
A
I
,
o
f
f
er
in
g
f
aster
an
d
m
o
r
e
ef
f
icien
t
s
o
lu
tio
n
s
f
o
r
p
r
ac
tical
ap
p
licatio
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
B
u
ild
in
g
a
p
h
o
to
n
ic
n
e
u
r
a
l n
et
w
o
r
k
b
a
s
ed
o
n
mu
lti
-
o
p
era
n
d
mu
ltimo
d
e
in
terf
eren
ce
…
(
Th
a
n
h
Tien
Do
)
317
F
UNDIN
G
I
NF
O
RM
AT
I
O
N
T
h
e
au
th
o
r
s
th
a
n
k
th
e
M
in
i
s
t
r
y
o
f
Scie
n
ce
an
d
T
ec
h
n
o
lo
g
y
o
f
Viet
n
a
m
f
o
r
f
u
n
d
in
g
t
h
i
s
r
esear
ch
u
n
d
er
p
r
o
j
ec
t n
u
m
b
er
ĐT
ĐL
.
C
N
-
9
2
/2
1
.
AUTHO
R
CO
NT
RIB
UT
I
O
NS ST
A
T
E
M
E
NT
T
h
is
w
o
r
k
f
o
llo
w
s
th
e
C
o
n
tr
i
b
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
e
d
iT)
.
T
h
e
s
p
ec
if
ic
co
n
tr
ib
u
tio
n
s
o
f
ea
ch
au
th
o
r
ar
e
d
etailed
b
elo
w
:
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
T
h
an
h
T
ien
Do
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
T
r
u
n
g
T
h
an
h
L
e
✓
✓
✓
✓
✓
✓
✓
✓
Hai
Yen
P
h
a
m
✓
✓
✓
✓
✓
✓
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
si
s
I
:
I
n
v
e
st
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
si
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
t
h
o
r
s
s
tate
n
o
co
n
f
lic
t o
f
i
n
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
au
th
o
r
s
co
n
f
ir
m
t
h
at
all
d
ata
s
u
p
p
o
r
tin
g
t
h
e
f
in
d
i
n
g
s
o
f
th
is
s
tu
d
y
ar
e
av
ailab
le
w
it
h
i
n
th
e
ar
ticle
an
d
its
s
u
p
p
le
m
e
n
tar
y
m
ater
ia
ls
RE
F
E
R
E
NC
E
S
[
1
]
A
.
K
.
Jai
n
,
J.
M
a
o
,
a
n
d
K
.
M
.
M
o
h
i
u
d
d
i
n
,
“
A
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s:
A
t
u
t
o
r
i
a
l
,
”
C
o
m
p
u
t
e
r
,
v
o
l
.
2
9
,
n
o
.
3
,
p
p
.
3
1
–
4
4
,
M
a
r
.
1
9
9
6
,
d
o
i
:
1
0
.
1
1
0
9
/
2
.
4
8
5
8
9
1
.
[
2
]
N
.
S
h
a
b
a
i
r
o
u
,
E.
C
o
h
e
n
,
O
.
W
a
g
n
e
r
,
D
.
M
a
l
k
a
,
a
n
d
Z
.
Z
a
l
e
v
sk
y
,
“
C
o
l
o
r
i
mag
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
a
n
d
r
e
c
o
n
s
t
r
u
c
t
i
o
n
u
s
i
n
g
a
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
o
n
mu
l
t
i
mo
d
e
f
i
b
e
r
i
mag
e
s:
t
o
w
a
r
d
s
a
n
a
l
l
-
o
p
t
i
c
a
l
d
e
si
g
n
,
”
O
p
t
i
c
s
L
e
t
t
e
rs
,
v
o
l
.
4
3
,
n
o
.
2
2
,
p
p
.
5
6
0
3
–
5
6
0
6
,
N
o
v
.
2
0
1
8
,
d
o
i
:
1
0
.
1
3
6
4
/
o
l
.
4
3
.
0
0
5
6
0
3
.
[
3
]
A
.
V
o
u
l
o
d
i
mo
s,
N
.
D
o
u
l
a
mi
s
,
A
.
D
o
u
l
a
mi
s,
a
n
d
E
.
P
r
o
t
o
p
a
p
a
d
a
k
i
s,
“
D
e
e
p
L
e
a
r
n
i
n
g
f
o
r
C
o
mp
u
t
e
r
V
i
s
i
o
n
:
A
B
r
i
e
f
R
e
v
i
e
w
,
”
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
N
e
u
r
o
sc
i
e
n
c
e
,
p
p
.
1
–
1
3
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
1
8
/
7
0
6
8
3
4
9
.
[
4
]
M
.
N
a
y
l
o
r
a
n
d
C
.
R
u
n
c
i
m
a
n
,
“
T
h
e
R
e
d
u
c
e
r
o
n
:
W
i
d
e
n
i
n
g
t
h
e
v
o
n
N
e
u
ma
n
n
b
o
t
t
l
e
n
e
c
k
f
o
r
g
r
a
p
h
r
e
d
u
c
t
i
o
n
u
s
i
n
g
a
n
F
P
G
A
,
”
i
n
L
e
c
t
u
re
N
o
t
e
s
i
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
(
i
n
c
l
u
d
i
n
g
s
u
b
ser
i
e
s
L
e
c
t
u
r
e
N
o
t
e
s
i
n
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
L
e
c
t
u
re
N
o
t
e
s
i
n
Bi
o
i
n
f
o
rm
a
t
i
c
s)
,
v
o
l
.
5
0
8
3
,
p
p
.
1
2
9
–
1
4
6
,
2
0
0
8
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
54
0
-
8
5
3
7
3
-
2
_
8
.
[
5
]
D
.
A
.
B
.
M
i
l
l
e
r
,
“
A
t
t
o
j
o
u
l
e
O
p
t
o
e
l
e
c
t
r
o
n
i
c
s
f
o
r
L
o
w
-
En
e
r
g
y
I
n
f
o
r
mat
i
o
n
P
r
o
c
e
ssi
n
g
a
n
d
C
o
mm
u
n
i
c
a
t
i
o
n
s,”
J
o
u
rn
a
l
o
f
L
i
g
h
t
w
a
v
e
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
3
5
,
n
o
.
3
,
p
p
.
3
4
6
–
3
9
6
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
JL
T
.
2
0
1
7
.
2
6
4
7
7
7
9
.
[
6
]
T
.
N
.
T
h
e
i
s
a
n
d
H
.
S
.
P
.
W
o
n
g
,
“
T
h
e
E
n
d
o
f
M
o
o
r
e
’
s
L
a
w
:
A
N
e
w
B
e
g
i
n
n
i
n
g
f
o
r
I
n
f
o
r
mat
i
o
n
T
e
c
h
n
o
l
o
g
y
,
”
C
o
m
p
u
t
i
n
g
i
n
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
9
,
n
o
.
2
,
p
p
.
4
1
–
5
0
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
M
C
S
E.
2
0
1
7
.
2
9
.
[
7
]
H
.
M
a
r
k
r
a
m
e
t
a
l
.
,
“
R
e
c
o
n
st
r
u
c
t
i
o
n
a
n
d
S
i
mu
l
a
t
i
o
n
o
f
N
e
o
c
o
r
t
i
c
a
l
M
i
c
r
o
c
i
r
c
u
i
t
r
y
,
”
C
e
l
l
,
v
o
l
.
1
6
3
,
n
o
.
2
,
p
p
.
4
5
6
–
4
9
2
,
2
0
1
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
e
l
l
.
2
0
1
5
.
0
9
.
0
2
9
.
[
8
]
J.
C
o
n
g
a
n
d
B
.
X
i
a
o
,
“
M
i
n
i
mi
z
i
n
g
c
o
mp
u
t
a
t
i
o
n
i
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
s,”
L
e
c
t
u
re
N
o
t
e
s i
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
(
i
n
c
l
u
d
i
n
g
su
b
s
e
ri
e
s
L
e
c
t
u
re
N
o
t
e
s
i
n
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
L
e
c
t
u
re
N
o
t
e
s
i
n
Bi
o
i
n
f
o
rm
a
t
i
c
s)
,
v
o
l
.
8
6
8
1
L
N
C
S
,
p
p
.
2
8
1
–
2
9
0
,
2
0
1
4
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
3
1
9
-
1
1
1
7
9
-
7
_
3
6
.
[
9
]
M
.
N
a
z
i
r
z
a
d
e
h
,
M
.
S
h
a
ms
a
b
a
r
d
e
h
,
a
n
d
S
.
J.
B
.
Y
o
o
,
“
En
e
r
g
y
-
Ef
f
i
c
i
e
n
t
a
n
d
H
i
g
h
-
T
h
r
o
u
g
h
p
u
t
N
a
n
o
p
h
o
t
o
n
i
c
N
e
u
r
o
mo
r
p
h
i
c
C
o
mp
u
t
i
n
g
,
”
i
n
C
o
n
f
e
re
n
c
e
o
n
L
a
s
e
rs
a
n
d
E
l
e
c
t
r
o
-
O
p
t
i
c
s
,
W
a
sh
i
n
g
t
o
n
,
D
.
C
.
:
O
S
A
,
2
0
1
8
,
d
o
i
:
1
0
.
1
3
6
4
/
C
L
EO
_
A
T
.
2
0
1
8
.
A
T
h
3
Q
.
2
.
[
1
0
]
Y
.
F
e
i
e
t
a
l
.
,
“
D
e
si
g
n
o
f
t
h
e
l
o
w
-
l
o
ss
w
a
v
e
g
u
i
d
e
c
o
i
l
f
o
r
i
n
t
e
r
f
e
r
o
me
t
r
i
c
i
n
t
e
g
r
a
t
e
d
o
p
t
i
c
g
y
r
o
sco
p
e
s,”
J
o
u
r
n
a
l
o
f
S
e
m
i
c
o
n
d
u
c
t
o
rs
,
v
o
l
.
3
8
,
n
o
.
4
,
p
.
0
4
4
0
0
9
,
A
p
r
.
2
0
1
7
,
d
o
i
:
1
0
.
1
0
8
8
/
1
6
7
4
-
4
9
2
6
/
3
8
/
4
/
0
4
4
0
0
9
.
[
1
1
]
R
.
S
l
a
v
í
k
,
Y
.
P
a
r
k
,
M
.
K
u
l
i
sh
o
v
,
R
.
M
o
r
a
n
d
o
t
t
i
,
a
n
d
J.
A
z
a
ñ
a
,
“
U
l
t
r
a
f
a
st
a
l
l
-
o
p
t
i
c
a
l
d
i
f
f
e
r
e
n
t
i
a
t
o
r
s,
”
O
p
t
i
c
s
E
x
p
ress
,
v
o
l
.
1
4
,
n
o
.
2
2
,
p
p
.
1
0
6
9
9
–
1
0
7
0
7
,
2
0
0
6
,
d
o
i
:
1
0
.
1
3
6
4
/
o
e
.
1
4
.
0
1
0
6
9
9
.
[
1
2
]
J.
H
u
a
n
g
,
C
.
L
i
,
R
.
L
u
,
L
.
L
i
,
a
n
d
Z
.
C
a
o
,
“
B
e
y
o
n
d
t
h
e
1
0
0
G
b
a
u
d
d
i
r
e
c
t
l
y
mo
d
u
l
a
t
e
d
l
a
se
r
f
o
r
sh
o
r
t
r
e
a
c
h
a
p
p
l
i
c
a
t
i
o
n
s,
”
J
o
u
r
n
a
l
o
f
S
e
m
i
c
o
n
d
u
c
t
o
rs
,
v
o
l
.
4
2
,
n
o
.
4
,
p
.
0
4
1
3
0
6
,
A
p
r
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
8
8
/
1
6
7
4
-
4
9
2
6
/
4
2
/
4
/
0
4
1
3
0
6
.
[
1
3
]
J.
C
h
a
n
g
,
V
.
S
i
t
z
ma
n
n
,
X
.
D
u
n
,
W
.
H
e
i
d
r
i
c
h
,
a
n
d
G
.
W
e
t
z
st
e
i
n
,
“
H
y
b
r
i
d
o
p
t
i
c
a
l
-
e
l
e
c
t
r
o
n
i
c
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
w
i
t
h
o
p
t
i
m
i
z
e
d
d
i
f
f
r
a
c
t
i
v
e
o
p
t
i
c
s
f
o
r
i
mag
e
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
S
c
i
e
n
t
i
f
i
c
Re
p
o
r
t
s
,
v
o
l
.
8
,
n
o
.
1
,
p
p
.
1
–
1
0
,
2
0
1
8
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
5
9
8
-
018
-
3
0
6
1
9
-
y.
[
1
4
]
M
.
M
i
sc
u
g
l
i
o
e
t
a
l
.
,
“
M
a
ssi
v
e
l
y
p
a
r
a
l
l
e
l
a
mp
l
i
t
u
d
e
-
o
n
l
y
F
o
u
r
i
e
r
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
O
p
t
i
c
a
,
v
o
l
.
7
,
n
o
.
1
2
,
p
p
.
1
8
1
2
–
1
8
1
9
,
2
0
2
0
,
d
o
i
:
1
0
.
1
3
6
4
/
o
p
t
i
c
a
.
4
0
8
6
5
9
.
[
1
5
]
Z
.
H
u
e
t
a
l
.
,
“
H
i
g
h
-
T
h
r
o
u
g
h
p
u
t
M
u
l
t
i
c
h
a
n
n
e
l
P
a
r
a
l
l
e
l
i
z
e
d
D
i
f
f
r
a
c
t
i
o
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
A
c
c
e
l
e
r
a
t
o
r
,
”
L
a
se
r
a
n
d
Ph
o
t
o
n
i
c
s
Re
v
i
e
w
s
,
v
o
l
.
1
6
,
n
o
.
1
2
,
p
p
.
1
–
9
,
D
e
c
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
2
/
l
p
o
r
.
2
0
2
2
0
0
2
1
3
.
[
1
6
]
W
.
S
h
i
e
t
a
l
.
,
“
L
O
EN
:
L
e
n
sl
e
ss o
p
t
o
-
e
l
e
c
t
r
o
n
i
c
n
e
u
r
a
l
n
e
t
w
o
r
k
e
mp
o
w
e
r
e
d
mac
h
i
n
e
v
i
si
o
n
,
”
L
i
g
h
t
:
S
c
i
e
n
c
e
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
14
,
No
.
2
,
J
u
l
y
20
25
:
311
-
3
1
9
318
1
1
,
n
o
.
1
,
p
.
1
2
1
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
3
7
7
-
0
2
2
-
0
0
8
0
9
-
5.
[
1
7
]
H
.
H
.
Z
h
u
e
t
a
l
.
,
“
S
p
a
c
e
-
e
f
f
i
c
i
e
n
t
o
p
t
i
c
a
l
c
o
mp
u
t
i
n
g
w
i
t
h
a
n
i
n
t
e
g
r
a
t
e
d
c
h
i
p
d
i
f
f
r
a
c
t
i
v
e
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
N
a
t
u
r
e
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
1
3
,
n
o
.
1
,
p
.
1
0
4
4
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
4
6
7
-
022
-
2
8
7
0
2
-
0.
[
1
8
]
X
.
X
u
,
L
.
Z
h
u
,
W
.
Z
h
u
a
n
g
,
L
.
L
u
,
a
n
d
P
.
Y
u
a
n
,
“
A
C
o
n
v
o
l
u
t
i
o
n
N
e
u
r
a
l
N
e
t
w
o
r
k
I
mp
l
e
m
e
n
t
e
d
b
y
T
h
r
e
e
3
×
3
P
h
o
t
o
n
i
c
I
n
t
e
g
r
a
t
e
d
R
e
c
o
n
f
i
g
u
r
a
b
l
e
L
i
n
e
a
r
P
r
o
c
e
sso
r
s,”
P
h
o
t
o
n
i
c
s
,
v
o
l
.
9
,
n
o
.
2
,
p
.
8
0
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
p
h
o
t
o
n
i
c
s9
0
2
0
0
8
0
.
[
1
9
]
S
.
X
u
e
t
a
l
.
,
“
O
p
t
i
c
a
l
c
o
h
e
r
e
n
t
d
o
t
-
p
r
o
d
u
c
t
c
h
i
p
f
o
r
so
p
h
i
s
t
i
c
a
t
e
d
d
e
e
p
l
e
a
r
n
i
n
g
r
e
g
r
e
ssi
o
n
,
”
L
i
g
h
t
:
S
c
i
e
n
c
e
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
0
,
n
o
.
1
,
p
.
2
2
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
3
7
7
-
0
2
1
-
0
0
6
6
6
-
8.
[
2
0
]
C
.
F
e
n
g
e
t
a
l
.
,
“
A
C
o
mp
a
c
t
B
u
t
t
e
r
f
l
y
-
S
t
y
l
e
S
i
l
i
c
o
n
P
h
o
t
o
n
i
c
-
El
e
c
t
r
o
n
i
c
N
e
u
r
a
l
C
h
i
p
f
o
r
H
a
r
d
w
a
r
e
-
Ef
f
i
c
i
e
n
t
D
e
e
p
L
e
a
r
n
i
n
g
,
”
A
C
S
Ph
o
t
o
n
i
c
s
,
v
o
l
.
9
,
n
o
.
1
2
,
p
p
.
3
9
0
6
–
3
9
1
6
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
2
1
/
a
c
sp
h
o
t
o
n
i
c
s
.
2
c
0
1
1
8
8
.
[
2
1
]
X
.
M
e
n
g
e
t
a
l
.
,
“
C
o
mp
a
c
t
o
p
t
i
c
a
l
c
o
n
v
o
l
u
t
i
o
n
p
r
o
c
e
ssi
n
g
u
n
i
t
b
a
se
d
o
n
mu
l
t
i
mo
d
e
i
n
t
e
r
f
e
r
e
n
c
e
,
”
N
a
t
u
re
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
1
4
,
n
o
.
1
,
p
.
3
0
0
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
4
6
7
-
0
2
3
-
3
8
7
8
6
-
x.
[
2
2
]
V
.
B
a
n
g
a
r
i
e
t
a
l
.
,
“
D
i
g
i
t
a
l
El
e
c
t
r
o
n
i
c
s
a
n
d
A
n
a
l
o
g
P
h
o
t
o
n
i
c
s
f
o
r
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
(
D
EA
P
-
C
N
N
s
)
,
”
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
Q
u
a
n
t
u
m
E
l
e
c
t
r
o
n
i
c
s
,
v
o
l
.
2
6
,
n
o
.
1
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
JS
T
Q
E.
2
0
1
9
.
2
9
4
5
5
4
0
.
[
2
3
]
S
.
X
u
,
J.
W
a
n
g
,
S
.
Y
i
,
a
n
d
W
.
Z
o
u
,
“
H
i
g
h
-
o
r
d
e
r
t
e
n
so
r
f
l
o
w
p
r
o
c
e
ssi
n
g
u
s
i
n
g
i
n
t
e
g
r
a
t
e
d
p
h
o
t
o
n
i
c
c
i
r
c
u
i
t
s
,
”
N
a
t
u
re
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
1
3
,
n
o
.
1
,
p
.
7
9
7
0
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
4
6
7
-
0
2
2
-
3
5
7
2
3
-
2.
[
2
4
]
J.
F
e
l
d
man
n
e
t
a
l
.
,
“
P
a
r
a
l
l
e
l
c
o
n
v
o
l
u
t
i
o
n
a
l
p
r
o
c
e
ssi
n
g
u
s
i
n
g
a
n
i
n
t
e
g
r
a
t
e
d
p
h
o
t
o
n
i
c
t
e
n
so
r
c
o
r
e
,
”
N
a
t
u
r
e
,
v
o
l
.
5
8
9
,
n
o
.
7
8
4
0
,
p
p
.
52
–
5
8
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
5
8
6
-
0
2
0
-
0
3
0
7
0
-
1.
[
2
5
]
S
.
X
u
,
J.
W
a
n
g
,
R
.
W
a
n
g
,
J.
C
h
e
n
,
a
n
d
W
.
Z
o
u
,
“
H
i
g
h
-
a
c
c
u
r
a
c
y
o
p
t
i
c
a
l
c
o
n
v
o
l
u
t
i
o
n
u
n
i
t
a
r
c
h
i
t
e
c
t
u
r
e
f
o
r
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
b
y
c
a
sca
d
e
d
a
c
o
u
st
o
-
o
p
t
i
c
a
l
mo
d
u
l
a
t
o
r
a
r
r
a
y
s,”
O
p
t
i
c
s
Ex
p
r
e
ss
,
v
o
l
.
2
7
,
n
o
.
1
4
,
p
p
.
1
9
7
7
8
–
1
9
7
8
7
,
2
0
1
9
,
d
o
i
:
1
0
.
1
3
6
4
/
o
e
.
2
7
.
0
1
9
7
7
8
.
[
2
6
]
F
.
A
sh
t
i
a
n
i
,
A
.
J.
G
e
e
r
s,
a
n
d
F
.
A
f
l
a
t
o
u
n
i
,
“
A
n
o
n
-
c
h
i
p
p
h
o
t
o
n
i
c
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
f
o
r
i
ma
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
N
a
t
u
re
,
v
o
l
.
6
0
6
,
n
o
.
7
9
1
4
,
p
p
.
5
0
1
–
5
0
6
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
8
6
-
0
2
2
-
0
4
7
1
4
-
0.
[
2
7
]
M
.
R
e
c
k
,
A
.
Z
e
i
l
i
n
g
e
r
,
H
.
J.
B
e
r
n
st
e
i
n
,
a
n
d
P
.
B
e
r
t
a
n
i
,
“
Ex
p
e
r
i
me
n
t
a
l
R
e
a
l
i
z
a
t
i
o
n
o
f
A
n
y
D
i
scre
t
e
U
n
i
t
a
r
y
O
p
e
r
a
t
o
r
,
”
P
h
y
si
c
a
l
Re
v
i
e
w
L
e
t
t
e
rs
,
v
o
l
.
7
3
,
n
o
.
2
0
,
p
p
.
5
8
–
6
1
,
1
9
9
4
,
d
o
i
:
1
0
.
1
1
0
3
/
P
h
y
sR
e
v
L
e
t
t
.
7
3
.
5
8
.
[
2
8
]
A
.
R
i
b
e
i
r
o
,
A
.
R
u
o
c
c
o
,
L
.
V
a
n
a
c
k
e
r
,
a
n
d
W
.
B
o
g
a
e
r
t
s,
“
D
e
mo
n
st
r
a
t
i
o
n
o
f
a
4
×
4
-
p
o
r
t
u
n
i
v
e
r
sal
l
i
n
e
a
r
c
i
r
c
u
i
t
,
”
O
p
t
i
c
a
,
v
o
l
.
3
,
n
o
.
1
2
,
p
p
.
1
3
4
8
–
1
3
5
7
,
D
e
c
.
2
0
1
6
,
d
o
i
:
1
0
.
1
3
6
4
/
O
P
TI
C
A
.
3
.
0
0
1
3
4
8
.
[
2
9
]
Y
.
S
h
e
n
,
N
.
C
.
H
a
r
r
i
s,
D
.
E
n
g
l
u
n
d
,
a
n
d
M
.
S
o
l
j
a
c
i
c
,
“
D
e
e
p
l
e
a
r
n
i
n
g
w
i
t
h
c
o
h
e
r
e
n
t
n
a
n
o
p
h
o
t
o
n
i
c
c
i
r
c
u
i
t
s,
”
N
a
t
u
re
p
h
o
t
o
n
i
c
s
,
p
p
.
4
4
1
–
4
4
6
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
E
3
S
.
2
0
1
7
.
8
2
4
6
1
9
0
.
[
3
0
]
Z
.
Z
h
a
o
e
t
a
l
.
,
“
H
a
r
d
w
a
r
e
-
so
f
t
w
a
r
e
C
o
-
d
e
si
g
n
o
f
sl
i
mm
e
d
o
p
t
i
c
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s,”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
A
si
a
a
n
d
S
o
u
t
h
P
a
c
i
f
i
c
D
e
si
g
n
A
u
t
o
m
a
t
i
o
n
C
o
n
f
e
re
n
c
e
,
AS
P
-
D
AC
,
N
e
w
Y
o
r
k
,
N
Y
,
U
S
A
:
A
C
M
,
J
a
n
.
2
0
1
9
,
p
p
.
7
0
5
–
710
,
d
o
i
:
1
0
.
1
1
4
5
/
3
2
8
7
6
2
4
.
3
2
8
7
7
2
0
.
[
3
1
]
J.
G
u
,
Z
.
Z
h
a
o
,
C
.
F
e
n
g
,
M
.
L
i
u
,
R
.
T
.
C
h
e
n
,
a
n
d
D
.
Z
.
P
a
n
,
“
T
o
w
a
r
d
s
A
r
e
a
-
Ef
f
i
c
i
e
n
t
O
p
t
i
c
a
l
N
e
u
r
a
l
N
e
t
w
o
r
k
s:
A
n
F
F
T
-
b
a
se
d
A
r
c
h
i
t
e
c
t
u
r
e
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
Asi
a
a
n
d
S
o
u
t
h
P
a
c
i
f
i
c
D
e
si
g
n
A
u
t
o
m
a
t
i
o
n
C
o
n
f
e
re
n
c
e
,
A
S
P
-
D
AC
,
2
0
2
0
,
p
p
.
4
7
6
–
4
8
1
,
d
o
i
:
1
0
.
1
1
0
9
/
A
S
P
-
D
A
C
4
7
7
5
6
.
2
0
2
0
.
9
0
4
5
1
5
6
.
[
3
2
]
A
.
B
a
l
d
o
mi
n
o
s,
Y
.
S
a
e
z
,
a
n
d
P
.
I
sasi
,
“
A
su
r
v
e
y
o
f
h
a
n
d
w
r
i
t
t
e
n
c
h
a
r
a
c
t
e
r
r
e
c
o
g
n
i
t
i
o
n
w
i
t
h
M
N
I
S
T
a
n
d
E
M
N
I
S
T
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
9
,
n
o
.
1
5
,
p
.
3
1
6
9
,
2
0
1
9
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
9
1
5
3
1
6
9
.
[
3
3
]
A
.
El
S
a
w
y
,
H
.
El
-
B
a
k
r
y
,
a
n
d
M
.
L
o
e
y
,
“
C
N
N
f
o
r
h
a
n
d
w
r
i
t
t
e
n
a
r
a
b
i
c
d
i
g
i
t
s
r
e
c
o
g
n
i
t
i
o
n
b
a
se
d
o
n
L
e
N
e
t
-
5
,
”
i
n
A
d
v
a
n
c
e
s
i
n
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
a
n
d
C
o
m
p
u
t
i
n
g
,
2
0
1
7
,
p
p
.
5
6
5
–
5
7
5
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
3
1
9
-
4
8
3
0
8
-
5
_
5
4
.
[
3
4
]
S
.
N
a
mas
u
d
r
a
,
R
.
C
h
a
k
r
a
b
o
r
t
y
,
S
.
K
a
d
r
y
,
G
.
M
a
n
o
g
a
r
a
n
,
a
n
d
B
.
S
.
R
a
w
a
l
,
“
F
A
S
T
:
F
a
st
A
c
c
e
ssi
n
g
S
c
h
e
me
f
o
r
d
a
t
a
T
r
a
n
smiss
i
o
n
i
n
c
l
o
u
d
c
o
mp
u
t
i
n
g
,
”
Pe
e
r
-
to
-
Pe
e
r
N
e
t
w
o
r
k
i
n
g
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
4
,
n
o
.
4
,
p
p
.
2
4
3
0
–
2
4
4
2
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
2
0
8
3
-
020
-
0
0
9
5
9
-
6.
[
3
5
]
T.
-
T
.
L
e
a
n
d
D
.
-
T
.
L
e
,
“
H
i
g
h
F
S
R
a
n
d
C
r
i
t
i
c
a
l
C
o
u
p
l
i
n
g
C
o
n
t
r
o
l
o
f
M
i
c
r
o
r
i
n
g
R
e
so
n
a
t
o
r
B
a
se
d
o
n
G
r
a
p
h
e
n
e
-
S
i
l
i
c
o
n
M
u
l
t
i
mo
d
e
W
a
v
e
g
u
i
d
e
s,”
W
a
v
e
g
u
i
d
e
T
e
c
h
n
o
l
o
g
i
e
s
i
n
P
h
o
t
o
n
i
c
s
a
n
d
Mi
c
r
o
w
a
v
e
E
n
g
i
n
e
e
ri
n
g
[
W
o
rk
i
n
g
T
i
t
l
e
]
,
2
0
2
0
,
d
o
i
:
1
0
.
5
7
7
2
/
i
n
t
e
c
h
o
p
e
n
.
9
2
2
1
0
.
[
3
6
]
P
.
R
.
P
r
u
c
n
a
l
a
n
d
B
h
a
v
i
n
J
.
S
h
a
st
r
i
,
“
N
e
u
ro
m
o
r
p
h
i
c
p
h
o
t
o
n
i
c
s
,
”
C
R
C
P
r
e
ss,
2
0
1
7
.
d
o
i
:
1
0
.
1
2
0
1
/
9
7
8
1
3
1
5
3
7
3
2
3
4
.
[
3
7
]
C
.
D
i
n
g
e
t
a
l
.
,
“
C
I
R
C
N
N
:
A
c
c
e
l
e
r
a
t
i
n
g
a
n
d
c
o
m
p
r
e
ssi
n
g
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s
u
si
n
g
b
l
o
c
k
-
c
i
r
c
u
l
a
n
t
w
e
i
g
h
t
m
a
t
r
i
c
e
s,”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
A
n
n
u
a
l
I
n
t
e
rn
a
t
i
o
n
a
l
S
y
m
p
o
s
i
u
m
o
n
Mi
c
ro
a
rc
h
i
t
e
c
t
u
re,
MICR
O
,
2
0
1
7
,
p
p
.
3
9
5
–
4
0
8
,
d
o
i
:
1
0
.
1
1
4
5
/
3
1
2
3
9
3
9
.
3
1
2
4
5
5
2
.
[
3
8
]
R
.
C
h
a
u
h
a
n
,
K
.
K
.
G
h
a
n
s
h
a
l
a
,
a
n
d
R
.
C
.
Jo
sh
i
,
“
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
)
f
o
r
I
mag
e
D
e
t
e
c
t
i
o
n
a
n
d
R
e
c
o
g
n
i
t
i
o
n
,
”
D
e
c
.
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
C
C
C
.
2
0
1
8
.
8
7
0
3
3
1
6
.
[
3
9
]
K
.
A
u
l
i
a
sari
,
M
.
W
a
se
f
,
a
n
d
M
.
K
e
r
t
a
n
i
n
g
t
y
a
s,
“
L
e
v
e
r
a
g
i
n
g
V
G
G
1
6
f
o
r
f
i
sh
c
l
a
ss
i
f
i
c
a
t
i
o
n
i
n
a
l
a
r
g
e
-
sca
l
e
d
a
t
a
se
t
,
”
Br
i
l
l
i
a
n
c
e
,
v
o
l
.
3
,
n
o
.
2
,
2
0
2
3
,
d
o
i
:
1
0
.
4
7
7
0
9
/
b
r
i
l
l
i
a
n
c
e
.
v
3
i
2
.
3
2
7
0
[
4
0
]
A
.
S
a
i
n
i
,
K
.
G
u
l
e
r
i
a
,
a
n
d
S
.
S
h
a
r
ma,
“
A
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
f
i
n
e
-
t
u
n
e
d
R
e
sN
e
t
5
0
mo
d
e
l
f
o
r
so
y
b
e
a
n
se
e
d
s
mu
l
t
i
c
l
a
ss
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s o
f
t
h
e
I
EEE
G
l
o
b
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
i
e
s
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
(
G
C
I
T
C
)
,
p
p
.
1
–
6
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
g
c
i
t
c
6
0
4
0
6
.
2
0
2
3
.
1
0
4
2
6
0
7
9
.
[
4
1
]
K
.
B
a
n
a
sz
e
w
sk
a
a
n
d
M
.
P
l
e
c
h
a
w
sk
a
-
W
ó
j
c
i
k
,
“
C
o
mp
a
r
a
t
i
v
e
a
n
a
l
y
si
s
o
f
C
N
N
mo
d
e
l
s
f
o
r
h
a
n
d
w
r
i
t
t
e
n
d
i
g
i
t
r
e
c
o
g
n
i
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
s In
s
t
i
t
u
t
e
,
v
o
l
.
3
2
,
p
p
.
1
7
9
–
1
8
5
,
2
0
2
4
.
[
4
2
]
F
.
P
.
S
u
n
n
y
,
E.
T
a
h
e
r
i
,
M
.
N
i
k
d
a
s
t
,
a
n
d
S
.
P
a
sr
i
c
h
a
,
“
A
S
u
r
v
e
y
o
n
S
i
l
i
c
o
n
P
h
o
t
o
n
i
c
s
f
o
r
D
e
e
p
L
e
a
r
n
i
n
g
,
”
AC
M
J
o
u
r
n
a
l
o
n
Em
e
rg
i
n
g
T
e
c
h
n
o
l
o
g
i
e
s
i
n
C
o
m
p
u
t
i
n
g
S
y
s
t
e
m
s
,
v
o
l
.
1
7
,
n
o
.
4
,
p
p
.
1
–
5
7
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
4
5
/
3
4
5
9
0
0
9
.
[
4
3
]
L
.
D
e
M
a
r
i
n
i
s
,
M
.
C
o
c
o
c
c
i
o
n
i
,
P
.
C
a
st
o
l
d
i
,
a
n
d
N
.
A
n
d
r
i
o
l
l
i
,
“
P
h
o
t
o
n
i
c
N
e
u
r
a
l
N
e
t
w
o
r
k
s:
A
S
u
r
v
e
y
,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
7
,
p
p
.
1
7
5
8
2
7
–
1
7
5
8
4
1
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
9
.
2
9
5
7
2
4
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
B
u
ild
in
g
a
p
h
o
to
n
ic
n
e
u
r
a
l n
et
w
o
r
k
b
a
s
ed
o
n
mu
lti
-
o
p
era
n
d
mu
ltimo
d
e
in
terf
eren
ce
…
(
Th
a
n
h
Tien
Do
)
319
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Th
a
n
h
Tie
n
Do
h
a
s
b
e
e
n
a
wa
r
d
e
d
th
e
M
.
A
.
d
e
g
re
e
f
ro
m
S
w
in
b
u
rn
e
Un
iv
e
rsity
o
f
T
e
c
h
n
o
lo
g
y
,
A
u
stra
li
a
,
sin
c
e
2
0
1
8
.
He
is
c
u
rre
n
tl
y
w
o
rk
in
g
to
w
a
rd
a
P
h
.
D
.
d
e
g
re
e
a
t
th
e
In
tern
a
ti
o
n
a
l
S
c
h
o
o
l
–
V
ietn
a
m
Na
ti
o
n
a
l
Un
iv
e
rsity
Ha
n
o
i.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
IT
a
rc
h
it
e
c
tu
re
s,
e
-
g
o
v
e
rn
m
e
n
t
m
o
d
e
ls,
m
a
c
h
in
e
lea
rn
in
g
a
n
d
o
p
ti
c
a
l
c
o
m
m
u
n
ica
ti
o
n
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
th
a
n
h
d
o
@v
n
u
is.ed
u
.
v
n
.
H
a
i
Ye
n
Ph
a
m
re
c
e
iv
e
d
th
e
M
.
A
.
d
e
g
re
e
f
ro
m
th
e
M
il
it
a
ry
T
e
c
h
n
o
lo
g
y
A
c
a
d
e
m
y
,
V
ietn
a
m
,
in
2
0
0
9
.
S
h
e
is
c
u
rre
n
tl
y
w
o
rk
in
g
to
w
a
rd
a
P
h
.
D
.
d
e
g
re
e
a
t
th
e
In
tern
a
ti
o
n
a
l
S
c
h
o
o
l
–
V
iet
n
a
m
Na
ti
o
n
a
l
Un
iv
e
rsit
y
Ha
n
o
i.
He
r
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
m
a
c
h
in
e
lea
rn
in
g
a
n
d
o
p
ti
c
a
l
c
o
m
m
u
n
ica
ti
o
n
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
y
e
n
p
h
@v
n
u
.
e
d
u
.
v
n
.
Tr
u
n
g
Th
a
n
h
Le
is
a
p
r
o
f
e
ss
o
r
a
n
d
I
n
tern
a
ti
o
n
a
l
S
c
h
o
o
l
-
V
NU
p
r
in
c
ip
a
l
.
He
re
c
e
iv
e
d
a
P
h
.
D
.
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
f
ro
m
L
a
T
ro
b
e
,
Au
stra
li
a
,
in
2
0
0
9
.
W
it
h
o
v
e
r
1
0
0
re
f
e
re
e
d
p
u
b
li
c
a
ti
o
n
s,
h
e
m
a
d
e
se
m
in
a
l
c
o
n
tri
b
u
ti
o
n
s
to
p
h
o
t
o
n
ic
in
teg
ra
ted
c
ircu
it
s
a
n
d
q
u
a
n
t
u
m
o
p
ti
c
s in
c
lu
d
in
g
las
e
r
a
n
d
f
ib
e
r
-
o
p
ti
c
a
l
c
o
m
m
u
n
ica
ti
o
n
sy
ste
m
s,
a
ll
-
o
p
ti
c
a
l
c
o
m
p
u
ti
n
g
a
n
d
o
p
ti
c
a
l
i
n
terc
o
n
n
e
c
ts,
se
n
so
rs
a
n
d
w
irele
ss
se
n
so
r
n
e
tw
o
rk
s,
a
n
d
in
teg
ra
ted
c
ircu
it
sy
ste
m
s.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
th
a
n
h
.
le@
v
n
u
.
e
d
u
.
v
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.