I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
1
6
,
No
.
4
,
A
u
g
u
s
t
20
2
6
,
p
p
.
2
2
2
0
~
2
2
2
9
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
1
6
i
4
.
pp
2
2
2
0
-
2
2
2
9
2220
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
A
co
nv
o
lutiona
l n
eura
l net
wo
rk
-
ba
sed driv
er mon
i
toring
sy
stem f
o
r drows
i
ness
and dis
trac
ti
o
n det
ec
tion
Sa
ra
B
enk
o
uid
er
,
Na
s
re
dd
in
e
L
a
g
ra
a
La
b
o
r
a
t
o
i
r
e
d
'I
n
f
o
r
ma
t
i
q
u
e
e
t
d
e
M
a
t
h
é
mat
i
q
u
e
s,
U
n
i
v
e
r
s
i
t
é
A
mar
Te
l
i
d
j
i
d
e
La
g
h
o
u
a
t
,
La
g
h
o
u
a
t
,
A
l
g
e
r
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
1
,
2
0
2
6
R
ev
is
ed
Ap
r
2
9
,
2
0
2
6
Acc
ep
ted
J
u
l 2
2
,
2
0
2
6
Ro
a
d
a
c
c
id
e
n
ts
c
a
u
se
d
b
y
d
ri
v
e
r
d
ro
ws
in
e
ss
a
n
d
d
istrac
ti
o
n
a
re
a
m
a
jo
r
g
lo
b
a
l
c
o
n
c
e
rn
,
a
s
fa
ti
g
u
e
a
n
d
i
n
a
tt
e
n
ti
o
n
sig
n
ifi
c
a
n
tl
y
slo
w
re
a
c
ti
o
n
ti
m
e
a
n
d
in
c
re
a
se
a
c
c
id
e
n
t
risk
.
To
a
d
d
re
ss
th
is
issu
e
,
th
is
p
a
p
e
r
p
ro
p
o
se
s
a
v
isio
n
-
b
a
se
d
d
ri
v
e
r
m
o
n
it
o
ri
n
g
sy
ste
m
u
sin
g
fa
c
ial
a
n
a
l
y
sis
fro
m
a
n
o
n
-
b
o
a
rd
c
a
m
e
ra
.
Th
e
sy
ste
m
d
e
te
c
ts
th
e
fa
c
e
a
n
d
e
x
trac
ts
k
e
y
r
e
g
io
n
s
o
f
in
tere
st,
in
c
l
u
d
i
n
g
th
e
e
y
e
s,
m
o
u
th
,
a
n
d
h
e
a
d
,
wh
ich
a
re
a
n
a
ly
z
e
d
in
d
e
p
e
n
d
e
n
t
ly
u
sin
g
c
o
n
v
o
lu
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
s.
Tem
p
o
ra
l
i
n
fo
r
m
a
ti
o
n
is
c
a
p
tu
re
d
b
y
a
g
g
re
g
a
ti
n
g
th
e
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
(CNN
)
o
u
t
p
u
ts
o
v
e
r
a
fi
x
e
d
ti
m
e
win
d
o
w.
Dr
o
ws
in
e
ss
is
e
stim
a
ted
b
y
f
u
sin
g
e
y
e
a
n
d
m
o
u
t
h
fe
a
tu
re
s
with
a
m
u
lt
il
a
y
e
r
p
e
rc
e
p
tro
n
,
wh
i
le
d
istrac
ti
o
n
is
d
e
tec
ted
b
a
se
d
o
n
h
e
a
d
m
o
v
e
m
e
n
ts.
A
n
imp
o
rtan
t
a
d
v
a
n
tag
e
o
f
t
h
e
p
r
o
p
o
se
d
a
p
p
ro
a
c
h
is
it
s
ro
b
u
stn
e
ss
to
p
a
rti
a
l
in
p
u
t
lo
ss
,
a
ll
o
win
g
t
h
e
sy
ste
m
to
re
m
a
in
f
u
n
c
ti
o
n
a
l
e
v
e
n
wh
e
n
s
o
m
e
fa
c
ial
re
g
io
n
s
a
re
m
issin
g
o
r
o
c
c
lu
d
e
d
,
su
c
h
a
s
wh
e
n
we
a
rin
g
su
n
g
las
se
s
o
r
fa
c
e
m
a
sk
s
.
Ex
p
e
rime
n
tal
re
su
l
ts
sh
o
w
h
ig
h
d
e
tec
ti
o
n
a
c
c
u
ra
c
y
,
re
a
c
h
in
g
9
7
.
3
%
fo
r
d
r
o
ws
in
e
ss
a
n
d
9
8
%
fo
r
d
istrac
ti
o
n
u
n
d
e
r
id
e
a
l
c
o
n
d
i
ti
o
n
s,
with
o
n
l
y
li
m
it
e
d
p
e
rfo
rm
a
n
c
e
d
e
g
ra
d
a
ti
o
n
i
n
c
h
a
ll
e
n
g
in
g
sc
e
n
a
rio
s.
Th
e
se
re
su
lt
s
c
o
n
firm
th
e
su
it
a
b
il
it
y
o
f
t
h
e
p
r
o
p
o
se
d
s
y
ste
m
fo
r
re
a
l
-
ti
m
e
d
riv
e
r
m
o
n
it
o
ri
n
g
a
p
p
l
ica
ti
o
n
s.
K
ey
w
o
r
d
s
:
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
Dis
tr
ac
tio
n
d
etec
tio
n
Dr
iv
er
m
o
n
it
o
r
in
g
Dr
o
wsi
n
ess
d
etec
tio
n
F
ac
ial
an
aly
s
is
Mu
ltil
ay
er
p
er
ce
p
tr
o
n
R
ea
l
-
tim
e
s
y
s
tem
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
:
Sar
a
B
en
k
o
u
id
er
L
ab
o
r
ato
ir
e
d
'
I
n
f
o
r
m
atiq
u
e
et
d
e
Ma
th
ém
atiq
u
es,
Un
iv
e
r
s
ité
Am
ar
T
elid
ji d
e
L
ag
h
o
u
at
M’
k
am
,
L
ag
h
o
u
at,
Alg
er
ia
E
m
ail:
s
.
b
en
k
o
u
i
d
er
@
lag
h
-
u
n
iv
.
d
z
1.
I
NT
RO
D
UCT
I
O
N
R
o
ad
tr
af
f
ic
ac
cid
en
ts
ar
e
a
m
ajo
r
g
lo
b
al
p
r
o
b
lem
.
Acc
o
r
d
in
g
to
th
e
W
o
r
ld
Hea
lth
Or
g
an
izatio
n
(
W
HO)
,
m
o
r
e
th
an
1
.
1
9
m
illi
o
n
p
e
o
p
le
d
ie
ev
er
y
y
ea
r
d
u
e
to
r
o
ad
tr
af
f
ic
ac
ci
d
en
ts
.
Ma
n
y
o
f
t
h
ese
ac
cid
en
ts
ar
e
ca
u
s
ed
b
y
h
u
m
an
f
ac
to
r
s
,
esp
ec
ially
d
r
iv
er
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
[
1
]
.
Dr
o
wsi
n
ess
r
ed
u
ce
s
atten
tio
n
,
s
lo
ws
r
ea
ctio
n
tim
e,
an
d
a
f
f
ec
ts
d
ec
is
io
n
-
m
ak
in
g
,
wh
ich
in
cr
ea
s
es
th
e
r
is
k
o
f
s
er
io
u
s
ac
cid
en
ts
[
2
]
,
[
3
]
.
Dis
tr
ac
tio
n
also
in
cr
ea
s
es
ac
ci
d
en
t
r
is
k
b
y
s
h
if
tin
g
th
e
d
r
iv
er
’
s
atten
tio
n
awa
y
f
r
o
m
th
e
r
o
a
d
an
d
d
r
iv
in
g
task
s
[
4
]
,
[
5
]
.
Fo
r
th
is
r
ea
s
o
n
,
d
etec
tin
g
d
r
iv
er
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
h
as
b
ec
o
m
e
an
im
p
o
r
tan
t
g
o
al
o
f
ad
v
an
ce
d
d
r
iv
e
r
ass
is
tan
ce
s
y
s
tem
s
(
ADAS)
.
T
h
ese
s
y
s
tem
s
aim
to
im
p
r
o
v
e
r
o
ad
s
af
ety
b
y
m
o
n
ito
r
in
g
t
h
e
d
r
iv
er
’
s
s
tate,
esp
ec
ially
with
th
e
r
ec
en
t u
s
e
o
f
d
ee
p
lear
n
in
g
m
eth
o
d
s
[
6
]
.
I
n
r
ec
e
n
t
y
ea
r
s
,
m
a
n
y
s
tu
d
ies
h
av
e
u
s
ed
d
ee
p
n
eu
r
al
n
etwo
r
k
s
to
d
etec
t
d
r
i
v
er
d
r
o
wsi
n
ess
[
7
]
–
[
1
1
]
an
d
d
is
tr
ac
tio
n
[
1
2
]
–
[
1
6
]
f
r
o
m
im
ag
es
ca
p
tu
r
ed
b
y
in
-
v
e
h
icle
ca
m
er
as.
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NNs)
ar
e
co
m
m
o
n
ly
u
s
ed
to
an
aly
ze
f
ac
ial
cu
es
s
u
ch
as
ey
e
clo
s
u
r
e,
b
lin
k
r
ate
[
1
7
]
,
y
awn
in
g
[
1
8
]
,
an
d
h
ea
d
m
o
v
e
m
en
ts
[
1
4
]
.
T
o
in
cl
u
d
e
tim
e
-
r
elate
d
in
f
o
r
m
atio
n
,
s
o
m
e
wo
r
k
s
co
m
b
in
e
C
NNs with
r
ec
u
r
r
en
t n
eu
r
al
n
etwo
r
k
s
[
1
9
]
–
[
2
1
]
s
u
ch
as
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
ST
M
)
[
2
2
]
–
[
2
4
]
.
T
h
ese
m
o
d
els
ca
n
lear
n
tem
p
o
r
al
p
atter
n
s
lin
k
ed
to
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
a
n
d
h
a
v
e
s
h
o
w
n
g
o
o
d
r
esu
lts
o
n
r
ea
l d
r
iv
i
n
g
d
ata
[
6
]
,
[
1
7
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
co
n
vo
l
u
tio
n
a
l
n
eu
r
a
l n
etw
o
r
k
-
b
a
s
ed
d
r
iver m
o
n
ito
r
in
g
s
y
s
tem
fo
r
…
(
S
a
r
a
B
en
ko
u
i
d
er
)
2221
Ho
wev
er
,
th
ese
ap
p
r
o
ac
h
es
h
av
e
s
ev
er
al
lim
itatio
n
s
.
Mo
d
e
ls
b
ased
o
n
co
m
p
le
x
tem
p
o
r
a
l
n
etwo
r
k
s
o
r
3
D
co
n
v
o
lu
tio
n
ar
e
co
m
p
u
tatio
n
ally
ex
p
en
s
iv
e
a
n
d
r
eq
u
ir
e
s
ig
n
if
ican
t
m
em
o
r
y
r
eso
u
r
ce
s
,
wh
ich
m
ak
es
th
eir
d
ep
l
o
y
m
e
n
t
o
n
r
ea
l
-
tim
e
em
b
ed
d
ed
s
y
s
tem
s
d
if
f
icu
lt
[
2
5
]
.
I
n
ad
d
itio
n
,
m
a
n
y
ex
is
tin
g
m
eth
o
d
s
ass
u
m
e
th
e
av
ailab
ilit
y
o
f
all
v
is
u
al
in
p
u
ts
an
d
a
r
e
th
er
e
f
o
r
e
h
ig
h
l
y
s
en
s
itiv
e
to
ch
an
g
es
in
lig
h
tin
g
co
n
d
itio
n
s
an
d
p
ar
tial
f
ac
ial
o
cc
lu
s
io
n
s
,
s
u
ch
as
s
u
n
g
lass
es
o
r
f
ac
e
m
as
k
s
.
W
h
en
s
o
m
e
f
ac
ial
in
f
o
r
m
atio
n
is
m
is
s
in
g
o
r
u
n
r
eliab
le,
th
e
p
er
f
o
r
m
an
ce
o
f
th
ese
m
et
h
o
d
s
o
f
ten
d
e
g
r
ad
es
s
ig
n
if
ican
tly
o
r
t
h
e
s
y
s
tem
m
ay
f
ail
t
o
o
p
er
ate
p
r
o
p
er
l
y
.
T
h
ese
lim
itatio
n
s
h
i
g
h
lig
h
t
th
e
n
ee
d
f
o
r
s
im
p
le,
e
f
f
icien
t,
an
d
r
o
b
u
s
t
ap
p
r
o
ac
h
e
s
th
at
ca
n
o
p
er
ate
r
eliab
ly
in
r
ea
l
d
r
iv
i
n
g
c
o
n
d
itio
n
s
,
ev
en
wh
e
n
s
o
m
e
i
n
p
u
t
in
f
o
r
m
atio
n
is
m
is
s
in
g
,
w
h
ile
s
atis
f
y
in
g
th
e
co
n
s
tr
ain
ts
o
f
em
b
e
d
d
ed
s
y
s
tem
s
.
I
n
t
h
i
s
p
a
p
e
r
,
w
e
p
r
o
p
o
s
e
a
m
o
d
u
l
a
r
v
i
s
i
o
n
-
b
a
s
e
d
s
y
s
te
m
u
s
i
n
g
s
e
p
a
r
a
t
e
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
o
r
t
h
e
e
y
e
s
,
m
o
u
t
h
,
a
n
d
h
e
a
d
r
e
g
i
o
n
s
.
A
li
g
h
t
w
e
i
g
h
t
t
e
m
p
o
r
a
l
a
g
g
r
e
g
a
t
i
o
n
m
e
t
h
o
d
a
n
d
a
f
u
s
i
o
n
s
t
r
at
e
g
y
a
r
e
e
m
p
l
o
y
e
d
t
o
c
o
m
b
i
n
e
i
n
f
o
r
m
a
t
i
o
n
f
r
o
m
t
h
e
s
e
r
e
g
i
o
n
s
.
A
n
i
m
p
o
r
t
a
n
t
a
d
v
a
n
t
a
g
e
o
f
t
h
e
p
r
o
p
o
s
e
d
a
r
c
h
i
t
e
ct
u
r
e
l
i
es
i
n
i
t
s
r
o
b
u
s
t
n
es
s
t
o
p
a
r
t
i
a
l
i
n
f
o
r
m
a
t
i
o
n
l
o
s
s
.
S
i
n
c
e
e
ac
h
f
a
c
i
a
l
r
e
g
i
o
n
i
s
a
n
a
l
y
z
e
d
i
n
d
e
p
e
n
d
e
n
t
l
y
,
t
h
e
s
y
s
t
e
m
c
a
n
c
o
n
t
i
n
u
e
t
o
o
p
e
r
a
t
e
e
v
e
n
w
h
e
n
o
n
e
o
r
m
o
r
e
r
e
g
i
o
n
s
o
f
i
n
t
e
r
e
s
t
(
R
O
I
s
)
a
r
e
m
is
s
i
n
g
o
r
p
a
r
t
i
a
lly
o
c
c
l
u
d
e
d
,
s
u
c
h
as
i
n
t
h
e
p
r
e
s
e
n
c
e
o
f
s
u
n
g
l
a
s
s
es
o
r
f
a
c
e
m
a
s
k
s
.
T
h
e
p
r
o
p
o
s
e
d
s
y
s
t
e
m
a
i
m
s
t
o
a
c
h
i
e
v
e
r
e
l
i
a
b
l
e
d
e
t
e
ct
i
o
n
p
e
r
f
o
r
m
a
n
c
e
w
h
i
l
e
r
e
m
a
i
n
i
n
g
c
o
m
p
u
t
a
t
i
o
n
a
l
l
y
e
f
f
i
ci
e
n
t
a
n
d
s
u
i
t
a
b
l
e
f
o
r
r
e
a
l
-
t
i
m
e
e
m
b
e
d
d
e
d
ap
p
l
i
c
a
t
i
o
n
s
.
T
h
e
m
a
i
n
c
o
n
t
r
i
b
u
t
i
o
n
s
o
f
t
h
is
p
a
p
e
r
ca
n
b
e
s
u
m
m
a
r
i
z
e
d
a
s
f
o
l
l
o
ws
:
i
)
a
m
o
d
u
l
a
r
m
u
l
ti
-
R
O
I
C
N
N
-
b
a
s
e
d
a
r
c
h
i
t
e
c
t
u
r
e
f
o
r
s
i
m
u
l
ta
n
e
o
u
s
d
r
i
v
e
r
d
r
o
w
s
i
n
e
s
s
a
n
d
d
i
s
t
r
a
c
t
i
o
n
d
e
t
e
c
ti
o
n
,
w
h
e
r
e
e
a
c
h
f
a
c
i
a
l
r
e
g
i
o
n
(
e
y
e
s
,
m
o
u
t
h
,
a
n
d
h
e
a
d
)
i
s
p
r
o
c
e
s
s
e
d
i
n
d
e
p
e
n
d
e
n
t
l
y
,
i
i
)
a
l
i
g
h
t
w
ei
g
h
t
t
e
m
p
o
r
a
l
a
g
g
r
e
g
a
t
i
o
n
s
t
r
a
t
e
g
y
t
h
a
t
c
a
p
t
u
r
e
s
d
y
n
a
m
i
c
d
r
i
v
e
r
b
e
h
a
v
i
o
r
s
w
i
t
h
o
u
t
r
e
l
y
i
n
g
o
n
c
o
m
p
u
t
a
t
i
o
n
a
ll
y
e
x
p
e
n
s
i
v
e
r
e
c
u
r
r
e
n
t
o
r
3
D
c
o
n
v
o
l
u
t
i
o
n
a
l
m
o
d
e
l
s
,
i
i
i
)
a
r
o
b
u
s
t
f
u
s
i
o
n
m
e
c
h
a
n
is
m
t
h
a
t
c
o
m
b
i
n
e
s
c
o
m
p
l
e
m
e
n
ta
r
y
i
n
f
o
r
m
a
t
i
o
n
f
r
o
m
d
i
f
f
e
r
e
n
t
f
a
c
i
a
l
r
e
g
i
o
n
s
t
o
i
m
p
r
o
v
e
d
e
t
e
c
t
i
o
n
r
el
i
a
b
il
i
t
y
,
i
v
)
a
n
o
c
c
l
u
s
i
o
n
-
r
e
s
i
l
i
e
n
t
d
es
i
g
n
t
h
a
t
e
n
a
b
l
es
t
h
e
s
y
s
t
e
m
t
o
m
a
i
n
t
ai
n
s
t
a
b
l
e
p
e
r
f
o
r
m
a
n
c
e
e
v
e
n
i
n
t
h
e
p
r
e
s
e
n
c
e
o
f
s
u
n
g
l
a
s
s
e
s
,
f
a
c
e
m
a
s
k
s
,
o
r
p
a
r
t
i
a
l
f
a
c
i
al
o
c
c
l
u
s
i
o
n
s
,
a
n
d
v
)
a
c
o
m
p
u
t
a
t
i
o
n
a
l
l
y
e
f
f
i
ci
e
n
t
f
r
a
m
e
w
o
r
k
s
u
i
t
a
b
l
e
f
o
r
r
e
a
l
-
t
i
m
e
d
e
p
l
o
y
m
e
n
t
o
n
e
m
b
e
d
d
e
d
s
y
s
t
em
s
w
h
i
l
e
ac
h
i
e
v
i
n
g
c
o
m
p
e
t
i
t
i
v
e
p
e
r
f
o
r
m
a
n
c
e
.
T
h
e
r
e
s
t
o
f
t
h
i
s
p
a
p
e
r
i
s
o
r
g
a
n
i
ze
d
a
s
f
o
l
l
o
w
s
.
Se
c
t
i
o
n
2
p
r
e
s
e
n
t
s
t
h
e
p
r
o
p
o
s
e
d
m
e
t
h
o
d
.
S
e
c
t
i
o
n
3
r
e
p
o
r
ts
t
h
e
ex
p
e
r
i
m
e
n
t
a
l
r
es
u
l
ts
.
S
e
ct
i
o
n
4
co
n
c
l
u
d
e
s
t
h
e
p
a
p
e
r
a
n
d
d
i
s
c
u
s
s
e
s
f
u
t
u
r
e
w
o
r
k
.
2.
O
UR
AP
P
RO
ACH
Ou
r
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
aim
s
to
d
etec
t
d
r
iv
er
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
v
ia
f
ac
ial
a
n
aly
s
is
u
s
in
g
a
v
id
eo
s
tr
ea
m
ca
p
tu
r
e
d
b
y
an
on
-
b
o
ar
d
ca
m
er
a
.
I
t
r
elies
o
n
a
m
u
ltimo
d
al
C
NN
-
b
ased
ar
ch
itectu
r
e
th
at
ex
p
lo
its
co
m
p
lem
en
tar
y
in
f
o
r
m
atio
n
f
r
o
m
th
e
ey
es,
m
o
u
th
,
an
d
h
ea
d
m
o
v
em
e
n
ts
.
E
ac
h
in
f
o
r
m
atio
n
s
o
u
r
ce
is
in
d
ep
en
d
en
tly
an
aly
ze
d
b
y
a
s
p
ec
ialized
n
etwo
r
k
,
an
d
th
e
e
x
tr
ac
ted
in
f
o
r
m
atio
n
is
th
en
f
u
s
ed
to
esti
m
ate
th
e
d
r
iv
er
’
s
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
s
tates.
T
h
is
m
o
d
u
lar
d
esig
n
im
p
r
o
v
es
r
o
b
u
s
tn
ess
,
as
th
e
s
y
s
tem
ca
n
co
n
tin
u
e
to
o
p
er
ate
ev
e
n
wh
e
n
o
n
e
o
r
m
o
r
e
f
ac
ial
in
p
u
ts
ar
e
m
is
s
in
g
o
r
p
ar
tially
o
cc
lu
d
ed
,
wh
ich
is
co
m
m
o
n
in
r
ea
l
d
r
iv
i
n
g
co
n
d
itio
n
s
.
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
is
s
tr
u
ctu
r
e
d
in
to
th
r
ee
m
ain
m
o
d
u
les:
i)
Face
d
etec
tio
n
an
d
R
OI
s
ex
tr
ac
tio
n
m
o
d
u
le,
ii)
C
NN
-
m
u
ltil
ay
er
p
er
ce
p
tr
o
n
(
M
L
P)
-
b
ased
d
r
o
wsi
n
ess
d
etec
tio
n
m
o
d
u
le
,
an
d
iii)
C
NN
-
b
ased
d
is
tr
ac
tio
n
d
etec
tio
n
m
o
d
u
le
as sh
o
wn
in
Fig
u
r
e
1
.
Fig
u
r
e
1
.
Ar
c
h
itectu
r
e
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
4
,
Au
g
u
s
t
20
2
6
:
2
2
2
0
-
2229
2222
2
.
1
.
F
a
ce
d
et
ec
t
io
n a
nd
RO
I
s
ex
t
ra
ct
io
n mo
du
le
T
h
is
m
o
d
u
le
co
n
s
titu
tes
th
e
f
i
r
s
t
s
tep
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
.
I
ts
o
b
jectiv
e
is
t
o
a
u
to
m
ati
ca
lly
lo
ca
te
th
e
d
r
iv
er
'
s
f
ac
e
f
r
o
m
th
e
v
id
e
o
s
tr
ea
m
ac
q
u
ir
ed
b
y
an
o
n
-
b
o
ar
d
ca
m
er
a,
th
e
n
to
ex
tr
ac
t
th
e
r
eg
io
n
s
o
f
in
te
r
est
(
R
OI
s
)
n
ec
ess
ar
y
f
o
r
th
e
d
etec
tio
n
o
f
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
t
io
n
.
2
.
1
.
1
.
F
a
ce
d
et
ec
t
i
o
n
Fo
r
ea
ch
v
id
eo
f
r
a
m
e
,
f
ac
e
d
etec
tio
n
is
p
er
f
o
r
m
ed
to
lo
ca
lize
th
e
d
r
iv
er
’
s
f
ac
e
th
r
o
u
g
h
a
b
o
u
n
d
in
g
b
o
x
d
e
f
in
ed
as:
=
ℱ
(
)
=
(
,
,
,
)
W
h
er
e
(
,
)
d
en
o
te
th
e
co
o
r
d
in
ates o
f
th
e
to
p
-
lef
t c
o
r
n
er
,
an
d
(
,
)
co
r
r
esp
o
n
d
to
th
e
wid
th
a
n
d
h
eig
h
t
o
f
th
e
d
etec
ted
f
ac
e,
r
esp
ec
tiv
ely
.
I
n
o
u
r
ap
p
r
o
ac
h
,
a
YOL
O
-
b
a
s
ed
f
ac
e
d
etec
to
r
is
em
p
lo
y
e
d
d
u
e
to
its
ab
ilit
y
to
ac
h
iev
e
r
ea
l
-
tim
e
p
er
f
o
r
m
an
ce
wh
ile
m
ain
tain
i
n
g
h
ig
h
d
etec
tio
n
ac
cu
r
ac
y
u
n
d
er
ch
allen
g
in
g
co
n
d
itio
n
s
s
u
ch
as
illu
m
in
atio
n
v
ar
iatio
n
s
,
o
cc
lu
s
io
n
s
,
an
d
h
ea
d
p
o
s
e
ch
a
n
g
es.
Un
lik
e
tr
ad
itio
n
al
m
u
lti
-
s
tag
e
d
etec
to
r
s
,
YOL
O
p
er
f
o
r
m
s
d
etec
tio
n
in
a
s
in
g
le
f
o
r
war
d
p
ass
b
y
d
ir
ec
tl
y
r
e
g
r
ess
in
g
b
o
u
n
d
in
g
b
o
x
c
o
o
r
d
in
ates
an
d
ass
o
ciate
d
co
n
f
i
d
en
ce
s
co
r
es
o
v
er
th
e
e
n
tire
im
ag
e.
T
o
en
s
u
r
e
r
eliab
le
l
o
ca
lizatio
n
,
r
ed
u
n
d
a
n
t
d
etec
tio
n
s
ar
e
eli
m
in
ated
u
s
in
g
n
o
n
-
m
ax
im
u
m
s
u
p
p
r
ess
io
n
(
NM
S),
wh
ich
r
etain
s
th
e
b
o
u
n
d
in
g
b
o
x
with
t
h
e
h
ig
h
est
co
n
f
id
en
ce
s
co
r
e
wh
ile
s
u
p
p
r
ess
in
g
o
v
e
r
lap
p
in
g
p
r
e
d
ictio
n
s
.
2
.
1
.
2
.
Reg
io
ns
o
f
inte
re
s
t
(
R
O
I
s
)
ex
t
ra
ct
io
n
Sin
ce
YOL
O
d
o
es
n
o
t
in
h
e
r
en
tly
p
r
o
v
id
e
f
i
n
e
-
g
r
ai
n
ed
f
a
cial
s
tr
u
ctu
r
e
i
n
f
o
r
m
atio
n
,
a
d
ed
icate
d
lan
d
m
ar
k
d
etec
tio
n
m
o
d
el
is
in
co
r
p
o
r
ated
to
esti
m
ate
k
ey
f
a
cial
p
o
in
ts
:
=
ℱ
(
,
)
wh
er
e
r
ep
r
esen
ts
th
e
s
et
o
f
f
ac
ial
lan
d
m
ar
k
s
co
r
r
esp
o
n
d
in
g
to
cr
itical
r
eg
io
n
s
s
u
ch
as
th
e
ey
es
an
d
m
o
u
th
.
A
d
en
s
e
lan
d
m
ar
k
d
etec
to
r
(
Me
d
iaPip
e
Face
Me
s
h
)
is
ad
o
p
ted
to
e
n
h
an
ce
s
p
atial
p
r
ec
is
io
n
an
d
r
o
b
u
s
tn
ess
ag
ain
s
t h
ea
d
m
o
v
em
en
ts
.
B
ased
o
n
th
e
d
etec
ted
f
ac
e
a
n
d
th
e
esti
m
ated
lan
d
m
ar
k
s
,
th
r
ee
s
em
an
tically
m
ea
n
in
g
f
u
l
R
OI
s
ar
e
ex
tr
ac
ted
:
−
E
y
e
R
OI
: Cap
tu
r
es th
e
ey
e
r
eg
io
n
f
o
r
an
aly
zin
g
ey
elid
cl
o
s
u
r
e
an
d
b
lin
k
in
g
p
atter
n
s
.
−
Mo
u
th
R
OI
: Cap
tu
r
es th
e
m
o
u
th
ar
ea
,
m
ain
ly
u
s
ed
f
o
r
d
etec
t
in
g
y
awn
in
g
.
−
Hea
d
R
OI
:
in
clu
d
es
t
h
e
f
u
ll
h
ea
d
r
eg
io
n
to
esti
m
ate
h
ea
d
o
r
ie
n
tatio
n
a
n
d
m
o
v
e
m
en
ts
r
elate
d
to
d
is
tr
ac
tio
n
.
T
h
e
ex
tr
ac
tio
n
o
f
a
r
eg
io
n
o
f
i
n
ter
est (
R
OI
)
at
tim
e
f
o
r
a
m
o
d
ality
(
ey
es,
m
o
u
th
o
r
h
ea
d
)
i
s
d
ef
in
ed
b
y
:
=
(
:
+
,
:
+
)
,
∈
{
,
ℎ
,
ℎ
}
2
.
1
.
3
.
Sp
a
t
ia
l
no
rma
liza
t
io
n
T
o
g
u
ar
a
n
tee
a
co
n
s
is
ten
t
in
p
u
t
r
ep
r
esen
tatio
n
f
o
r
s
u
b
s
eq
u
en
t
C
NN
-
b
ased
p
r
o
ce
s
s
in
g
,
e
ac
h
R
OI
is
r
esized
to
a
p
r
e
d
ef
in
ed
r
eso
lu
t
io
n
u
s
in
g
b
ilin
ea
r
in
ter
p
o
latio
n
:
=
b
i
l
i
n
ea
r
(
,
,
)
,
∈
{
,
ℎ
,
ℎ
}
T
h
e
ad
o
p
ted
r
eso
lu
tio
n
s
ar
e
6
4
×1
2
8
f
o
r
ey
es,
6
4
×
6
4
f
o
r
th
e
m
o
u
th
,
an
d
1
2
8
×
1
2
8
f
o
r
th
e
h
ea
d
.
B
ilin
ea
r
in
ter
p
o
latio
n
is
s
elec
ted
as
it
p
r
o
v
id
es
an
ef
f
ec
tiv
e
co
m
p
r
o
m
is
e
b
etwe
en
co
m
p
u
tatio
n
al
ef
f
icien
cy
an
d
p
r
eser
v
atio
n
o
f
s
p
atial
s
tr
u
ctu
r
es.
2
.
1
.
4
.
I
nte
ns
it
y
no
rm
a
liza
t
io
n
:
T
o
r
e
d
u
ce
th
e
i
n
f
lu
en
ce
o
f
li
g
h
tin
g
v
ar
iatio
n
s
an
d
im
p
r
o
v
e
s
y
s
tem
r
o
b
u
s
tn
ess
,
th
e
p
ix
e
l
v
alu
es
o
f
ea
ch
R
OI
ar
e
n
o
r
m
alize
d
with
in
th
e
in
ter
v
al
[
0
,
1
]
:
=
255
,
∈
{
,
ℎ
,
ℎ
}
T
h
is
o
p
er
atio
n
e
n
s
u
r
es a
u
n
if
o
r
m
in
p
u
t
r
ep
r
esen
tatio
n
b
ef
o
r
e
p
r
o
ce
s
s
in
g
b
y
th
e
C
NN
m
o
d
e
ls
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
co
n
vo
l
u
tio
n
a
l
n
eu
r
a
l n
etw
o
r
k
-
b
a
s
ed
d
r
iver m
o
n
ito
r
in
g
s
y
s
tem
fo
r
…
(
S
a
r
a
B
en
ko
u
i
d
er
)
2223
2
.
2
.
CNN
–
MLP
-
ba
s
ed
dro
w
s
ines
s
det
ec
t
io
n m
o
du
le
T
h
is
m
o
d
u
le
aim
s
to
d
etec
t
d
r
iv
er
d
r
o
wsi
n
ess
th
r
o
u
g
h
th
e
co
m
b
in
e
d
an
aly
s
is
o
f
e
y
e
an
d
m
o
u
th
m
o
v
em
en
ts
o
v
er
a
tem
p
o
r
al
win
d
o
w
o
f
T
co
n
s
ec
u
tiv
e
f
r
a
m
es.
T
h
is
tem
p
o
r
al
f
o
r
m
u
latio
n
en
a
b
les
th
e
s
y
s
tem
to
ca
p
tu
r
e
th
e
d
y
n
am
ic
ev
o
lu
tio
n
o
f
d
r
o
wsi
n
ess
-
r
elate
d
p
atte
r
n
s
s
u
ch
as b
lin
k
in
g
an
d
y
awn
in
g
.
T
h
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
a
d
o
p
ts
a
co
m
p
u
tatio
n
ally
ef
f
icien
t
two
-
s
tag
e
s
tr
ateg
y
.
Fi
r
s
t,
C
NN
s
in
d
ep
en
d
en
tly
p
r
o
ce
s
s
ea
ch
f
r
am
e
to
ex
tr
ac
t
s
p
atial
f
ea
tu
r
es
f
r
o
m
f
ac
ial
R
OI
s
.
T
h
en
,
tem
p
o
r
al
in
f
o
r
m
atio
n
is
m
o
d
eled
b
y
a
ML
P,
w
h
ich
ag
g
r
eg
ates
th
e
s
eq
u
e
n
ce
o
f
C
NN
o
u
tp
u
ts
o
v
er
th
e
tem
p
o
r
al
win
d
o
w
as
s
h
o
wn
i
n
Fig
u
r
e
2
.
T
h
is
d
ec
o
u
p
led
d
esig
n
b
et
wee
n
s
p
atial
an
d
tem
p
o
r
al
m
o
d
elin
g
s
im
p
lifie
s
tr
ain
in
g
,
im
p
r
o
v
es
s
tab
ilit
y
,
an
d
s
ig
n
if
ican
tly
r
e
d
u
ce
s
co
m
p
u
tatio
n
al
co
m
p
lex
ity
co
m
p
a
r
ed
to
r
ec
u
r
r
e
n
t
o
r
3
D
co
n
v
o
lu
tio
n
a
l
ap
p
r
o
ac
h
es.
I
n
ad
d
itio
n
,
th
e
in
d
ep
en
d
en
t
p
r
o
ce
s
s
in
g
o
f
ea
ch
f
ac
ial
r
eg
io
n
en
h
an
ce
s
r
o
b
u
s
tn
ess
to
p
ar
tial
o
cc
lu
s
io
n
s
an
d
m
is
s
in
g
in
f
o
r
m
atio
n
.
Fig
u
r
e
2
.
C
NN
-
MLP
-
b
ased
d
r
o
wsi
n
ess
d
etec
tio
n
alg
o
r
ith
m
2
.
2
.
1
.
CNN
-
ba
s
ed
f
ea
t
ure
ex
t
ra
ct
io
n
Fo
r
ea
ch
m
o
d
ality
∈
{
,
ℎ
}
,
a
tem
p
o
r
al
s
eq
u
en
ce
o
f
R
OI
s
is
co
n
s
id
er
ed
:
=
{
1
,
2
,
…
…
.
.
,
}
E
ac
h
R
OI
is
p
r
o
ce
s
s
ed
in
d
ep
en
d
en
tly
b
y
a
d
ed
icate
d
C
NN
b
r
an
ch
.
T
h
e
C
NN
ar
ch
itectu
r
e
co
n
s
is
ts
o
f
:
i)
C
o
n
v
2
D
(
3
×3
,
3
2
f
ilter
s
)
+
R
eL
U
,
ii)
Ma
x
Po
o
lin
g
(
2
×2
)
,
i
ii)
C
o
n
v
2
D
(
3
×
3
,
6
4
f
ilter
s
)
+
R
eL
U
,
iv
)
Glo
b
al
Av
er
ag
e
Po
o
lin
g
,
v
)
Fu
lly
C
o
n
n
ec
ted
lay
er
,
an
d
v
i)
Sig
m
o
id
ac
tiv
atio
n
.
Fo
r
ea
ch
f
r
a
m
e
,
th
e
C
NN
o
u
tp
u
ts
a
co
n
tin
u
o
u
s
co
n
f
i
d
en
ce
s
c
o
r
e:
=
(
)
,
∈
[
0
,
1
]
Af
ter
p
r
o
ce
s
s
in
g
th
e
f
u
ll tem
p
o
r
al
win
d
o
w,
a
tem
p
o
r
al
s
co
r
e
v
ec
to
r
is
o
b
tain
e
d
:
=
[
1
,
2
,
…
…
.
.
,
]
2
.
2
.
2
.
T
em
po
ra
l
f
us
io
n a
nd
dro
wsi
nes
s
estim
a
t
io
n
T
h
e
tem
p
o
r
al
s
co
r
e
v
ec
to
r
s
ex
tr
ac
ted
f
r
o
m
th
e
e
y
e
an
d
m
o
u
th
C
NNs
ar
e
co
n
ca
ten
ated
t
o
f
o
r
m
th
e
in
p
u
t o
f
th
e
ML
P:
=
[
,
ℎ
]
∈
ℝ
2
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
4
,
Au
g
u
s
t
20
2
6
:
2
2
2
0
-
2229
2224
T
h
e
ML
P
lear
n
s
ab
o
u
t
tem
p
o
r
al
d
ep
en
d
e
n
cies
an
d
in
ter
m
o
d
al
r
elatio
n
s
h
ip
s
b
etwe
en
ey
e
a
n
d
m
o
u
th
b
eh
av
io
r
s
.
I
t p
r
o
d
u
ce
s
a
g
l
o
b
a
l d
r
o
wsi
n
ess
co
n
f
id
en
ce
s
co
r
e:
=
(
2
∙
(
1
+
1
)
+
2
)
w
h
er
e
1
a
n
d
2
ar
e
weig
h
t
m
atr
i
ce
s
,
1
,
2
ar
e
b
ias
ter
m
s
a
n
d
t
h
e
o
u
tp
u
t
∈
[
0
,
1
]
r
ep
r
esen
ts
th
e
p
r
o
b
a
b
ilit
y
o
f
d
r
o
wsi
n
ess
o
v
er
th
e
tem
p
o
r
al
win
d
o
w.
2
.
2
.
3
.
Dec
is
io
n
s
t
ra
t
eg
y
T
h
e
f
in
al
d
ec
is
io
n
is
o
b
tain
ed
b
y
th
r
esh
o
l
d
in
g
t
h
e
ML
P o
u
tp
u
t:
=
(
1
,
≥
0
,
ℎ
)
w
h
er
e
i
s
a
p
r
ed
ef
in
ed
th
r
esh
o
ld
(
ty
p
ically
=
0
.
5
).
T
h
is
d
elay
ed
d
ec
is
io
n
s
tr
ateg
y
im
p
r
o
v
es
r
o
b
u
s
tn
ess
an
d
av
o
id
s
f
r
am
e
-
le
v
el
n
o
is
e.
2
.
3
.
CNN
-
ba
s
ed
dis
t
ra
ct
io
n det
ec
t
io
n m
o
du
le
T
h
e
p
r
o
p
o
s
ed
C
NN
-
b
ased
d
is
tr
ac
tio
n
d
etec
tio
n
m
o
d
u
le
aim
s
to
id
en
tif
y
d
r
iv
er
d
is
tr
ac
tio
n
b
y
an
aly
zin
g
h
ea
d
p
o
s
e
an
d
m
o
tio
n
p
atter
n
s
o
v
er
a
tem
p
o
r
al
win
d
o
w
o
f
T
co
n
s
ec
u
tiv
e
f
r
am
es.
Un
lik
e
d
r
o
wsi
n
ess
d
etec
tio
n
,
wh
ich
r
elies
o
n
m
u
ltip
le
f
ac
ial
cu
es,
d
is
tr
ac
tio
n
d
etec
tio
n
p
r
im
ar
il
y
d
ep
e
n
d
s
o
n
h
ea
d
o
r
ien
tatio
n
d
y
n
am
ics,
wh
ich
a
r
e
s
tr
o
n
g
in
d
icato
r
s
o
f
v
is
u
al
a
n
d
co
g
n
itiv
e
d
is
tr
ac
tio
n
.
2
.
3
.
1
.
I
np
ut
da
t
a
a
nd
t
em
po
r
a
l w
ind
o
w
A
s
eq
u
en
ce
o
f
h
ea
d
R
eg
io
n
s
o
f
I
n
ter
est (
R
OI
s
)
is
ex
tr
ac
ted
o
v
er
a
tem
p
o
r
al
win
d
o
w:
ℎ
=
{
ℎ
1
,
ℎ
2
,
…
…
.
.
,
ℎ
}
2
.
3
.
2
.
CNN
-
ba
s
ed
f
ea
t
ure
ex
t
ra
ct
io
n
E
ac
h
ℎ
is
p
r
o
ce
s
s
ed
in
d
ep
en
d
e
n
tly
b
y
a
d
ed
icate
d
C
N
N
ℎ
b
r
an
ch
d
esig
n
ed
to
ex
tr
ac
t
d
is
cr
im
in
ativ
e
s
p
atial
f
ea
tu
r
es
r
elate
d
to
h
ea
d
o
r
ien
tatio
n
an
d
m
o
v
em
en
t
as
s
h
o
wn
in
Fig
u
r
e
3
.
T
h
e
C
N
N
ℎ
ar
ch
itectu
r
e
f
o
llo
ws
th
e
s
am
e
d
esig
n
p
r
in
cip
les
as
t
h
e
o
th
er
m
o
d
ality
-
s
p
ec
if
ic
n
etwo
r
k
s
(
C
N
N
C
N
N
)
.
Fo
r
ea
c
h
f
r
am
e
,
th
e
C
NN
p
r
o
d
u
ce
s
a
d
is
tr
ac
tio
n
co
n
f
i
d
en
ce
s
co
r
e:
ℎ
=
ℎ
(
ℎ
)
,
ℎ
∈
[
0
,
1
]
T
h
is
s
co
r
e
r
ep
r
esen
ts
th
e
p
r
o
b
ab
ilit
y
o
f
d
is
tr
ac
tio
n
at
tim
e
.
Fig
u
r
e
3
.
C
NN
-
b
ased
d
is
tr
ac
tio
n
d
etec
tio
n
alg
o
r
ith
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
co
n
vo
l
u
tio
n
a
l
n
eu
r
a
l n
etw
o
r
k
-
b
a
s
ed
d
r
iver m
o
n
ito
r
in
g
s
y
s
tem
fo
r
…
(
S
a
r
a
B
en
ko
u
i
d
er
)
2225
2
.
3
.
3
.
T
em
po
ra
l
a
g
g
re
g
a
t
io
n
T
o
in
co
r
p
o
r
ate
tem
p
o
r
al
in
f
o
r
m
atio
n
,
t
h
e
f
r
am
e
-
lev
el
s
co
r
es
ar
e
ag
g
r
eg
ated
u
s
in
g
an
av
er
ag
in
g
s
tr
ateg
y
:
ℎ
=
∑
(
ℎ
)
=
T
h
is
ag
g
r
eg
atio
n
r
ed
u
ce
s
n
o
is
e
an
d
p
r
o
v
id
es a
s
tab
le
esti
m
atio
n
o
f
d
is
tr
ac
tio
n
o
v
e
r
th
e
tem
p
o
r
al
win
d
o
w.
2
.
3
.
4
.
Dec
is
io
n
r
ule
A
f
in
al
b
in
ar
y
d
is
tr
ac
tio
n
d
ec
i
s
io
n
is
o
b
tain
ed
b
y
ap
p
ly
in
g
a
th
r
esh
o
ld
to
th
e
ag
g
r
eg
ated
s
co
r
e:
=
(
1
,
ℎ
≥
0
,
ℎ
)
w
h
er
e,
1
in
d
icate
s
a
d
is
tr
ac
ted
s
tate,
0
in
d
icate
s
a
n
o
r
m
al
s
tate,
an
d
is
a
p
r
ed
ef
in
e
d
d
e
cisi
o
n
th
r
esh
o
ld
(
ty
p
ically
=
0
.
5
).
Un
lik
e
f
r
a
m
e
-
b
a
s
ed
d
ec
is
io
n
s
,
th
is
d
elay
ed
b
in
ar
iatio
n
im
p
r
o
v
es
r
o
b
u
s
tn
ess
b
y
lev
er
ag
in
g
tem
p
o
r
al
c
o
n
s
is
ten
cy
.
3.
SI
M
UL
A
T
I
O
N
R
E
S
UL
T
S
T
o
ev
alu
ate
th
e
p
er
f
o
r
m
a
n
c
e
o
f
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
,
m
u
ltip
le
ex
p
e
r
im
en
tal
s
ce
n
ar
io
s
wer
e
d
esig
n
ed
,
r
a
n
g
in
g
f
r
o
m
id
ea
l
co
n
d
itio
n
s
to
m
o
r
e
ch
allen
g
in
g
s
itu
atio
n
s
in
v
o
lv
in
g
f
ac
ial
o
cc
lu
s
io
n
s
,
s
u
ch
as
d
r
iv
er
s
wea
r
in
g
s
u
n
g
lass
es
o
r
s
u
r
g
ical
m
ask
s
.
T
h
e
e
x
p
er
im
en
tal
ev
alu
atio
n
is
co
n
d
u
cted
u
s
in
g
v
id
eo
s
eq
u
en
ce
s
ac
q
u
ir
e
d
u
n
d
er
d
if
f
er
en
t
c
o
n
d
itio
n
s
,
in
clu
d
in
g
n
o
r
m
al
d
r
iv
in
g
,
d
r
o
wsi
n
ess
,
d
is
tr
ac
tio
n
,
an
d
o
cc
lu
s
io
n
s
ce
n
ar
io
s
.
T
h
e
v
id
e
o
s
ar
e
s
eg
m
en
ted
in
to
tem
p
o
r
al
win
d
o
ws
o
f
len
g
th
T
,
r
an
g
i
n
g
f
r
o
m
5
0
to
2
0
0
0
f
r
am
es,
u
s
in
g
a
s
lid
in
g
win
d
o
w
s
tr
ateg
y
th
at
g
en
e
r
ates
a
lar
g
e
n
u
m
b
er
o
f
tem
p
o
r
al
s
am
p
les.
T
h
e
d
ataset
in
clu
d
es
r
ec
o
r
d
in
g
s
f
r
o
m
a
s
in
g
le
d
r
i
v
er
,
allo
win
g
a
c
o
n
tr
o
ll
ed
an
al
y
s
is
o
f
tem
p
o
r
al
m
o
d
el
in
g
an
d
r
o
b
u
s
tn
ess
to
o
cc
lu
s
io
n
s
.
A
n
n
o
tatio
n
s
ar
e
d
ef
in
e
d
at
t
h
e
tem
p
o
r
al
win
d
o
w
lev
el,
wh
er
e
d
r
o
wsi
n
ess
is
lab
eled
b
ased
o
n
ey
e
clo
s
u
r
e
an
d
y
awn
in
g
b
eh
av
io
r
,
a
n
d
d
is
tr
ac
tio
n
is
d
e
f
in
ed
b
ased
o
n
h
ea
d
p
o
s
e
d
ev
iat
io
n
.
T
h
e
s
im
u
latio
n
an
d
d
ataset
p
ar
am
eter
s
u
s
ed
i
n
all
ex
p
er
im
en
ts
ar
e
s
u
m
m
ar
i
ze
d
in
T
ab
le
1
.
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
is
ev
alu
ated
u
n
d
er
t
h
r
ee
d
is
tin
ct
test
co
n
d
it
io
n
s
in
o
r
d
e
r
to
e
v
alu
ate
its
r
o
b
u
s
tn
ess
.
T
ab
le
1
.
Simu
latio
n
p
ar
am
eter
s
C
a
t
e
g
o
r
y
P
a
r
a
me
t
e
r
s
V
a
l
u
e
s
D
a
t
a
s
e
t
& E
x
p
e
r
i
m
e
n
t
a
l
S
e
t
u
p
N
u
mb
e
r
o
f
d
r
i
v
e
r
s
1
C
P
U
I
n
t
e
l
C
o
r
e
i
7
R
A
M
1
6
G
B
F
r
a
me
r
a
t
e
3
0
f
p
s
Te
mp
o
r
a
l
s
e
g
m
e
n
t
a
t
i
o
n
S
l
i
d
i
n
g
w
i
n
d
o
w
D
r
o
w
si
n
e
ss
c
l
a
sses
D
r
o
w
sy
/
N
o
n
-
d
r
o
w
s
y
D
i
st
r
a
c
t
i
o
n
c
l
a
ss
e
s
D
i
st
r
a
c
t
e
d
/
A
t
t
e
n
t
i
v
e
S
c
e
n
a
r
i
o
s
N
o
r
mal
,
S
u
n
g
l
a
sses
,
S
u
r
g
i
c
a
l
M
a
s
k
R
e
g
i
o
n
s
o
f
I
n
t
e
r
e
s
t
Te
mp
o
r
a
l
w
i
n
d
o
w
si
z
e
T
[
5
0
-
2
0
0
0
]
f
r
a
mes
Ey
e
R
O
I
si
z
e
6
4
×
1
2
8
M
o
u
t
h
R
O
I
si
z
e
6
4
×
6
4
H
e
a
d
R
O
I
si
z
e
1
2
8
×
1
2
8
C
N
N
P
a
r
a
m
e
t
e
r
s
K
e
r
n
e
l
si
z
e
(
3
×
3
)
N
u
mb
e
r
o
f
f
i
l
t
e
r
s
3
2
(
C
o
n
v
1
)
,
6
4
(
C
o
n
v
2
)
A
c
t
i
v
a
t
i
o
n
f
u
n
c
t
i
o
n
R
e
LU
P
o
o
l
i
n
g
M
a
x
P
o
o
l
i
n
g
(
2
×
2
)
F
e
a
t
u
r
e
a
g
g
r
e
g
a
t
i
o
n
G
l
o
b
a
l
A
v
e
r
a
g
e
P
o
o
l
i
n
g
O
u
t
p
u
t
a
c
t
i
v
a
t
i
o
n
S
i
g
m
o
i
d
M
LP
I
n
p
u
t
s
S
e
ye
s
,
S
m
ou
t
h
H
i
d
d
e
n
l
a
y
e
r
s
1
H
i
d
d
e
n
u
n
i
t
s
16
O
u
t
p
u
t
a
c
t
i
v
a
t
i
o
n
S
i
g
m
o
i
d
3
.
1
.
I
dea
l
co
nd
it
io
n
I
n
th
is
s
ce
n
ar
io
,
th
e
d
r
iv
er
d
o
es
n
o
t
wea
r
s
u
n
g
lass
es
o
r
a
s
u
r
g
ical
m
ask
.
Un
d
er
t
h
ese
co
n
d
itio
n
s
,
th
e
f
ac
ial
r
eg
io
n
s
o
f
in
ter
est
(
e
y
es,
m
o
u
th
,
an
d
h
ea
d
)
ar
e
clea
r
ly
v
is
ib
le,
allo
win
g
th
e
p
r
o
p
o
s
e
d
s
y
s
tem
to
u
s
e
all
v
is
u
al
in
f
o
r
m
atio
n
f
o
r
d
etec
tio
n
.
Fig
u
r
e
4
s
h
o
ws
th
e
p
e
r
f
o
r
m
an
ce
o
f
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
d
etec
tio
n
i
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
4
,
Au
g
u
s
t
20
2
6
:
2
2
2
0
-
2229
2226
id
ea
l
co
n
d
itio
n
s
,
wh
er
e
t
h
e
d
r
i
v
er
d
o
es
n
o
t
wea
r
s
u
n
g
lass
es
o
r
a
s
u
r
g
ical
m
ask
.
W
e
o
b
s
er
v
e
th
at
as
th
e
s
ize
o
f
th
e
tim
e
win
d
o
w
T
in
cr
ea
s
es,
th
e
ac
cu
r
ac
y
im
p
r
o
v
es f
o
r
b
o
th
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
d
et
ec
tio
n
.
Dis
tr
ac
tio
n
d
etec
tio
n
g
i
v
es
s
lig
h
tly
b
etter
r
esu
lts
,
r
ea
ch
i
n
g
ab
o
u
t
98%
,
wh
ile
d
r
o
wsi
n
ess
d
etec
tio
n
r
ea
ch
es
ar
o
u
n
d
9
7
.
3
%
at
lar
g
e
T
.
T
h
is
ca
n
b
e
ex
p
lain
e
d
b
y
t
h
e
f
ac
t
th
at
d
is
tr
ac
tio
n
s
ar
e
p
r
im
ar
ily
r
elate
d
t
o
h
ea
d
m
o
v
em
en
ts
,
wh
ich
a
r
e
e
asier
to
d
etec
t th
an
th
e
s
u
b
tle
f
ac
ial
ch
an
g
es a
s
s
o
ciate
d
with
d
r
o
wsi
n
ess
,
s
u
ch
as
ey
e
clo
s
u
r
e
a
n
d
y
aw
n
in
g
.
B
o
th
cu
r
v
es
b
ec
o
m
e
s
tab
le
f
o
r
lar
g
e
tem
p
o
r
al
win
d
o
ws,
s
h
o
win
g
th
at
lo
n
g
e
r
win
d
o
ws
b
r
in
g
litt
le
im
p
r
o
v
e
m
en
t.
T
h
ese
r
esu
lts
co
n
f
ir
m
th
e
r
o
b
u
s
tn
ess
an
d
ef
f
ec
tiv
e
n
e
s
s
o
f
o
u
r
p
r
o
p
o
s
ed
s
y
s
tem
in
id
ea
l d
r
iv
in
g
co
n
d
itio
n
s
.
Fig
u
r
e
4
.
Dis
tr
ac
tio
n
v
s
d
r
o
wsi
n
ess
d
etec
tio
n
p
er
f
o
r
m
an
ce
(
id
ea
l c
o
n
d
itio
n
s
)
3
.
2
.
Su
ng
la
s
s
es
c
o
nd
it
io
n
I
n
r
ea
l
d
r
iv
in
g
s
itu
atio
n
s
,
m
a
n
y
d
r
i
v
er
s
wea
r
s
u
n
g
lass
es
to
p
r
o
tect
th
eir
e
y
es
f
r
o
m
s
u
n
lig
h
t.
Vis
u
al
in
f
o
r
m
atio
n
f
r
o
m
th
e
ey
es
,
s
u
ch
as
e
y
e
clo
s
u
r
e
an
d
b
lin
k
in
g
,
p
la
y
a
n
im
p
o
r
tan
t
r
o
le
in
d
etec
tin
g
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
,
as
in
Fig
u
r
e
5
.
Ho
wev
er
,
s
u
n
g
lass
es
m
ak
e
t
h
is
task
m
o
r
e
c
h
allen
g
in
g
b
y
p
ar
tially
h
id
i
n
g
th
e
ey
es,
ca
u
s
in
g
r
e
f
lectio
n
s
,
an
d
r
ed
u
cin
g
im
a
g
e
q
u
ality
.
T
h
er
ef
o
r
e,
th
is
s
ec
tio
n
an
al
y
ze
s
th
e
im
p
ac
t
o
f
s
u
n
g
lass
es
o
n
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
.
Fig
u
r
e
5
(
a)
s
h
o
ws
th
e
ef
f
ec
t
o
f
s
u
n
g
lass
es
o
n
d
r
o
wsi
n
ess
d
etec
tio
n
ac
cu
r
ac
y
f
o
r
d
if
f
er
en
t
tem
p
o
r
al
win
d
o
w
s
izes
T
.
W
ith
o
u
t
s
u
n
g
lass
es,
ac
cu
r
ac
y
in
c
r
ea
s
es
f
r
o
m
ab
o
u
t
95%
to
ap
p
r
o
x
im
ately
9
7
.
3
%
,
as
T
in
cr
ea
s
es.
W
e
n
o
tice
th
at
wi
th
s
u
n
g
lass
es,
th
e
ac
cu
r
ac
y
is
s
ig
n
if
ican
tly
lo
wer
,
r
ea
c
h
in
g
ar
o
u
n
d
9
0
%
,
at
lar
g
e
T
.
T
h
is
d
if
f
er
en
ce
(
ab
o
u
t
7%
)
is
to
s
tr
o
n
g
e
y
e
o
cc
lu
s
io
n
an
d
r
ef
lectio
n
s
ca
u
s
ed
b
y
s
u
n
g
lass
es
.
W
e
o
b
s
er
v
e
th
at
ev
e
n
with
s
u
n
g
lass
es,
th
e
s
y
s
tem
s
till
p
er
f
o
r
m
s
well.
W
h
en
en
o
u
g
h
f
r
a
m
es
ar
e
u
s
ed
,
th
e
ac
cu
r
ac
y
b
ec
o
m
es
s
tab
le
an
d
r
em
ai
n
s
h
ig
h
.
T
h
is
s
h
o
w
s
th
at
th
e
p
r
o
p
o
s
ed
s
y
s
tem
is
r
o
b
u
s
t a
n
d
ca
n
d
etec
t d
r
o
wsi
n
ess
ev
en
wh
en
th
e
d
r
iv
er
wea
r
s
s
u
n
g
lass
es.
Fig
u
r
e
5
(
b
)
illu
s
tr
ates
th
e
ef
f
ec
t
o
f
s
u
n
g
lass
es
o
n
d
is
tr
ac
tio
n
d
etec
tio
n
ac
cu
r
ac
y
.
W
e
o
b
s
er
v
e
th
at
wh
en
th
e
d
r
iv
er
is
n
o
t
wea
r
in
g
s
u
n
g
lass
es,
th
e
s
y
s
tem
a
ch
iev
es
in
cr
ea
s
ed
ac
cu
r
ac
y
f
o
r
all
o
b
s
er
v
atio
n
d
u
r
atio
n
s
.
T
h
e
ac
cu
r
ac
y
in
cr
e
ases
g
r
ad
u
ally
with
an
d
r
ea
ch
es
ab
o
u
t
9
8
%
f
o
r
lar
g
e
wi
n
d
o
ws.
Ho
wev
er
,
w
h
en
s
u
n
g
lass
es
ar
e
wo
r
n
,
th
e
ac
cu
r
ac
y
is
s
lig
h
tly
lo
wer
,
esp
ec
ially
f
o
r
s
m
all
v
alu
es
o
f
.
Ho
wev
er
,
as
in
cr
ea
s
es,
th
e
p
er
f
o
r
m
a
n
ce
i
m
p
r
o
v
es
an
d
s
tab
ilizes
ar
o
u
n
d
9
6
.
5
%
.
T
h
e
p
er
f
o
r
m
a
n
ce
g
ap
b
etwe
en
th
e
two
ca
s
es
r
em
ain
s
lim
ited
(
ar
o
u
n
d
1
.
5
%
)
.
T
h
ese
r
esu
lts
in
d
icate
th
at
s
u
n
g
lass
es
h
av
e
a
m
in
o
r
im
p
ac
t
o
n
d
is
tr
ac
tio
n
d
etec
tio
n
,
a
n
d
th
e
p
r
o
p
o
s
ed
s
y
s
tem
r
em
ain
s
r
o
b
u
s
t e
v
en
wh
en
ey
e
v
is
ib
ilit
y
is
p
ar
tially
h
id
d
e
n
.
3
.
3
.
Su
rg
ica
l
m
a
s
k
c
o
nd
it
io
n
Sin
ce
th
e
C
OVI
D
-
1
9
p
an
d
e
m
ic,
wea
r
in
g
f
ac
e
m
ask
s
(
s
u
r
g
i
ca
l
m
ask
s
o
r
cl
o
th
m
ask
s
)
h
a
s
b
ec
o
m
e
co
m
m
o
n
in
v
e
h
icles,
p
ar
ticu
la
r
ly
f
o
r
p
r
o
f
ess
io
n
al
d
r
iv
e
r
s
s
u
ch
as
tax
is
an
d
p
u
b
lic
tr
an
s
p
o
r
t.
Face
m
ask
s
h
id
e
th
e
m
o
u
th
r
e
g
io
n
an
d
p
ar
t
o
f
t
h
e
f
ac
e,
wh
ich
is
a
cr
itical
s
o
u
r
ce
o
f
in
f
o
r
m
atio
n
f
o
r
d
etec
t
in
g
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
.
T
h
is
s
ec
tio
n
an
aly
ze
s
th
e
im
p
ac
t o
f
s
u
r
g
ical
m
ask
s
o
n
th
e
p
r
o
p
o
s
ed
s
y
s
te
m
,
s
ee
Fig
u
r
e
6
.
Fig
u
r
e
6
(
a)
illu
s
tr
ates
th
e
im
p
ac
t
o
f
wea
r
in
g
a
s
u
r
g
ical
m
ask
o
n
d
r
o
wsi
n
ess
d
etec
tio
n
ac
c
u
r
ac
y
f
o
r
d
if
f
er
en
t
tem
p
o
r
al
win
d
o
w
s
izes
T
.
I
n
b
o
th
s
ce
n
ar
io
s
,
we
o
b
s
er
v
e
th
at
ac
c
u
r
ac
y
i
n
cr
ea
s
es
as
T
g
r
o
ws.
W
ith
o
u
t
a
m
ask
,
th
e
ac
cu
r
ac
y
r
ap
id
l
y
in
cr
ea
s
es
an
d
s
tab
ilizes
at
ar
o
u
n
d
97%
f
o
r
lar
g
e
tem
p
o
r
al
win
d
o
ws.
W
h
en
th
e
d
r
iv
e
r
wea
r
s
a
s
u
r
g
i
ca
l m
ask
,
th
e
ac
cu
r
ac
y
is
lo
wer
an
d
c
o
n
v
e
r
g
es
to
o
a
p
p
r
o
x
im
ately
91%
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
co
n
vo
l
u
tio
n
a
l
n
eu
r
a
l n
etw
o
r
k
-
b
a
s
ed
d
r
iver m
o
n
ito
r
in
g
s
y
s
tem
fo
r
…
(
S
a
r
a
B
en
ko
u
i
d
er
)
2227
T
h
is
d
ec
r
ea
s
e
o
f
ab
o
u
t
6%
is
m
ain
ly
d
u
e
to
t
h
e
lo
s
s
o
f
m
o
u
th
-
r
elate
d
v
is
u
al
in
f
o
r
m
atio
n
,
s
u
ch
as
y
awn
in
g
,
wh
ich
is
h
id
d
en
b
y
th
e
m
ask
.
Ho
wev
er
,
o
u
r
s
y
s
tem
s
till
p
er
f
o
r
m
s
well
w
ith
a
m
ask
wo
r
n
,
esp
ec
ially
f
o
r
lar
g
e
v
al
u
es o
f
T
.
T
h
ese
r
esu
lts
d
em
o
n
s
tr
ate
th
at
th
e
p
r
o
p
o
s
ed
m
eth
o
d
is
r
o
b
u
s
t e
v
en
wh
en
p
ar
t o
f
th
e
f
ac
e
is
h
id
d
en
.
Fig
u
r
e
6
(
b
)
s
h
o
ws
th
e
i
n
f
lu
en
ce
o
f
a
s
u
r
g
ical
m
ask
o
n
d
is
tr
ac
tio
n
d
etec
tio
n
ac
cu
r
ac
y
.
I
n
b
o
th
ca
s
es,
th
e
ac
cu
r
ac
y
i
n
cr
ea
s
es
wh
en
th
e
tem
p
o
r
al
win
d
o
w
s
ize
T
b
ec
o
m
es
lar
g
er
.
W
ith
o
u
t
a
s
u
r
g
ical
m
ask
,
th
e
ac
cu
r
ac
y
in
cr
ea
s
es
q
u
ick
ly
a
n
d
s
tab
ilizes
at
ar
o
u
n
d
98%
f
o
r
lar
g
e
tem
p
o
r
al
win
d
o
ws.
W
h
en
th
e
d
r
iv
er
wea
r
s
a
s
u
r
g
ical
m
ask
,
th
e
ac
c
u
r
ac
y
is
s
lig
h
tly
lo
wer
an
d
co
n
v
er
g
e
s
to
ap
p
r
o
x
im
ately
97%
.
T
h
e
p
er
f
o
r
m
an
ce
d
r
o
p
is
th
er
ef
o
r
e
lim
ited
to
ab
o
u
t
1%
,
wh
ich
in
d
icate
s
th
at
m
ask
wea
r
in
g
h
as
a
v
er
y
s
m
all
im
p
ac
t
o
n
o
u
r
d
is
tr
ac
tio
n
d
etec
tio
n
s
y
s
tem
.
B
ec
au
s
e
it
m
ain
ly
d
ep
en
d
s
o
n
h
ea
d
m
o
v
em
e
n
ts
,
wh
ich
ar
e
n
o
t
s
tr
o
n
g
ly
af
f
ec
te
d
b
y
wea
r
in
g
a
m
ask
.
T
h
ese
r
e
s
u
lts
s
h
o
w
th
at
th
e
p
r
o
p
o
s
ed
s
y
s
tem
wo
r
k
s
well
f
o
r
d
is
tr
ac
tio
n
d
etec
tio
n
e
v
en
wh
en
th
e
d
r
iv
er
wea
r
s
a
s
u
r
g
ical
m
ask
.
(
a)
(
b
)
Fig
u
r
e
5
.
E
f
f
ec
t o
f
s
u
n
g
lass
es o
n
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
d
etec
tio
n
p
er
f
o
r
m
a
n
ce
: (
a)
d
r
o
wsi
n
ess
d
etec
tio
n
an
d
(
b
)
d
is
tr
ac
tio
n
d
etec
tio
n
(
a)
(
b
)
Fig
u
r
e
6
.
E
f
f
ec
t o
f
s
u
r
g
ical
m
a
s
k
o
n
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
d
etec
tio
n
p
er
f
o
r
m
an
ce
:
(
a)
d
r
o
wsi
n
ess
d
etec
tio
n
an
d
(
b
)
d
is
tr
ac
tio
n
d
etec
tio
n
4.
CO
NCLU
SI
O
N
I
n
th
is
p
ap
e
r
,
we
p
r
o
p
o
s
ed
a
s
y
s
tem
to
d
etec
t
d
r
iv
er
d
r
o
wsi
n
ess
an
d
d
is
tr
ac
tio
n
u
s
in
g
f
ac
i
al
an
aly
s
is
f
r
o
m
an
o
n
-
b
o
ar
d
ca
m
er
a.
O
u
r
s
y
s
tem
u
s
es
m
o
d
u
lar
ar
ch
ite
ctu
r
e
with
C
NNs
to
an
aly
ze
th
e
ey
es,
m
o
u
t
h
,
an
d
h
ea
d
r
eg
i
o
n
s
,
co
m
b
in
ed
with
a
s
im
p
le
tem
p
o
r
al
f
u
s
io
n
m
eth
o
d
to
ca
p
tu
r
e
d
r
i
v
er
b
e
h
av
io
r
o
v
er
tim
e.
Du
e
to
its
m
o
d
u
lar
d
esig
n
,
th
e
s
y
s
tem
r
em
ain
s
r
o
b
u
s
t
to
p
ar
t
ial
in
f
o
r
m
atio
n
lo
s
s
an
d
ca
n
c
o
n
tin
u
e
to
o
p
er
ate
e
v
en
wh
e
n
s
o
m
e
f
ac
i
al
r
eg
io
n
s
a
r
e
m
is
s
in
g
o
r
o
c
clu
d
ed
,
w
h
ich
is
a
c
o
m
m
o
n
lim
itatio
n
o
f
m
a
n
y
ex
is
tin
g
ap
p
r
o
ac
h
es.
E
x
p
er
im
en
tal
r
esu
lts
d
em
o
n
s
tr
ate
th
at
th
e
p
r
o
p
o
s
ed
s
y
s
tem
ac
h
iev
es
h
ig
h
ac
cu
r
ac
y
an
d
s
tr
o
n
g
r
o
b
u
s
tn
ess
u
n
d
er
v
ar
i
o
u
s
d
r
iv
in
g
co
n
d
itio
n
s
.
I
n
p
ar
ticu
lar
,
it
attain
s
u
p
to
9
7
.
3
%
ac
cu
r
ac
y
f
o
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
4
,
Au
g
u
s
t
20
2
6
:
2
2
2
0
-
2229
2228
d
r
o
wsi
n
ess
d
etec
tio
n
an
d
9
8
%
f
o
r
d
is
tr
ac
tio
n
d
etec
tio
n
i
n
id
ea
l
s
ce
n
ar
i
o
s
.
E
v
en
wh
e
n
th
e
d
r
iv
er
wea
r
s
s
u
n
g
lass
es
o
r
a
s
u
r
g
ical
m
as
k
,
th
e
s
y
s
tem
m
ain
tain
s
r
elia
b
le
p
er
f
o
r
m
an
ce
,
with
ac
cu
r
a
cy
r
em
ain
in
g
ab
o
v
e
9
0
%
f
o
r
d
r
o
wsi
n
ess
an
d
9
7
%
f
o
r
d
is
tr
ac
tio
n
,
co
n
f
ir
m
in
g
its
s
u
itab
ili
ty
f
o
r
r
ea
l
-
tim
e
d
r
iv
er
m
o
n
ito
r
in
g
ap
p
licatio
n
s
.
As
f
u
tu
r
e
wo
r
k
,
we
p
r
o
p
o
s
e
to
in
teg
r
ate
p
er
s
o
n
alize
d
a
d
a
p
tatio
n
s
tr
ateg
ies
th
at
tak
e
in
to
ac
co
u
n
t
in
d
iv
id
u
al
d
r
iv
er
ch
ar
ac
ter
is
t
ics,
wh
ich
co
u
ld
im
p
r
o
v
e
s
y
s
tem
ac
cu
r
ac
y
an
d
r
e
d
u
ce
d
etec
tio
n
er
r
o
r
s
.
I
n
ad
d
itio
n
,
in
tr
o
d
u
cin
g
tem
p
o
r
al
weig
h
tin
g
o
r
atten
tio
n
m
ec
h
a
n
is
m
s
wo
u
ld
allo
w
m
o
r
e
ef
f
ec
tiv
e
ex
p
lo
itatio
n
o
f
th
e
m
o
s
t
r
elev
an
t
in
f
o
r
m
atio
n
with
in
th
e
tem
p
o
r
al
win
d
o
w.
Fin
ally
,
a
co
m
p
r
e
h
en
s
iv
e
ev
alu
atio
n
u
n
d
er
r
ea
l
d
r
iv
in
g
co
n
d
itio
n
s
,
i
n
clu
d
in
g
d
if
f
er
en
t
d
r
iv
e
r
p
r
o
f
iles
a
n
d
v
ar
io
u
s
d
r
iv
in
g
s
ce
n
ar
io
s
s
u
c
h
as
n
ig
h
ttime
d
r
iv
in
g
,
ad
v
er
s
e
wea
th
er
co
n
d
itio
n
s
,
a
n
d
d
en
s
e
tr
af
f
ic,
is
ess
en
tial to
v
alid
ate
th
e
r
o
b
u
s
tn
ess
an
d
p
r
ac
tical
ap
p
licab
ilit
y
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
.
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
Sar
a
B
en
k
o
u
id
er
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Nasr
ed
d
in
e
L
ag
r
aa
✓
✓
✓
✓
✓
✓
✓
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
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
.
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
D
a
t
a
a
v
a
il
a
b
i
li
t
y
is
n
o
t
a
p
p
l
i
ca
b
l
e
t
o
t
h
is
p
a
p
e
r
a
s
n
o
n
e
w
d
at
a
w
e
r
e
c
r
e
a
t
e
d
o
r
a
n
al
y
z
e
d
i
n
t
h
is
s
t
u
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
W
o
r
l
d
H
e
a
l
t
h
O
r
g
a
n
i
z
a
t
i
o
n
,
“
G
l
o
b
a
l
s
t
a
t
u
s r
e
p
o
r
t
o
n
r
o
a
d
s
a
f
e
t
y
,
”
G
e
n
e
v
a
,
S
w
i
t
z
e
r
l
a
n
d
,
2
0
2
3
.
[
2
]
J.
A
.
H
o
r
n
e
a
n
d
L
.
A
.
R
e
y
n
e
r
,
“
S
l
e
e
p
r
e
l
a
t
e
d
v
e
h
i
c
l
e
a
c
c
i
d
e
n
t
s
,
”
B
MJ
,
v
o
l
.
3
1
0
,
n
o
.
6
9
7
9
,
p
p
.
5
6
5
–
5
6
7
,
M
a
r
.
1
9
9
5
,
d
o
i
:
1
0
.
1
1
3
6
/
b
mj
.
3
1
0
.
6
9
7
9
.
5
6
5
.
[
3
]
A
.
S
a
h
a
y
a
d
h
a
s
,
K
.
S
u
n
d
a
r
a
j
,
a
n
d
M
.
M
u
r
u
g
a
p
p
a
n
,
“
D
e
t
e
c
t
i
n
g
d
r
i
v
e
r
d
r
o
w
si
n
e
ss
b
a
se
d
o
n
se
n
so
r
s
:
A
r
e
v
i
e
w
,
”
S
e
n
so
rs
,
v
o
l
.
1
2
,
n
o
.
1
2
,
p
p
.
1
6
9
3
7
–
1
6
9
5
3
,
D
e
c
.
2
0
1
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s
1
2
1
2
1
6
9
3
7
.
[
4
]
N
a
t
i
o
n
a
l
H
i
g
h
w
a
y
Tr
a
f
f
i
c
S
a
f
e
t
y
A
d
mi
n
i
st
r
a
t
i
o
n
,
“
D
i
s
t
r
a
c
t
e
d
d
r
i
v
i
n
g
r
e
p
o
r
t
,
”
W
a
s
h
i
n
g
t
o
n
,
D
C
,
U
S
A
,
2
0
2
2
.
[
5
]
G
.
A
l
h
a
m
a
d
a
n
d
M
.
-
B
.
K
u
r
d
y
,
“
A
r
e
v
i
e
w
o
f
d
r
i
v
e
r
d
i
s
t
r
a
c
t
i
o
n
d
e
t
e
c
t
i
o
n
w
h
i
l
e
d
r
i
v
i
n
g
b
a
s
e
d
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
s,”
I
AES
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
(
I
J
-
AI
)
,
v
o
l
.
1
4
,
n
o
.
6
,
p
p
.
4
4
1
5
–
4
4
2
6
,
D
e
c
.
2
0
2
5
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
a
i
.
v
1
4
.
i
6
.
p
p
4
4
1
5
-
4
4
2
6
.
[
6
]
T.
F
o
n
se
c
a
a
n
d
S
.
F
e
r
r
e
i
r
a
,
“
D
r
o
w
s
i
n
e
ss
d
e
t
e
c
t
i
o
n
i
n
d
r
i
v
e
r
s:
a
s
y
st
e
m
a
t
i
c
r
e
v
i
e
w
o
f
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
mo
d
e
l
s
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
5
,
n
o
.
1
6
,
p
.
9
0
1
8
,
A
u
g
.
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
5
1
6
9
0
1
8
.
[
7
]
R
.
F
l
o
r
e
z
,
F
.
P
a
l
o
m
i
n
o
-
Q
u
i
s
p
e
,
R
.
J
.
C
o
a
q
u
i
r
a
-
C
a
st
i
l
l
o
,
J.
C
.
H
e
r
r
e
r
a
-
Le
v
a
n
o
,
T.
P
a
i
x
ã
o
,
a
n
d
A
.
B
.
A
l
v
a
r
e
z
,
“
A
C
N
N
-
b
a
s
e
d
a
p
p
r
o
a
c
h
f
o
r
d
r
i
v
e
r
d
r
o
w
s
i
n
e
ss
d
e
t
e
c
t
i
o
n
b
y
r
e
a
l
-
t
i
m
e
e
y
e
st
a
t
e
i
d
e
n
t
i
fi
c
a
t
i
o
n
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
3
,
n
o
.
1
3
.
p
p
.
7
8
4
9
–
7
8
6
6
,
Ju
n
.
0
5
,
2
0
2
3
,
d
o
i
:
1
0
.
2
0
9
4
4
/
p
r
e
p
r
i
n
t
s
2
0
2
3
0
6
.
0
2
6
7
.
v
1
.
[
8
]
A
.
S
o
h
a
i
l
,
A
.
A
.
S
h
a
h
,
S
.
I
l
y
a
s
,
a
n
d
N
.
A
l
sh
a
mm
r
y
,
“
A
C
N
N
-
b
a
s
e
d
d
e
e
p
l
e
a
r
n
i
n
g
f
r
a
mew
o
r
k
f
o
r
d
r
i
v
e
r
’
s
d
r
o
w
si
n
e
ss
d
e
t
e
c
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
A
d
v
a
n
c
e
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
5
,
n
o
.
3
,
p
p
.
1
6
9
–
1
7
8
,
2
0
2
4
,
d
o
i
:
1
0
.
1
4
5
6
9
/
I
JA
C
S
A
.
2
0
2
4
.
0
1
5
0
3
1
7
.
[
9
]
D
.
S
a
l
e
m a
n
d
M
.
W
a
l
e
e
d
,
“
D
r
o
w
si
n
e
ss
d
e
t
e
c
t
i
o
n
i
n
r
e
a
l
-
t
i
me
v
i
a
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 a
n
d
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
,
”
J
o
u
r
n
a
l
o
f
En
g
i
n
e
e
ri
n
g
a
n
d
A
p
p
l
i
e
d
S
c
i
e
n
c
e
,
v
o
l
.
7
1
,
n
o
.
1
,
p
.
1
2
2
,
D
e
c
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
8
6
/
s
4
4
1
4
7
-
0
2
4
-
0
0
4
5
7
-
z.
[
1
0
]
O
.
F
.
H
a
ss
a
n
,
A
.
F
.
I
b
r
a
h
i
m,
A
.
G
o
m
a
a
,
M
.
A
.
M
a
k
h
l
o
u
f
,
a
n
d
B
.
H
a
f
i
z
,
“
R
e
a
l
-
t
i
m
e
d
r
i
v
e
r
d
r
o
w
si
n
e
ss
d
e
t
e
c
t
i
o
n
u
si
n
g
t
r
a
n
sf
o
r
m
e
r
a
r
c
h
i
t
e
c
t
u
r
e
s
:
a
n
o
v
e
l
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
rt
s
,
v
o
l
.
1
5
,
n
o
.
1
,
p
.
1
7
4
9
3
,
M
a
y
2
0
2
5
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
5
9
8
-
025
-
0
2
1
1
1
-
x.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
co
n
vo
l
u
tio
n
a
l
n
eu
r
a
l n
etw
o
r
k
-
b
a
s
ed
d
r
iver m
o
n
ito
r
in
g
s
y
s
tem
fo
r
…
(
S
a
r
a
B
en
ko
u
i
d
er
)
2229
[
1
1
]
A
.
M
a
d
a
n
e
,
A
.
S
i
n
g
h
,
S
.
F
a
r
g
a
d
e
,
a
n
d
A
.
D
o
n
g
a
r
e
,
“
R
e
a
l
-
t
i
me
d
r
i
v
e
r
d
r
o
w
si
n
e
ss
d
e
t
e
c
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
c
o
mp
u
t
e
r
v
i
s
i
o
n
t
e
c
h
n
i
q
u
e
s
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
S
c
i
e
n
t
i
f
i
c
Re
se
a
rc
h
i
n
S
c
i
e
n
c
e
,
E
n
g
i
n
e
e
r
i
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
2
,
n
o
.
3
,
p
p
.
2
2
2
–
2
2
9
,
M
a
y
2
0
2
5
,
d
o
i
:
1
0
.
3
2
6
2
8
/
I
JS
R
S
ET2
5
1
2
3
3
5
.
[
1
2
]
K
.
A
.
A
l
S
h
a
l
f
a
n
a
n
d
M
.
Za
k
a
r
i
a
h
,
“
D
e
t
e
c
t
i
n
g
d
r
i
v
e
r
d
i
st
r
a
c
t
i
o
n
u
s
i
n
g
d
e
e
p
-
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
C
o
m
p
u
t
e
rs,
Ma
t
e
r
i
a
l
s
&
C
o
n
t
i
n
u
a
,
v
o
l
.
6
8
,
n
o
.
1
,
p
p
.
6
8
9
–
7
0
4
,
2
0
2
1
,
d
o
i
:
1
0
.
3
2
6
0
4
/
c
m
c
.
2
0
2
1
.
0
1
5
9
8
9
.
[
1
3
]
R
.
S
a
l
a
k
a
p
u
r
i
,
N
.
K
.
N
a
v
u
r
i
,
T.
V
o
b
b
i
l
i
n
e
n
i
,
G
.
R
a
v
i
,
K
.
K
a
r
ma
k
o
n
d
a
,
a
n
d
K
.
A
.
V
a
r
d
h
a
n
,
“
I
n
t
e
g
r
a
t
e
d
d
e
e
p
l
e
a
r
n
i
n
g
f
r
a
mew
o
r
k
f
o
r
d
r
i
v
e
r
d
i
st
r
a
c
t
i
o
n
d
e
t
e
c
t
i
o
n
a
n
d
r
e
a
l
-
t
i
m
e
r
o
a
d
o
b
j
e
c
t
r
e
c
o
g
n
i
t
i
o
n
i
n
a
d
v
a
n
c
e
d
d
r
i
v
e
r
a
ssi
s
t
a
n
c
e
s
y
st
e
ms,
”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
r
t
s
,
v
o
l
.
1
5
,
n
o
.
1
,
p
.
2
5
1
2
5
,
J
u
l
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
5
9
8
-
025
-
0
8
4
7
5
-
4.
[
1
4
]
B
.
B
a
h
e
t
i
,
S
.
G
a
j
r
e
,
a
n
d
S
.
T
a
l
b
a
r
,
“
D
e
t
e
c
t
i
o
n
o
f
d
i
st
r
a
c
t
e
d
d
r
i
v
e
r
u
si
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
i
n
2
0
1
8
I
EEE
/
C
V
F
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
s
i
o
n
a
n
d
P
a
t
t
e
r
n
R
e
c
o
g
n
i
t
i
o
n
W
o
r
k
sh
o
p
s
(
C
V
PRW
)
,
Ju
n
.
2
0
1
8
,
p
p
.
1
1
4
5
–
1
1
4
5
6
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
W
.
2
0
1
8
.
0
0
1
5
0
.
[
1
5
]
M
.
A
l
j
a
si
m
a
n
d
R
.
K
a
s
h
e
f
,
“
E
2
D
R
:
A
d
e
e
p
l
e
a
r
n
i
n
g
e
n
s
e
m
b
l
e
-
b
a
s
e
d
d
r
i
v
e
r
d
i
st
r
a
c
t
i
o
n
d
e
t
e
c
t
i
o
n
w
i
t
h
r
e
c
o
mm
e
n
d
a
t
i
o
n
s
m
o
d
e
l
,
”
S
e
n
so
rs
,
v
o
l
.
2
2
,
n
o
.
5
,
p
.
1
8
5
8
,
F
e
b
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s2
2
0
5
1
8
5
8
.
[
1
6
]
S.
-
J.
Ji
a
o
,
L.
-
Y
.
L
i
u
,
a
n
d
Q
.
L
i
u
,
“
A
h
y
b
r
i
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
f
o
r
r
e
c
o
g
n
i
z
i
n
g
a
c
t
i
o
n
s
o
f
d
i
s
t
r
a
c
t
e
d
d
r
i
v
e
r
s,
”
S
e
n
so
rs
,
v
o
l
.
2
1
,
n
o
.
2
1
,
p
.
7
4
2
4
,
N
o
v
.
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
1
2
1
7
4
2
4
.
[
1
7
]
R
.
G
h
o
d
d
o
o
s
i
a
n
,
M
.
G
a
l
i
b
,
a
n
d
V
.
A
t
h
i
t
s
o
s,
“
A
r
e
a
l
i
s
t
i
c
d
a
t
a
se
t
a
n
d
b
a
se
l
i
n
e
t
e
mp
o
r
a
l
m
o
d
e
l
f
o
r
e
a
r
l
y
d
r
o
w
s
i
n
e
ss
d
e
t
e
c
t
i
o
n
,
”
i
n
2
0
1
9
I
EE
E/
C
VF
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
si
o
n
a
n
d
Pa
t
t
e
rn
Re
c
o
g
n
i
t
i
o
n
Wo
r
k
sh
o
p
s
(
C
VPR
W)
,
Ju
n
.
2
0
1
9
,
p
p
.
1
7
8
–
1
8
7
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
W
.
2
0
1
9
.
0
0
0
2
7
.
[
1
8
]
R
.
J
a
b
b
a
r
,
K
.
A
l
-
K
h
a
l
i
f
a
,
M
.
K
h
a
r
b
e
c
h
e
,
W
.
A
l
h
a
j
y
a
se
e
n
,
M
.
J
a
f
a
r
i
,
a
n
d
S
.
Ji
a
n
g
,
“
R
e
a
l
-
t
i
me
d
r
i
v
e
r
d
r
o
w
si
n
e
ss
d
e
t
e
c
t
i
o
n
f
o
r
A
n
d
r
o
i
d
a
p
p
l
i
c
a
t
i
o
n
u
s
i
n
g
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s
,
”
Pr
o
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
3
0
,
p
p
.
4
0
0
–
4
0
7
,
2
0
1
8
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s
.
2
0
1
8
.
0
4
.
0
6
0
.
[
1
9
]
F
.
F
a
r
a
j
i
,
F
L
o
t
f
i
,
J.
K
h
o
r
r
a
md
e
l
,
A
.
N
a
j
a
f
i
,
a
n
d
A
.
G
h
a
f
f
a
r
i
,
“
D
r
o
w
s
i
n
e
ss d
e
t
e
c
t
i
o
n
b
a
s
e
d
o
n
d
r
i
v
e
r
t
e
m
p
o
r
a
l
b
e
h
a
v
i
o
r
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
a
r
Xi
v
p
re
p
ri
n
t
a
r
Xi
v
:
2
1
0
4
.
0
0
1
2
5
,
v
o
l
.
a
b
s/
2
1
0
4
.
0
,
2
0
2
1
.
[
2
0
]
A
.
P
i
n
g
o
e
t
a
l
.
,
“
D
r
i
v
i
n
g
b
e
h
a
v
i
o
r
c
l
a
ssi
f
i
c
a
t
i
o
n
u
s
i
n
g
a
C
o
n
v
LST
M
,
”
Fu
t
u
re
T
r
a
n
s
p
o
rt
a
t
i
o
n
,
v
o
l
.
5
,
n
o
.
2
,
p
.
5
2
,
M
a
y
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
f
u
t
u
r
e
t
r
a
n
s
p
5
0
2
0
0
5
2
.
[
2
1
]
A
.
A
b
o
u
O
u
f
,
I
.
S
o
b
h
,
M
.
N
a
sser,
O
.
A
l
saq
a
,
O
.
El
e
z
a
b
y
,
a
n
d
J.
F
.
W
.
Za
k
i
,
“
M
u
l
t
i
m
o
d
e
l
sy
s
t
e
m
f
o
r
d
r
i
v
e
r
d
i
st
r
a
c
t
i
o
n
d
e
t
e
c
t
i
o
n
a
n
d
e
l
i
mi
n
a
t
i
o
n
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
0
,
p
p
.
7
2
4
5
8
–
7
2
4
6
9
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
2
.
3
1
8
8
7
1
5
.
[
2
2
]
M
.
L
i
u
,
X
.
X
u
,
J.
H
u
,
a
n
d
Q
.
J
i
a
n
g
,
“
R
e
a
l
-
t
i
me
d
e
t
e
c
t
i
o
n
o
f
d
r
i
v
e
r
f
a
t
i
g
u
e
b
a
s
e
d
o
n
C
N
N
–
LSTM
,
”
I
ET I
m
a
g
e
P
ro
c
e
ssi
n
g
,
v
o
l
.
1
6
,
n
o
.
2
,
p
p
.
5
7
6
–
5
9
5
,
F
e
b
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
4
9
/
i
p
r
2
.
1
2
3
7
3
.
[
2
3
]
O
.
N
a
si
r
e
t
a
l
.
,
“
S
A
F
E
-
D
R
I
V
E
-
A
I
:
A
C
N
N
–
LSTM
–
a
t
t
e
n
t
i
o
n
f
r
a
m
e
w
o
r
k
f
o
r
d
r
o
w
s
i
n
e
ss
d
e
t
e
c
t
i
o
n
,
”
En
g
i
n
e
e
ri
n
g
,
T
e
c
h
n
o
l
o
g
y
&
Ap
p
l
i
e
d
S
c
i
e
n
c
e
Re
s
e
a
rc
h
,
v
o
l
.
1
5
,
n
o
.
5
,
p
p
.
2
7
5
9
4
–
2
7
6
0
0
,
O
c
t
.
2
0
2
5
,
d
o
i
:
1
0
.
4
8
0
8
4
/
e
t
a
sr
.
1
2
7
2
5
.
[
2
4
]
J.
Li
u
,
Y
.
Li
u
,
D
.
L
i
,
H
.
W
a
n
g
,
X
.
H
u
a
n
g
,
a
n
d
L
.
S
o
n
g
,
“
D
S
D
C
LA
:
d
r
i
v
i
n
g
st
y
l
e
d
e
t
e
c
t
i
o
n
v
i
a
h
y
b
r
i
d
C
N
N
-
L
S
TM
w
i
t
h
m
u
l
t
i
-
l
e
v
e
l
a
t
t
e
n
t
i
o
n
f
u
s
i
o
n
,
”
A
p
p
l
i
e
d
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
5
3
,
n
o
.
1
6
,
p
p
.
1
9
2
3
7
–
1
9
2
5
4
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
4
8
9
-
0
2
3
-
0
4
4
5
1
-
5.
[
2
5
]
S
.
Ji
,
W
.
X
u
,
M
.
Y
a
n
g
,
a
n
d
K
.
Y
u
,
“
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
s
f
o
r
h
u
ma
n
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
Pa
t
t
e
r
n
An
a
l
y
si
s
a
n
d
Ma
c
h
i
n
e
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
3
5
,
n
o
.
1
,
p
p
.
2
2
1
–
2
3
1
,
Ja
n
.
2
0
1
3
,
d
o
i
:
1
0
.
1
1
0
9
/
TPA
M
I
.
2
0
1
2
.
5
9
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
S
a
r
a
Be
n
k
o
u
id
e
r
r
e
c
e
iv
e
d
h
e
r
M
.
S
c
d
e
g
re
e
i
n
c
o
m
p
u
ter
sc
ien
c
e
fro
m
Un
i
v
e
rsity
o
f
S
c
ien
c
e
s
a
n
d
Tec
h
n
o
l
o
g
y
Ho
u
a
ri
Bo
u
m
e
d
ien
e
,
(Al
g
iers
),
in
2
0
0
9
,
a
n
d
h
e
r
P
h
D
d
e
g
re
e
i
n
c
o
m
p
u
ter
sc
ien
c
e
fro
m
Un
iv
e
rsit
y
o
f
Lag
h
o
u
a
t,
Alg
e
ria,
i
n
2
0
2
0
.
S
h
e
is
c
u
rre
n
tl
y
a
m
e
m
b
e
r
a
t
th
e
In
fo
rm
a
ti
c
s
a
n
d
M
a
th
e
m
a
ti
c
s
Lab
o
ra
to
ry
(L
IM
)
o
f
t
h
e
U
n
iv
e
rsit
y
o
f
Lag
h
o
u
a
t.
He
r
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
Lo
c
a
li
z
a
ti
o
n
in
m
o
b
il
e
n
e
two
rk
s
(W
S
Ns
,
VA
Ne
ts
,
UA
Vs
),
Un
m
a
n
n
e
d
Ae
rial
Ve
h
icle
M
o
b
il
it
y
,
M
o
b
il
e
C
o
m
p
u
ti
n
g
,
Dri
v
e
r
m
o
n
it
o
rin
g
s
y
ste
m
s
a
n
d
c
o
ll
isio
n
a
v
o
id
a
n
c
e
i
n
a
sw
a
rm
o
f
UA
Vs
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
s.
b
e
n
k
o
u
i
d
e
r@lag
h
-
u
n
iv
.
d
z
.
Na
sr
e
d
d
in
e
La
g
r
a
a
r
e
c
e
iv
e
d
h
is
P
h
D
in
a
u
t
o
m
a
ti
o
n
fr
o
m
th
e
Eco
le
Na
ti
o
n
a
le
P
o
ly
tec
h
n
i
q
u
e
,
Alg
e
ria
(2
0
0
8
).
He
tea
c
h
e
s
v
a
rio
u
s
c
o
u
rse
s
o
n
c
o
m
p
u
ter
a
rc
h
it
e
c
tu
re
,
m
u
lt
ime
d
ia,
m
o
b
il
e
n
e
two
rk
s
a
n
d
n
e
two
rk
se
c
u
rit
y
.
His
re
se
a
rc
h
in
tere
sts
a
re
:
Ve
h
icu
lar
n
e
two
rk
s,
n
e
two
r
k
se
c
u
rit
y
,
d
e
c
e
n
tralize
d
c
o
n
tro
l,
f
u
z
z
y
,
a
n
d
a
rti
f
icia
l
n
e
u
ra
l
n
e
two
r
k
s.
He
i
s
se
rv
in
g
i
n
th
e
tec
h
n
ica
l
p
r
o
g
ra
m
c
o
m
m
it
tee
s
o
f
m
a
n
y
in
tern
a
ti
o
n
a
l
c
o
n
fe
re
n
c
e
s,
Wi
re
les
s
a
n
d
M
o
b
i
le
Co
m
p
u
ti
n
g
,
Ne
two
rk
in
g
a
n
d
C
o
m
m
u
n
ica
ti
o
n
s
(W
i
M
o
b
)
,
a
n
d
I
n
n
o
v
a
ti
o
n
s
i
n
In
fo
rm
a
ti
o
n
Tec
h
n
o
l
o
g
y
(IIT
)
.
H
e
c
a
n
b
e
c
o
n
tac
ted
a
t
n
.
la
g
ra
a
@la
g
h
-
u
n
i
v
.
d
z
.
Evaluation Warning : The document was created with Spire.PDF for Python.