I
nte
rna
t
io
na
l J
o
urna
l o
f
Adv
a
nces in Applie
d Science
s
(
I
J
AAS)
Vo
l.
15
,
No
.
1
,
Ma
r
ch
20
26
,
p
p
.
20
8
~
21
7
I
SS
N:
2252
-
8
8
1
4
,
DOI
:
1
0
.
1
1
5
9
1
/ijaas
.
v15.
i
1
.
pp
20
8
-
21
7
208
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
a
s
.
ia
esco
r
e.
co
m
Ro
bust mul
ti
-
fac
e
s recog
nition a
nd
tracking
via
f
uzzy
genetic
a
lg
o
rithms a
nd d
eep coupled
f
ea
t
u
res
Adil Ab
du
lh
ur
Abu
s
ha
na
1
,
Yo
us
if
Sa
m
er
M
ud
ha
f
a
r
1,
2
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
F
a
c
u
l
t
y
o
f
E
d
u
c
a
t
i
o
n
,
U
n
i
v
e
r
s
i
t
y
o
f
K
u
f
a
,
N
a
j
a
f
,
I
r
a
q
2
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
Te
c
h
n
i
q
u
e
s
En
g
i
n
e
e
r
i
n
g
,
F
a
c
u
l
t
y
o
f
Te
c
h
n
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
I
sl
a
mi
c
U
n
i
v
e
r
si
t
y
,
N
a
j
a
f
,
I
r
a
q
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
1
1
,
2
0
2
5
R
ev
is
ed
No
v
2
4
,
2
0
2
5
Acc
ep
ted
J
an
1
,
2
0
2
6
In
re
a
l
-
wo
rl
d
s
u
rv
e
il
lan
c
e
e
n
v
i
ro
n
m
e
n
ts,
fa
c
e
re
c
o
g
n
i
ti
o
n
a
n
d
trac
k
in
g
re
m
a
in
c
h
a
ll
e
n
g
in
g
d
u
e
to
p
a
rti
a
l
o
c
c
lu
sio
n
,
p
o
se
v
a
riati
o
n
,
il
l
u
m
in
a
ti
o
n
c
h
a
n
g
e
s,
a
n
d
b
a
c
k
g
ro
u
n
d
c
lu
tt
e
r.
Th
is
p
a
p
e
r
p
re
se
n
ts
a
ro
b
u
st
h
y
b
ri
d
fra
m
e
wo
rk
th
a
t
i
n
teg
ra
tes
f
u
z
z
y
g
e
n
e
ti
c
a
l
g
o
rit
h
m
s
(
F
G
A)
with
d
e
e
p
c
o
u
p
le
d
fe
a
tu
re
lea
rn
in
g
f
o
r
m
u
lt
i
-
fa
c
e
re
c
o
g
n
it
i
o
n
a
n
d
trac
k
in
g
.
T
h
e
p
ro
p
o
se
d
sy
ste
m
c
o
m
p
rise
s
th
re
e
m
a
in
m
o
d
u
les
:
i
)
fa
c
e
d
e
tec
ti
o
n
a
n
d
p
re
-
p
ro
c
e
ss
in
g
u
sin
g
t
h
e
m
u
lt
i
-
tas
k
c
a
sc
a
d
e
d
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
tw
o
rk
(
M
TCNN),
ii
)
d
e
e
p
c
o
u
p
led
Re
sN
e
t
e
m
b
e
d
d
in
g
s
t
h
a
t
jo
in
tl
y
lea
rn
id
e
n
ti
ty
a
n
d
a
p
p
e
a
ra
n
c
e
-
in
v
a
rian
t
re
p
re
se
n
tati
o
n
s,
a
n
d
iii
)
a
fu
z
z
y
ru
le
-
b
a
se
d
g
e
n
e
ti
c
o
p
ti
m
ize
r
th
a
t
a
d
a
p
ti
v
e
l
y
re
fin
e
s
trac
k
in
g
d
e
c
isio
n
s
b
a
se
d
o
n
u
n
c
e
rtain
ty
i
n
m
o
ti
o
n
,
a
p
p
e
a
ra
n
c
e
sim
il
a
rit
y
,
a
n
d
o
c
c
lu
sio
n
lev
e
ls.
Th
e
n
o
v
e
lt
y
o
f
th
is
wo
rk
li
e
s
in
t
h
e
f
u
sio
n
o
f
fu
z
z
y
i
n
fe
re
n
c
e
with
e
v
o
l
u
ti
o
n
a
ry
se
a
rc
h
to
g
u
id
e
th
e
g
e
n
e
ti
c
o
p
ti
m
iza
ti
o
n
p
r
o
c
e
ss
—
a
ll
o
win
g
d
y
n
a
m
ic
a
d
a
p
tati
o
n
t
o
n
o
isy
a
n
d
u
n
c
e
rtain
v
is
u
a
l
c
o
n
d
it
io
n
s.
M
o
r
e
o
v
e
r,
p
ro
b
a
b
il
isti
c
d
a
ta
a
ss
o
c
iat
io
n
fil
ters
(P
DA
F
)
a
n
d
c
o
n
d
i
ti
o
n
a
l
jo
in
t
l
i
k
e
li
h
o
o
d
fil
ters
(CJLF
)
a
re
e
m
p
lo
y
e
d
t
o
fu
rth
e
r
e
n
h
a
n
c
e
tem
p
o
ra
l
c
o
n
si
ste
n
c
y
u
n
d
e
r
o
c
c
lu
si
o
n
a
n
d
a
p
p
e
a
ra
n
c
e
v
a
riatio
n
.
Th
e
re
su
l
ts
c
o
n
firm
t
h
a
t
fu
z
z
y
e
v
o
lu
ti
o
n
a
ry
o
p
ti
m
iza
t
io
n
,
wh
e
n
c
o
u
p
le
d
wit
h
d
e
e
p
fe
a
tu
re
lea
rn
i
n
g
,
sig
n
ifi
c
a
n
tl
y
im
p
r
o
v
e
s
r
o
b
u
s
tn
e
ss
a
n
d
sta
b
il
it
y
fo
r
re
a
l
-
ti
m
e
fa
c
e
trac
k
in
g
in
c
o
m
p
lex
,
d
y
n
a
m
ic sc
e
n
e
s.
K
ey
w
o
r
d
s
:
Fu
zz
y
g
en
etic
alg
o
r
ith
m
s
Gen
etic
p
ar
ticle
f
ilter
in
g
J
o
in
t p
r
o
b
a
b
ilit
y
ass
o
ciatio
n
R
o
b
u
s
t
d
a
t
a
a
s
s
o
c
i
a
t
i
o
n
t
e
c
h
n
i
q
u
e
s
T
ex
tu
r
e
-
co
l
o
r
-
s
h
ap
e
f
ea
tu
r
es
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
:
Yo
u
s
if
Sam
er
Mu
d
h
a
f
ar
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
Facu
lty
o
f
E
d
u
ca
tio
n
,
Un
i
v
er
s
ity
o
f
Ku
f
a
Naja
f
,
I
r
aq
E
m
ail:
y
o
u
s
if
s
.
m
u
d
h
af
ar
@
u
o
k
u
f
a.
ed
u
.
iq
1.
I
NT
RO
D
UCT
I
O
N
Mu
lti
-
f
ac
e
r
ec
o
g
n
itio
n
an
d
tr
a
ck
in
g
a
r
e
ess
en
tial
in
s
u
r
v
eilla
n
ce
,
m
o
n
ito
r
in
g
s
y
s
tem
s
,
an
d
in
tellig
en
t
v
id
eo
a
n
aly
tics
[
1
]
,
[
2
]
,
b
u
t
p
er
f
o
r
m
an
ce
d
eg
r
a
d
es
u
n
d
er
o
cc
lu
s
io
n
,
p
o
s
e
ch
an
g
es,
v
is
u
al
s
im
ilar
ity
,
an
d
lo
w
-
r
eso
lu
tio
n
im
a
g
er
y
[
3
]
,
[
4
]
.
C
lass
ical
tr
ac
k
in
g
a
p
p
r
o
a
ch
es
r
ely
o
n
h
an
d
-
cr
af
te
d
f
e
atu
r
es
th
at
ar
e
n
o
t
r
o
b
u
s
t
to
s
u
ch
v
ar
iatio
n
s
,
w
h
ile
d
ee
p
lear
n
in
g
m
eth
o
d
s
h
av
e
im
p
r
o
v
ed
d
is
cr
im
in
ativ
e
f
ea
tu
r
e
ex
tr
ac
tio
n
th
r
o
u
g
h
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NN
s
)
[
5
]
,
[
6
]
.
Ho
wev
er
,
C
NN
-
b
ased
m
o
d
el
s
s
till
s
tr
u
g
g
le
wh
en
f
ac
es
ar
e
ca
p
tu
r
ed
at
lo
w
r
eso
lu
tio
n
o
r
u
n
d
e
r
p
o
o
r
lig
h
tin
g
co
n
d
itio
n
s
,
wh
ich
r
ed
u
ce
s
b
o
t
h
id
en
tific
atio
n
a
n
d
re
-
id
en
tific
atio
n
r
eliab
ilit
y
in
r
ea
l
-
wo
r
ld
s
u
r
v
eillan
ce
[
7
]
,
[
8
]
.
T
o
ad
d
r
ess
th
ese
ch
allen
g
es,
h
y
b
r
id
a
p
p
r
o
ac
h
es
co
m
b
in
in
g
d
e
ep
lear
n
in
g
with
o
p
tim
izat
io
n
-
b
ased
r
ea
s
o
n
in
g
h
av
e
g
ain
ed
atten
ti
o
n
.
Gen
etic
alg
o
r
ith
m
s
(
GAs)
p
r
o
v
id
e
ad
a
p
tiv
e
g
l
o
b
al
s
ea
r
c
h
,
an
d
wh
en
p
air
ed
with
f
u
zz
y
lo
g
ic,
ca
n
b
etter
h
an
d
le
u
n
ce
r
tain
t
y
an
d
p
ar
tial
v
is
ib
ilit
y
d
u
r
in
g
o
cc
l
u
s
io
n
[
9
]
.
T
h
u
s
,
in
teg
r
atin
g
d
ee
p
f
ea
tu
r
e
r
ep
r
esen
tatio
n
w
ith
f
u
zz
y
-
o
p
tim
ized
tr
ac
k
in
g
y
ield
s
a
m
o
r
e
r
o
b
u
s
t
an
d
a
d
a
p
tab
le
s
o
lu
tio
n
f
o
r
m
u
lti
-
f
ac
e
tr
ac
k
in
g
(
MFT
)
[
1
0
]
,
[
1
1
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
o
b
u
s
t m
u
lti
-
fa
ce
s
r
ec
o
g
n
itio
n
a
n
d
tr
a
ck
in
g
via
f
u
z
z
y
g
en
etic
a
lg
o
r
ith
ms a
n
d
…
(
A
d
il A
b
d
u
lh
u
r
A
b
u
s
h
a
n
a
)
209
R
ec
en
t
MFT
r
esear
ch
h
as
ev
o
lv
ed
f
r
o
m
d
etec
tio
n
-
b
ased
lin
k
in
g
to
ad
a
p
tiv
e
ass
o
ciatio
n
f
r
am
ewo
r
k
s
.
Ar
ac
h
ch
ilag
e
an
d
I
z
q
u
ier
d
o
[
1
2
]
im
p
r
o
v
ed
tem
p
o
r
al
co
n
s
is
ten
cy
th
r
o
u
g
h
ad
a
p
tiv
e
tr
ac
k
let
ag
g
r
eg
atio
n
,
wh
ile
B
ar
q
u
er
o
et
a
l.
[
1
3
]
a
d
d
r
ess
ed
cr
o
wd
ed
-
s
ce
n
e
r
ec
o
n
n
ec
ti
o
n
,
an
d
Z
h
an
g
et
a
l.
[
1
4
]
e
n
h
an
ce
d
ap
p
ea
r
an
ce
r
o
b
u
s
tn
ess
v
ia
u
n
s
u
p
er
v
is
ed
ad
ap
tatio
n
.
Fu
r
th
er
r
ef
in
em
e
n
ts
,
s
u
ch
as
v
er
if
icatio
n
-
b
ase
d
r
an
k
in
g
[
1
5
]
an
d
s
tr
u
ctu
r
ed
-
s
ce
n
e
o
p
tim
izatio
n
[
1
6
]
,
im
p
r
o
v
e
d
s
tab
ilit
y
b
u
t stru
g
g
led
u
n
d
er
s
ev
er
e
o
cc
lu
s
io
n
s
.
R
eg
io
n
al
s
im
p
le
o
n
lin
e
a
n
d
r
ea
l
-
tim
e
tr
ac
k
in
g
(
R
eSOR
T
)
in
tr
o
d
u
ce
d
I
D
r
ec
o
v
er
y
[
1
7
]
,
an
d
d
o
u
b
le
-
tr
ip
let
n
etwo
r
k
s
im
p
r
o
v
ed
cr
o
s
s
-
ca
m
er
a
co
n
s
is
ten
cy
[
1
8
]
.
Mo
r
e
r
ec
e
n
t
m
et
h
o
d
s
i
n
teg
r
ate
m
u
ltimo
d
al
cu
es,
u
s
in
g
b
o
th
f
ac
e
a
n
d
b
o
d
y
f
ea
tu
r
es
[
1
9
]
,
m
em
o
r
y
-
b
ased
m
atch
in
g
[
2
0
]
,
o
r
b
io
m
etr
ic
f
u
s
io
n
[
2
1
]
to
im
p
r
o
v
e
r
e
-
id
en
tific
atio
n
u
n
d
e
r
am
b
ig
u
ity
.
I
n
th
is
p
ap
er
,
we
p
r
o
p
o
s
e
a
n
o
v
el
h
y
b
r
id
f
r
am
ewo
r
k
th
at
in
teg
r
ates
th
e
f
u
z
zy
d
ata
ass
o
ciatio
n
en
g
in
e
f
r
o
m
L
i
an
d
Z
h
a
n
[
2
2
]
with
a
d
ee
p
co
u
p
le
d
R
esNe
t
f
o
r
f
ea
tu
r
e
ex
tr
ac
ti
o
n
an
d
a
GA
f
o
r
o
p
tim
izin
g
tr
ac
k
in
g
d
ec
is
io
n
s
.
Ou
r
p
r
im
a
r
y
co
n
tr
ib
u
tio
n
lies
in
th
e
s
y
n
er
g
is
t
ic
co
m
b
in
atio
n
o
f
th
ese
th
r
ee
co
m
p
o
n
en
ts
.
W
h
ile
th
e
wo
r
k
o
f
L
i
an
d
Z
h
an
[
2
2
]
p
r
o
v
i
d
es
th
e
co
r
e
f
u
zz
y
lo
g
ic
f
o
r
ass
o
ciatio
n
,
o
u
r
f
r
am
ewo
r
k
e
x
ten
d
s
it
b
y
f
ee
d
in
g
th
e
s
y
s
tem
r
ich
,
d
i
s
cr
im
in
ativ
e
f
ea
tu
r
es
f
r
o
m
a
d
ee
p
n
etwo
r
k
an
d
th
en
u
s
in
g
a
GA
to
d
y
n
am
ically
r
ef
in
e
an
d
o
p
tim
ize
th
e
tr
ac
k
in
g
h
y
p
o
th
eses
g
e
n
er
ated
b
y
th
e
f
u
zz
y
s
y
s
tem
.
T
h
e
in
n
o
v
ativ
e
asp
ec
t
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
is
illu
s
tr
ated
b
y
th
r
ee
p
r
in
cip
al
d
esig
n
elem
en
ts
:
i)
T
h
e
h
y
b
r
id
s
y
s
tem
co
m
b
in
e
s
d
ee
p
c
o
u
p
le
d
R
esNet
em
b
ed
d
in
g
s
[
5
]
with
a
GA
,
m
o
d
u
lated
b
y
a
Ma
m
d
an
i
f
u
zz
y
in
f
e
r
en
ce
s
y
s
tem
[
2
2
]
.
T
h
is
u
n
iq
u
e
f
e
n
ce
en
ab
les
s
im
u
ltan
eo
u
s
d
ee
p
f
ea
tu
r
e
lear
n
in
g
,
wh
ich
is
g
r
o
u
n
d
ed
i
n
s
to
ch
asti
city
,
with
a
s
to
ch
asti
c,
ev
o
lu
ti
o
n
-
b
ased
s
ea
r
ch
o
p
tim
izatio
n
.
ii)
T
h
er
e
is
a
f
ee
d
b
ac
k
-
d
r
iv
en
ad
ap
tatio
n
m
ec
h
an
is
m
,
wh
er
eb
y
f
u
zz
y
co
n
f
id
en
ce
o
u
tp
u
ts
d
y
n
a
m
ically
ad
ap
t
GA
p
ar
am
eter
s
(
e.
g
.
s
elec
tio
n
p
r
ess
u
r
e
an
d
m
u
tatio
n
r
at
e)
in
a
c
o
n
tex
t
-
s
p
ec
if
ic
m
a
n
n
er
d
u
r
in
g
a
r
ea
l
-
tim
e
ad
ap
tatio
n
t
o
o
cc
lu
s
i
o
n
s
ev
er
ity
,
d
etec
tio
n
co
n
f
id
e
n
ce
,
an
d
ap
p
ea
r
a
n
ce
s
im
ilar
ity
.
iii)
I
n
co
r
p
o
r
atin
g
p
r
o
b
ab
ilis
tic
d
ata
ass
o
ciatio
n
f
ilter
s
(
PDA
F)
,
th
e
co
n
d
itio
n
al
jo
in
t
lik
elih
o
o
d
f
ilter
s
(
C
J
L
F)
im
p
r
o
v
e
tem
p
o
r
al
co
n
s
is
ten
cy
ac
r
o
s
s
f
r
am
e
s
am
p
lin
g
r
ates
as
it
r
etain
s
id
e
n
tity
co
n
s
tan
cy
wh
ile
m
in
im
izin
g
d
r
i
f
t u
n
d
er
v
a
r
y
in
g
ap
p
ea
r
a
n
ce
s
an
d
w
h
en
o
v
er
l
ap
p
in
g
with
o
th
er
tar
g
ets.
T
o
g
eth
er
,
th
ese
th
r
ee
c
o
m
p
o
n
en
ts
co
n
tr
i
b
u
te
to
a
r
o
b
u
s
t
an
d
s
ca
lab
le
MFT
f
r
a
m
ewo
r
k
,
p
r
o
v
id
in
g
en
h
an
ce
d
ca
p
ab
ilit
y
to
m
ai
n
tain
h
ig
h
ac
c
u
r
ac
y
a
n
d
s
tab
ilit
y
in
d
y
n
am
ic
a
n
d
s
o
m
etim
es
ch
allen
g
in
g
r
ea
l
-
wo
r
ld
co
n
d
itio
n
s
.
2.
M
E
T
H
O
D
2
.
1
.
P
r
o
po
s
ed
f
ra
m
ewo
r
k
a
r
chit
ec
t
ure
T
h
e
s
u
g
g
ested
MFT
s
y
s
te
m
in
clu
d
es
f
o
u
r
p
r
i
m
ar
y
p
r
o
ce
s
s
in
g
s
tag
es:
f
ac
e
d
ete
ctio
n
an
d
s
eg
m
en
tatio
n
,
d
ee
p
co
u
p
led
R
esNet
-
b
ased
f
ea
tu
r
e
ex
tr
ac
tio
n
,
f
u
zz
y
g
en
etic
o
p
tim
izatio
n
,
an
d
p
r
o
b
a
b
ilis
tic
d
ata
ass
o
ciatio
n
.
T
h
e
s
tag
es,
illu
s
tr
ated
in
Fig
u
r
e
1
,
p
r
o
ce
s
s
in
co
m
in
g
v
id
eo
f
r
am
es
th
r
o
u
g
h
th
e
v
ar
io
u
s
co
m
p
o
n
en
ts
,
p
r
o
d
u
cin
g
s
tab
le
id
en
tity
tr
ac
k
in
g
o
u
t
p
u
ts
.
Face
d
etec
tio
n
o
cc
u
r
s
f
ir
s
t
th
r
o
u
g
h
HSV
-
b
ased
s
k
in
co
lo
r
class
if
icatio
n
an
d
th
en
o
cc
lu
s
io
n
-
awa
r
e
r
eg
io
n
s
eg
m
en
tatio
n
u
s
in
g
Ma
r
k
o
v
r
an
d
o
m
f
ield
s
(
MRF
)
.
T
h
e
s
eg
m
en
ted
r
eg
io
n
s
r
ep
r
esen
tin
g
th
e
d
etec
ted
s
u
b
jects’
f
ac
es
ar
e
en
co
d
ed
u
s
in
g
a
d
ee
p
co
u
p
led
R
esNet
ar
ch
itectu
r
e
tu
n
e
d
f
o
r
lo
w
-
r
es
o
lu
tio
n
s
u
r
v
eillan
ce
v
id
eo
im
a
g
er
y
.
Fig
u
r
e
1
.
Pro
p
o
s
ed
MFT
f
r
am
ewo
r
k
ar
ch
itectu
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
1
,
Ma
r
c
h
2
0
2
6
:
208
-
2
1
7
210
2
.
2
.
Dee
p
co
up
led
ResNet
f
o
r
lo
w
-
re
s
o
lutio
n f
a
ce
re
co
g
nitio
n
T
h
e
d
ee
p
c
o
u
p
le
d
R
esNet
ad
d
r
ess
es
lo
w
-
r
eso
lu
tio
n
f
ac
e
r
ec
o
g
n
itio
n
b
y
jo
in
tly
lear
n
in
g
f
ea
tu
r
e
m
ap
p
in
g
s
b
etwe
en
lo
w
-
an
d
h
ig
h
-
r
eso
lu
tio
n
d
o
m
ain
s
th
r
o
u
g
h
two
c
o
u
p
led
n
etwo
r
k
b
r
an
c
h
es.
T
h
e
lo
w
-
r
eso
lu
tio
n
p
ath
p
r
o
ce
s
s
es
th
e
d
eg
r
ad
ed
in
p
u
t,
wh
ile
th
e
h
ig
h
-
r
eso
lu
tio
n
b
r
a
n
ch
p
r
o
v
id
es
s
u
p
er
v
is
o
r
y
g
u
id
an
ce
f
o
r
f
ea
tu
r
e
alig
n
m
en
t.
T
h
is
co
u
p
led
r
ep
r
es
en
tatio
n
im
p
r
o
v
es
d
is
cr
im
in
a
b
ilit
y
u
n
d
er
q
u
ality
l
o
s
s
b
y
lev
er
ag
in
g
s
tatis
tical
s
k
in
-
co
lo
r
m
o
d
eli
n
g
,
wh
er
e
h
o
m
o
g
en
eo
u
s
r
eg
io
n
s
ar
e
r
ep
r
esen
ted
u
s
in
g
a
3
D
Gau
s
s
ian
PDF in
R
G
B
/HS
V
s
p
ac
e
as in
(
1
)
.
(
)
=
1
(
2
)
3
/
2
|
|
1
/
2
(
1
2
(
1
−
)
−
1
(
1
−
)
)
(
1
)
Her
e,
I
d
en
o
te
th
e
co
lo
r
v
ec
to
r
,
μ
th
e
m
ea
n
,
a
n
d
th
e
co
v
ar
ia
n
ce
m
atr
ix
,
an
d
b
in
ar
y
s
k
in
m
ask
s
ar
e
g
en
er
ated
v
ia
ad
ap
tiv
e
th
r
esh
o
l
d
in
g
o
f
th
e
Gau
s
s
ian
PDF.
Fig
u
r
e
2
s
h
o
ws
th
e
d
ee
p
co
u
p
led
R
esNet
ar
ch
itectu
r
e,
co
n
s
is
tin
g
o
f
a
tr
u
n
k
n
etwo
r
k
f
o
r
f
ea
tu
r
e
ex
tr
ac
tio
n
(
wit
h
co
n
v
o
lu
tio
n
al
lay
er
s
o
f
3
2
-
5
1
2
c
h
an
n
els)
an
d
b
r
an
ch
n
etwo
r
k
s
th
at
co
u
p
l
e
lo
w
-
an
d
h
ig
h
-
r
eso
lu
tio
n
f
ea
tu
r
e
s
p
ac
es.
T
h
e
m
o
d
el
in
teg
r
ates
s
o
f
tm
ax
,
ce
n
ter
,
a
n
d
cr
o
s
s
-
m
o
d
ality
(
CM
)
lo
s
s
es
to
jo
i
n
tly
e
n
f
o
r
ce
cl
ass
s
ep
ar
ab
ilit
y
an
d
cr
o
s
s
-
r
eso
lu
tio
n
c
o
n
s
is
ten
cy
.
R
esid
u
al
co
n
n
ec
tio
n
s
(
1
,
2
,
3
,
5
)
s
u
p
p
o
r
t
g
r
a
d
ien
t
s
tab
ilit
y
an
d
h
ier
ar
ch
ical
f
ea
tu
r
e
r
eu
s
e.
T
h
is
co
u
p
led
d
esig
n
im
p
r
o
v
es
r
e
co
g
n
itio
n
ac
cu
r
ac
y
f
o
r
lo
w
-
r
eso
lu
tio
n
s
u
r
v
e
illan
ce
im
a
g
er
y
co
m
p
ar
ed
to
s
in
g
le
-
p
ath
m
o
d
els.
Fig
u
r
e
2
.
I
ll
u
s
tr
atio
n
o
f
a
d
ee
p
co
u
p
l
ed
R
esNet
ar
ch
itectu
r
e
2
.
3
.
F
uzzy
g
enet
ic
a
lg
o
rit
hm
o
ptim
iza
t
io
n
T
h
e
f
u
zz
y
g
en
etic
o
p
tim
izatio
n
m
o
d
u
le
a
d
d
r
ess
es
lo
ca
l
m
in
im
a
an
d
u
n
ce
r
tain
ty
in
d
ata
a
s
s
o
ciatio
n
b
y
in
c
o
r
p
o
r
atin
g
th
e
Ma
m
d
a
n
i
f
u
zz
y
in
f
e
r
en
ce
f
r
am
ew
o
r
k
o
f
L
i
an
d
Z
h
an
[
2
2
]
,
w
h
ich
e
s
tim
ates
ass
o
ciatio
n
co
n
f
id
en
ce
u
s
in
g
m
o
tio
n
a
n
d
ap
p
ea
r
an
ce
af
f
in
ities
.
W
e
ad
o
p
t
th
e
s
am
e
f
u
zz
y
s
tr
u
ctu
r
e
b
u
t
r
elab
el
th
e
f
iv
e
in
p
u
t
s
ets
to
r
ef
lect
s
em
an
tic
co
n
f
id
en
ce
lev
els
,
e.
g
.
ze
r
o
co
n
f
id
en
ce
(
Z
C
)
an
d
lo
w
co
n
f
id
en
ce
(
L
C
)
.
Ou
r
co
n
tr
ib
u
tio
n
lies
in
in
teg
r
ati
n
g
th
ese
f
u
zz
y
o
u
tp
u
ts
in
to
a
GA
-
b
ased
o
p
tim
izer
t
o
i
m
p
r
o
v
e
ass
ig
n
m
en
t
s
tab
ilit
y
.
T
h
e
GA
u
s
es
elitis
m
with
r
o
u
lette
-
wh
ee
l
s
elec
t
io
n
,
a
cr
o
s
s
o
v
er
p
r
o
b
a
b
ilit
y
o
f
p
c
=0
.
5
,
an
d
a
m
u
tatio
n
p
r
o
b
a
b
ilit
y
o
f
p
m
=
0
.
1
,
with
B
h
attac
h
ar
y
y
a
d
is
tan
ce
u
s
ed
as
th
e
f
itn
ess
m
etr
ic.
A
p
p
ea
r
an
ce
af
f
in
ity
is
co
m
p
u
ted
u
s
in
g
co
lo
r
h
is
to
g
r
am
s
in
R
GB
/HS
V
s
p
ac
e
co
m
b
in
ed
with
lo
ca
l
b
in
ar
y
p
atter
n
(
L
B
P
)
te
x
tu
r
e
f
ea
tu
r
es
to
im
p
r
o
v
e
d
is
cr
im
in
atio
n
u
n
d
er
o
cc
lu
s
io
n
an
d
ill
u
m
in
atio
n
ch
an
g
es.
Un
lik
e
e
ar
lier
m
eth
o
d
s
th
at
r
elied
s
o
lely
o
n
h
a
n
d
cr
a
f
ted
h
is
to
g
r
am
s
,
th
is
h
y
b
r
i
d
r
ep
r
esen
tatio
n
s
tr
en
g
th
en
s
ass
o
ciatio
n
r
eliab
ilit
y
in
clu
tter
ed
en
v
ir
o
n
m
e
n
ts
b
y
jo
i
n
tly
ex
p
lo
itin
g
co
lo
r
an
d
tex
t
u
r
e
cu
es.
As
a
r
esu
lt,
th
e
f
r
a
m
ewo
r
k
m
ain
tai
n
s
r
o
b
u
s
t
tr
ac
k
let
co
n
tin
u
ity
d
i
r
ec
tly
f
r
o
m
r
aw
v
id
eo
with
o
u
t
r
eq
u
i
r
in
g
p
r
e
-
f
ilter
ed
d
e
tectio
n
s
o
r
m
an
u
al
f
alse
-
p
o
s
itiv
e
r
em
o
v
al
.
2
.
3
.
1
.
M
em
bersh
ip
f
un
ct
io
n
s
T
h
e
r
u
les
o
f
t
h
e
f
u
zz
y
in
f
er
e
n
ce
s
y
s
tem
f
o
r
f
u
zz
y
weig
h
t
.
I
n
th
is
p
ap
er
,
two
in
p
u
ts
an
d
o
n
e
o
u
tp
u
t a
r
e
r
ep
r
esen
te
d
as sh
o
wn
in
Fig
u
r
e
3
.
T
wo
in
p
u
t v
a
r
iab
les:
i)
=
(
,
)
,
=
(
,
)
,
=
(
,
)
m
ea
n
s
m
em
b
er
s
h
ip
d
en
o
t
e
th
e
m
o
tio
n
,
s
h
ap
e,
an
d
ap
p
ea
r
an
ce
a
f
f
in
ities
b
etwe
en
o
b
ject
f
ac
es
i
an
d
o
b
s
er
v
atio
n
j
,
r
esp
ec
tiv
ely
m
ea
n
s
m
em
b
er
s
h
ip
d
en
o
te
th
e
m
o
tio
n
,
s
h
ap
e,
an
d
ap
p
ea
r
an
ce
af
f
in
ities
b
etwe
en
o
b
ject
f
ac
es
i
an
d
o
b
s
er
v
atio
n
j
,
r
esp
ec
tiv
ely
.
ii)
̂
.
̂
,
̂
m
ea
n
s
n
o
n
-
m
em
b
e
r
s
h
ip
.
T
h
e
s
h
ap
e
af
f
in
ity
(
i
,
j
)
.
B
etwe
en
o
b
ject
i
an
d
o
b
s
er
v
atio
n
is
d
ef
in
ed
as (
2
)
.
(
−
(
ℎ
−
ℎ
)
2
2
2
+
(
−
)
2
2
2
)
(
2
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
o
b
u
s
t m
u
lti
-
fa
ce
s
r
ec
o
g
n
itio
n
a
n
d
tr
a
ck
in
g
via
f
u
z
z
y
g
en
etic
a
lg
o
r
ith
ms a
n
d
…
(
A
d
il A
b
d
u
lh
u
r
A
b
u
s
h
a
n
a
)
211
W
h
er
e
ℎ
an
d
ℎ
d
en
o
te
th
e
h
eig
h
t
s
o
f
o
b
ject
i
an
d
o
b
s
er
v
atio
n
,
r
esp
ec
tiv
ely
,
an
d
d
en
o
te
t
h
e
wid
th
s
o
f
o
b
ject
i
an
d
o
b
s
er
v
a
tio
n
,
r
esp
ec
tiv
ely
,
2
,
2
d
en
o
te
th
e
v
ar
ian
ce
f
o
r
th
e
h
eig
h
t
an
d
wid
th
,
r
esp
ec
tiv
ely
.
T
h
e
a
f
f
in
ity
b
et
wee
n
th
e
p
r
e
d
icted
s
tate
o
f
f
ac
e
i
an
d
o
b
s
er
v
atio
n
j
is
n
o
r
m
alize
d
to
a
v
alu
e
b
etwe
en
0
an
d
1
.
T
h
ese
n
o
r
m
alize
d
v
alu
es
ar
e
th
e
n
m
ap
p
ed
t
o
co
r
r
esp
o
n
d
i
n
g
f
u
z
zy
s
ets
with
in
th
e
f
u
zz
y
i
n
f
er
en
ce
s
y
s
tem
.
I
n
g
en
er
al,
in
cr
ea
s
in
g
th
e
n
u
m
b
er
o
f
f
u
zz
y
s
ets
ca
n
lea
d
to
h
ig
h
e
r
ac
cu
r
ac
y
,
th
o
u
g
h
it
also
r
aises
co
m
p
u
tatio
n
al
co
m
p
le
x
ity
.
T
h
er
ef
o
r
e
,
th
e
n
u
m
b
er
o
f
f
u
zz
y
s
ets
is
o
f
t
e
n
d
eter
m
in
ed
em
p
ir
ically
b
ased
o
n
th
e
tr
a
d
e
-
o
f
f
b
etwe
en
p
r
ec
i
s
io
n
an
d
ef
f
icien
cy
.
Fig
u
r
e
3
.
C
o
n
f
id
en
ce
m
e
m
b
er
s
h
ip
f
u
n
ctio
n
,
ad
a
p
ted
f
r
o
m
L
i
an
d
Z
h
a
n
[
2
2
]
I
n
th
is
p
ap
e
r
,
we
ch
o
o
s
e
f
iv
e
f
u
zz
y
s
ets
to
d
escr
ib
e
af
f
in
it
y
in
th
e
f
u
zz
y
i
n
f
er
en
ce
s
y
s
tem
,
wh
er
e
ea
ch
ex
p
licit
in
p
u
t
d
ata
(
=
(
i
,
j
)
,
=
(
i
,
j
)
,
=
(
i
,
j
)
)
is
ca
teg
o
r
ized
in
to
Z
C
,
L
C
,
m
ed
iu
m
co
n
f
id
en
ce
(
MC),
h
ig
h
c
o
n
f
id
en
ce
(
HC
)
,
an
d
v
er
y
h
ig
h
co
n
f
id
en
ce
(
VHC);
f
ea
t
u
r
e
af
f
in
ity
v
alu
es
less
th
a
n
o
r
eq
u
al
to
0
.
1
i
n
d
icate
u
n
r
eliab
l
e
f
ac
e
f
ea
tu
r
es,
wh
ile
v
alu
es
g
r
ea
ter
th
a
n
o
r
eq
u
al
t
o
0
.
9
s
ig
n
if
y
v
er
y
r
eliab
le
f
ea
tu
r
es.
C
o
n
s
eq
u
e
n
tly
,
ea
ch
f
u
zz
y
r
u
le
i
n
T
a
b
les
1
to
3
u
tili
ze
s
co
n
f
id
en
c
e
lev
els
f
r
o
m
a
p
p
ea
r
an
ce
,
m
o
tio
n
,
an
d
s
h
ap
e
to
m
an
ag
e
o
b
ject
m
er
g
in
g
,
s
p
litt
in
g
,
an
d
o
cc
lu
s
io
n
h
an
d
lin
g
ef
f
ec
tiv
ely
.
A
d
r
o
p
in
m
o
tio
n
af
f
in
ity
u
n
d
er
th
r
esh
o
l
d
α
r
ed
u
ce
s
th
e
in
f
lu
en
ce
o
f
a
p
p
ea
r
a
n
ce
af
f
in
ity
,
m
itig
atin
g
f
alse
o
b
s
er
v
ati
o
n
s
,
an
d
ca
u
s
es
all
ap
p
ea
r
an
ce
weig
h
ts
.
b
ein
g
s
et
to
VHC.
T
ab
le
1
.
Fu
zz
y
r
u
les b
ase
wei
g
h
t
M
VHC
HC
MC
LC
ZC
VHC
HC
HC
VHC
VHC
ZC
̂
M
VHC
HC
HC
HC
VHC
LC
UK
VHC
MC
HC
VHC
MC
UK
VHC
MC
MC
VHC
HC
T
ab
le
2
.
Fu
zz
y
r
u
les b
ase
wei
g
h
t
μ
ij
S
W
M
K
VHC
HC
MC
LC
ZC
LC
LC
ZC
ZC
ZC
ZC
μ
̂
ij
S
MC
LC
LC
ZC
ZC
LC
HC
MC
MC
LC
ZC
MC
VHC
HC
MC
MC
MC
HC
UK
HC
HC
HC
HC
VHC
T
ab
le
3
.
Fu
zz
y
r
u
les b
ase
wei
g
ht
A
VHC
HC
MC
LC
ZC
MC
MC
LC
LC
ZC
ZC
̂
A
HC
HC
MC
MC
LC
LC
VHC
HC
MC
MC
LC
MC
UK
VHC
HC
MC
HC
HC
UK
VHC
HC
HC
VHC
VHC
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
1
,
Ma
r
c
h
2
0
2
6
:
208
-
2
1
7
212
T
h
e
lin
g
u
is
tic
v
ar
iab
les
h
a
v
e
b
ee
n
r
elab
eled
to
r
e
p
r
esen
t
co
n
f
id
en
ce
lev
els
in
o
u
r
f
r
am
ew
o
r
k
.
W
h
en
p
r
ed
ictio
n
s
ar
e
ac
cu
r
ate,
t
h
e
m
o
tio
n
af
f
in
ity
f
o
r
ea
c
h
f
ac
e
g
ain
s
im
p
o
r
tan
ce
,
an
d
th
e
weig
h
t
o
f
th
e
ap
p
ea
r
an
ce
af
f
in
ity
s
h
o
u
ld
in
cr
ea
s
e
as
̂
.
R
is
es,
with
th
e
f
u
zz
y
r
u
les
i
n
t
h
e
f
o
u
r
th
co
lu
m
n
ad
ju
s
ted
to
L
C
,
V
HC
,
an
d
u
n
k
n
o
wn
(
UK)
,
r
esp
ec
tiv
ely
.
T
h
e
f
ir
s
t
an
d
s
ec
o
n
d
f
u
zz
y
r
u
les
in
th
e
f
if
th
co
l
u
m
n
a
d
d
r
ess
th
e
ch
allen
g
es
p
o
s
ed
b
y
o
cc
lu
d
e
d
f
ac
es
o
r
clu
tter
ed
en
v
ir
o
n
m
e
n
ts
,
wh
er
e
d
is
tin
g
u
is
h
in
g
d
if
f
er
en
ce
s
i
n
th
eir
ap
p
ea
r
an
ce
s
b
ec
o
m
es d
if
f
icu
lt; th
u
s
,
th
e
w
eig
h
ts
.
th
ey
ar
e
s
et
to
VHC,
wh
ile
o
th
er
r
u
les ar
e
d
esig
n
ated
as UK
.
Ad
d
itio
n
ally
,
t
h
e
f
u
zz
y
r
u
les
in
th
e
s
ec
o
n
d
an
d
th
ir
d
c
o
lu
m
n
s
ty
p
ically
m
an
a
g
e
s
ce
n
a
r
io
s
wh
er
e
p
r
ed
ictio
n
p
o
s
itio
n
s
f
o
r
m
u
ltip
le
f
ac
es
lack
ac
cu
r
ac
y
.
As
i
n
c
r
ea
s
es
th
e
ap
p
ea
r
an
ce
a
f
f
in
ity
,
in
im
p
o
r
tan
ce
,
ad
ju
s
tin
g
.
to
L
C
,
MC,
an
d
Z
C
,
r
esp
ec
tiv
ely
.
I
n
T
ab
les
2
an
d
3
,
th
e
f
o
u
r
th
an
d
f
if
th
f
u
z
zy
r
u
les
in
th
e
f
ir
s
t
co
lu
m
n
ad
d
r
ess
o
cc
lu
s
io
n
s
,
em
p
h
asizin
g
ap
p
ea
r
an
c
e
af
f
in
ity
wh
en
o
b
ject
p
o
s
i
tio
n
s
ar
e
clo
s
e
to
o
b
s
er
v
atio
n
s
,
with
weig
h
ts
a
n
d
s
et
to
HC
an
d
VHC,
r
esp
ec
tiv
ely
,
wh
ile
th
e
a
p
p
ea
r
an
c
e
af
f
in
ity
in
cr
ea
s
es a
s
,
.
An
u
p
war
d
tr
en
d
in
tr
u
e
p
o
s
it
iv
es
(
T
P)
an
d
tr
u
e
n
eg
ativ
es
(
T
N)
,
alo
n
g
with
r
ed
u
ce
d
f
alse
p
o
s
itiv
es
(
FP
)
an
d
f
alse
n
eg
ativ
es
(
FN)
,
r
ef
lects
im
p
r
o
v
ed
id
e
n
tific
atio
n
r
eliab
ilit
y
o
v
e
r
tim
e.
As
s
h
o
wn
in
T
ab
le
1
,
th
e
p
r
o
p
o
s
ed
m
eth
o
d
o
u
t
p
er
f
o
r
m
s
alter
n
atin
g
d
ir
ec
tio
n
m
et
h
o
d
o
f
m
u
ltip
lier
s
(
ADM
M
)
ac
r
o
s
s
r
ec
all,
p
r
ec
is
io
n
,
F1
-
s
co
r
e
,
m
u
ltip
le
o
b
jects
tr
a
ck
in
g
ac
c
u
r
ac
y
(
MO
T
A)
,
a
n
d
m
u
ltip
le
o
b
je
cts
tr
ac
k
i
n
g
p
r
ec
is
io
n
(
MO
T
P)
.
T
h
e
MO
T
A
-
b
ased
ev
alu
atio
n
,
w
h
ich
in
co
r
p
o
r
ates
m
o
s
t
tr
ac
k
ed
(
MT
)
,
m
o
s
t
lo
s
t
(
ML
)
,
f
r
ag
m
en
tatio
n
(
FG)
,
an
d
b
o
u
n
d
in
g
-
b
o
x
o
v
er
la
p
v
ia
MO
T
P,
co
n
f
ir
m
s
s
u
p
er
i
o
r
tr
ac
k
in
g
ef
f
ec
tiv
en
ess
f
o
r
m
u
ltip
le
f
ac
es
as in
(
3
)
.
MO
T
P
=
∑
∑
(
3
)
W
h
er
e
r
ep
r
esen
ts
th
e
to
tal
n
u
m
b
er
o
f
ass
o
ciate
d
o
b
jects a
t th
e
tim
e.
T
h
ese
eq
u
atio
n
s
,
as
d
ef
in
ed
b
y
MO
T
A
an
d
MO
T
P,
p
r
o
v
id
e
m
ath
em
atics f
o
r
ev
al
u
atin
g
tr
ac
k
in
g
p
e
r
f
o
r
m
an
ce
.
2
.
3
.
2
.
F
uzzy
s
y
s
t
em
(
inp
uts,
o
utput
s
,
a
nd
rules)
L
in
g
u
is
tic
v
ar
iab
les an
d
m
em
b
er
s
h
ip
f
u
n
ctio
n
s
in
clu
d
e:
i)
I
n
p
u
ts
(
p
e
r
ass
o
ciatio
n
h
y
p
o
th
esis
b
etwe
en
tr
ac
k
an
d
d
etec
tio
n
:
–
Occ
=
o
cc
lu
s
io
n
lev
el
∈
[
0
,
1
]
MF: {
lo
w,
m
ed
,
h
ig
h
}
v
ia
tr
ia
n
g
u
lar
/tra
p
ez
o
id
al
s
ets
lo
w
:
[
0
,
0
,
0
.
3
]
,
m
ed
:
[
0
.
2
,
0
.
5
,
0
.
8
]
,
h
ig
h
:
[
0
.
6
,
1
,
1
]
.
–
Sim
=
ap
p
ea
r
an
ce
s
im
ilar
ity
(
d
ee
p
-
co
u
p
led
c
o
s
in
e)
∈
[
0
,
1
]
MF: {
lo
w,
m
ed
,
h
ig
h
}
→
lo
w
:
[
0
,
0
,
0
.
4
]
,
m
ed
:
[
0
.
3
,
0
.
6
,
0
.
8
]
,
h
ig
h
:
[
0
.
7
,
1
,
1
]
.
–
C
o
n
f
=
d
etec
to
r
c
o
n
f
i
d
en
ce
∈
[
0
,
1
]
MF: {
lo
w,
m
ed
,
h
ig
h
}
→
lo
w
:
[
0
,
0
,
0
.
4
]
,
m
ed
:
[
0
.
3
,
0
.
6
,
0
.
85
]
,
h
ig
h
:
[
0
.
75
,
1
,
1
]
.
–
Mo
t =
m
o
tio
n
c
o
n
s
is
ten
cy
(
M
ah
alan
o
b
is
/KF r
esid
u
al
n
o
r
m
a
lized
)
∈
[
0
,
1
]
,
h
ig
h
er
is
b
etter
MF: {
p
o
o
r
,
f
air
,
g
o
o
d
}
→
p
o
o
r
:
[
0
,
0
,
0
.
4
]
,
f
air
:
[
0
.
3
,
0
.
6
,
0
.
85
]
,
g
o
o
d
:
[
0
.
75
,
1
,
1
]
.
ii)
Ou
tp
u
ts
(
d
ef
u
zz
i
f
ied
b
y
ce
n
tr
o
id
)
:
–
_
ap
p
∈
[
0
,
1
]
—
weig
h
t f
o
r
a
p
p
ea
r
a
n
c
e
ter
m
in
ass
o
ciatio
n
.
–
_
m
o
t
∈
[
0
,
1
]
—
weig
h
t f
o
r
m
o
tio
n
te
r
m
(
en
f
o
r
ce
α
_
ap
p
+
α
_
m
o
t =
1
af
ter
d
e
f
u
zz
)
.
–
_
ass
o
c
∈
[
0
,
1
]
—
ad
ap
tiv
e
ass
o
ciatio
n
th
r
esh
o
ld
.
–
GA
h
y
p
er
p
a
r
am
eter
s
p
er
g
en
e
r
atio
n
:
_
∈
[
0
.
01
,
0
.
3
]
,
_
∈
[
0
.
6
,
0
.
95
]
,
_
∈
[
1
.
2
,
2
.
0
]
.
2
.
3
.
3
.
Six
co
re
f
uzzy
rules
Use M
am
d
an
i r
u
les;
a
co
m
p
ac
t,
h
ig
h
-
im
p
ac
t su
b
s
et:
−
R
1
:
ℎ
ℎ
ℎ
,
,
ℎ
,
,
ℎ
.
−
R
2
:
ℎ
,
,
,
↑
,
ℎ
(
)
.
−
R
3
:
,
ℎ
,
−
ℎ
.
−
R
4
:
,
,
,
ℎ
(
)
.
−
R
5
:
ℎ
ℎ
,
,
−
ℎ
.
−
R
6
:
ℎ
,
−
ℎ
,
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
o
b
u
s
t m
u
lti
-
fa
ce
s
r
ec
o
g
n
itio
n
a
n
d
tr
a
ck
in
g
via
f
u
z
z
y
g
en
etic
a
lg
o
r
ith
ms a
n
d
…
(
A
d
il A
b
d
u
lh
u
r
A
b
u
s
h
a
n
a
)
213
2
.
3
.
4
.
Chro
m
o
s
o
m
e
enco
din
g
E
ac
h
ch
r
o
m
o
s
o
m
e
e
n
co
d
es p
e
r
-
fr
am
e
ass
o
ciatio
n
a
n
d
g
l
o
b
al
k
n
o
b
s
:
i)
Glo
b
al
g
en
es:
−
∈
[
0
,
1
]
: p
r
io
r
weig
h
t o
n
ap
p
ea
r
a
n
ce
(
b
ef
o
r
e
f
u
zz
y
a
d
ju
s
tm
en
t)
.
−
=
1
−
.
−
∈
[
0
,
1
]
: b
ase
ass
o
ciatio
n
th
r
esh
o
ld
.
−
∈
[
0
,
1
]
: I
o
U
g
atin
g
weig
h
t
.
−
ℎ
∈
[
0
,
1
]
:
tr
ajec
to
r
y
s
m
o
o
th
n
ess
r
eg
u
la
r
izatio
n
.
−
∈
[
0
,
1
]
: CJ
L
F b
len
d
in
g
with
PDAF
.
ii)
Per
-
tr
ac
k
g
en
es (
o
p
tio
n
al
c
o
m
p
ac
t f
o
r
m
u
s
in
g
s
h
ar
e
d
p
ar
a
m
s
b
y
clu
s
ter
s
)
:
−
∈
[
0
.
1
,
0
.
9
]
: g
atin
g
r
ad
i
u
s
s
ca
lin
g
.
−
∈
{
1
,
…
,
}
: m
em
o
r
y
le
n
g
th
f
o
r
f
ea
tu
r
e
g
a
ller
y
.
iii)
Ass
o
ciatio
n
g
en
es
(
f
o
r
to
p
-
k
ca
n
d
id
ate
p
air
s
p
er
f
r
a
m
e)
:
b
in
ar
y
v
ec
t
o
r
A
with
{
,
}
∈
{
0
,
1
}
u
n
d
e
r
1
-
to
-
1
co
n
s
tr
ain
ts
(
Hu
n
g
ar
ia
n
-
co
m
p
atib
le)
.
I
n
p
r
ac
tice,
GA
s
ea
r
ch
es
o
v
er
th
r
esh
o
l
d
s
/weig
h
ts
;
th
e
f
in
al
1
-
to
-
1
is
p
r
o
d
u
ce
d
b
y
Hu
n
g
ar
i
an
o
n
t
h
e
GA
-
weig
h
ted
c
o
s
t m
atr
ix
.
2
.
3
.
5
.
F
it
nes
s
f
un
ct
io
n (
s
ing
l
e
-
o
bje
ct
iv
e,
f
a
s
t
M
O
T
A
pro
x
y
)
Fo
r
a
v
alid
atio
n
c
h
u
n
k
(
e.
g
.
,
2
0
0
-
5
0
0
f
r
am
es),
co
m
p
u
te:
−
Ap
p
ea
r
an
ce
c
o
s
t:
=1
−Sim
(
co
s
in
e)
.
−
Mo
tio
n
co
s
t:
=
n
o
r
m
alize
d
K
F/J
PD
A
r
esid
u
al
.
−
I
o
U
p
en
alty
:
= 1
–
I
o
U
.
Per
h
y
p
o
t
h
esis
co
s
t
as in
(
4
)
.
=
∗
+
∗
+
∗
(
4
)
Af
ter
ass
ig
n
m
en
t (
Hu
n
g
ar
ian
)
,
ac
cu
m
u
late:
−
FN,
FP
,
I
DS,
Fra
g
(
o
n
lin
e
esti
m
ates)
.
−
Sm
o
o
th
n
ess
:
=
|
−
{
−
1
}
|
o
v
er
tr
a
ck
s
(
v
=
v
elo
city
)
.
−
R
u
n
tim
e
p
r
o
x
y
: R
=
#
o
p
s
/f
r
a
m
e
(
esti
m
ated
f
r
o
m
ac
tiv
e
tr
ac
k
s
,
g
aller
y
s
ize)
.
Fit
n
ess
to
m
ax
im
ize
(
co
n
v
er
t t
o
m
in
im
izatio
n
as n
ee
d
ed
)
as i
n
(
5
)
.
=
1
×
(
1
−
)
+
2
×
(
1
−
)
+
3
×
(
1
−
)
+
4
×
(
1
−
)
+
5
×
−
6
×
−
7
×
(
5
)
T
y
p
ical
weig
h
ts
as in
(
6
)
.
1
=
0
.
25
,
2
=
0
.
15
,
3
=
0
.
2
,
4
=
0
.
15
,
5
=
0
.
15
,
6
=
0
.
05
,
7
=
0
.
05
(
6
)
2
.
4
.
P
r
o
ba
bil
is
t
ic
da
t
a
a
s
s
o
cia
t
io
n f
ilte
r
(
P
DAF
/J
P
DAF
)
T
h
e
alg
o
r
ith
m
m
e
r
g
es
b
o
th
t
h
e
PDAF
an
d
th
e
jo
in
t
p
r
o
b
a
b
i
lis
tic
d
ata
ass
o
ciatio
n
f
ilter
(
J
PDAF)
f
o
r
m
u
lti
-
tar
g
et
en
v
i
r
o
n
m
e
n
t
s
tab
le
d
ata
ass
o
ciatio
n
.
PDAF
is
ap
p
lied
f
o
r
tr
ac
k
i
n
g
a
s
in
g
le
tar
g
et
with
clu
tter
,
wh
ile
J
PD
AF
i
s
th
e
m
u
lti
-
tar
g
et
ex
ten
s
io
n
with
p
o
ten
tial
in
ter
ac
tio
n
s
b
etwe
en
ta
r
g
ets.
Ass
o
ciatio
n
p
r
o
b
a
b
ilit
ies ar
e
co
m
p
u
ted
as
(
7
)
.
(
∣
∣
)
=
∑
∑
(
;
,
)
=
1
(
7
)
W
h
er
e
is
th
e
p
r
o
b
a
b
ilit
y
th
at
m
ea
s
u
r
em
en
t
i
is
o
f
th
e
tar
g
et
o
b
ject,
H
is
th
e
o
b
s
er
v
atio
n
m
atr
ix
,
an
d
R
is
th
e
m
ea
s
u
r
em
en
t
n
o
is
e
co
v
ar
ian
c
e.
J
o
in
t
lik
elih
o
o
d
s
ar
e
esti
m
ated
b
y
J
PDAF
f
o
r
m
u
ltip
le
o
v
er
lap
p
in
g
f
ac
es
to
m
ain
tain
tr
ac
k
in
te
g
r
ity
d
u
r
in
g
clo
s
e
in
ter
ac
tio
n
.
2
.
5
.
Co
nd
it
io
na
l
j
o
int
lik
eliho
o
d f
ilte
r
T
h
e
C
J
L
F
co
m
p
o
n
en
t
im
p
r
o
v
es
tr
ac
k
in
g
ac
cu
r
ac
y
b
y
m
o
d
e
lin
g
jo
in
t
o
b
jects
an
d
s
tates
p
r
o
b
ab
ilit
y
o
f
co
r
r
elate
d
o
b
jects
with
co
n
s
tr
ain
ts
o
n
s
p
atial
an
d
tem
p
o
r
al
lo
ca
tio
n
s
.
T
h
e
f
ilter
ad
d
r
es
s
es
th
e
o
cc
lu
s
io
n
an
d
clu
tter
p
r
o
b
lem
s
u
s
in
g
d
ep
th
o
r
d
er
in
g
an
d
v
is
ib
ilit
y
-
co
n
s
tr
ain
ed
tr
ac
k
lik
elih
o
o
d
u
p
d
ate.
T
r
ac
k
er
p
ar
am
eter
s
ar
e
u
p
d
ated
u
p
o
n
o
cc
lu
s
io
n
b
y
m
o
d
if
y
in
g
th
e
s
y
s
tem
u
s
in
g
g
r
ad
ie
n
t
ascen
t
o
p
tim
izatio
n
wit
h
d
er
iv
ativ
e
-
f
r
ee
Po
well'
s
o
p
tim
izatio
n
m
eth
o
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
1
,
Ma
r
c
h
2
0
2
6
:
208
-
2
1
7
214
2
.
6
.
F
e
a
t
ure
ex
t
r
a
ct
io
n a
nd
s
k
in det
ec
t
io
n
T
h
e
s
y
s
tem
u
tili
ze
d
m
u
lti
-
m
o
d
al
f
ea
tu
r
e
ex
tr
ac
tio
n
b
ased
o
n
co
lo
r
,
tex
tu
r
e,
a
n
d
s
h
ap
e
f
ea
tu
r
es.
Sk
in
p
ix
el
class
if
icatio
n
is
d
o
n
e
u
s
in
g
a
p
r
o
b
ab
ilis
tic
m
o
d
el
a
s
in
(
8
)
.
(
s
k
in
∣
,
,
)
=
(
,
,
∣
∣
s
k
in
)
⋅
(
sk
in
)
(
,
,
)
(
8
)
W
h
er
e
C
,
T
,
an
d
S
d
en
o
te
co
lo
r
,
tex
tu
r
e
,
an
d
s
h
ap
e
f
ea
tu
r
es,
r
esp
ec
tiv
ely
.
L
B
P
d
escr
ip
to
r
s
ar
e
u
s
ed
t
o
ca
p
tu
r
e
tex
tu
r
e
d
etails,
an
d
HSV
co
lo
r
h
is
to
g
r
am
s
ar
e
u
s
ed
f
o
r
in
s
en
s
itiv
ity
to
co
lo
r
r
ep
r
esen
tatio
n
u
n
d
er
ch
a
n
g
in
g
illu
m
in
atio
n
co
n
d
itio
n
s
.
2
.
7
.
E
v
a
lua
t
i
o
n
m
et
rics
Per
f
o
r
m
an
ce
ass
ess
m
en
t
em
p
lo
y
s
s
tan
d
ar
d
m
ea
s
u
r
es
f
o
r
m
u
lti
-
o
b
ject
tr
ac
k
in
g
,
i
n
clu
d
in
g
MO
T
A,
MO
T
P,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e.
MO
T
A
co
m
p
u
tes
o
v
er
all
tr
ac
k
in
g
p
r
ec
is
io
n
co
n
s
id
er
in
g
FP
,
FN
,
an
d
id
en
tity
s
witch
es
(
I
DS)
.
M
OT
A
=
1
−
∑
(
FP
+
FN
+
ID
)
∑
(
9
)
W
h
er
e
FP
,
FN,
an
d
I
D
r
ep
r
ese
n
t f
alse p
o
s
itiv
es,
f
alse n
eg
ativ
es,
an
d
I
DS
,
r
esp
ec
tiv
ely
,
an
d
m
ₜ
is
th
e
n
u
m
b
er
o
f
g
r
o
u
n
d
tr
u
th
o
b
jects a
t tim
e
t
.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
E
x
p
er
im
en
ts
wer
e
c
o
n
d
u
cted
o
n
th
e
m
u
s
ic
v
id
eo
d
ataset
in
tr
o
d
u
ce
d
b
y
Z
h
an
g
et
a
l.
[
1
1
]
,
wh
ic
h
co
n
tain
s
2
0
m
a
n
u
ally
an
n
o
tat
ed
m
u
lti
-
f
ac
e
v
i
d
eo
s
eq
u
e
n
ce
s
with
v
ar
iatio
n
s
in
illu
m
in
atio
n
,
o
cc
lu
s
io
n
,
an
d
p
o
s
e.
As
r
ep
o
r
te
d
in
T
a
b
le
4
,
th
e
p
r
o
p
o
s
ed
f
u
zz
y
g
e
n
etic
d
ee
p
-
co
u
p
led
f
r
am
ewo
r
k
a
ch
iev
es
an
av
er
a
g
e
F1
-
s
co
r
e
o
f
8
6
.
1
±
1
.
2
%
an
d
MO
T
A
o
f
6
6
.
5
±
1
.
7
%,
o
u
tp
e
r
f
o
r
m
in
g
p
r
io
r
m
o
d
els
with
lo
w
v
ar
ian
ce
ac
r
o
s
s
s
eq
u
en
ce
s
,
in
d
icatin
g
s
tr
o
n
g
g
en
er
aliza
tio
n
.
T
h
e
s
y
s
tem
a
ls
o
ac
h
iev
es
3
1
.
8
FP
S,
d
em
o
n
s
tr
atin
g
r
ea
l
-
tim
e
ca
p
ab
ilit
y
f
o
r
s
u
r
v
eillan
ce
an
d
ed
g
e
-
b
ased
d
ep
lo
y
m
en
t.
T
ab
le
4
.
Statis
tical
p
er
f
o
r
m
a
n
ce
o
f
th
e
p
r
o
p
o
s
ed
f
u
zz
y
g
e
n
et
ic
d
ee
p
-
co
u
p
led
f
r
am
ewo
r
k
d
u
r
in
g
tr
ain
in
g
an
d
test
in
g
p
h
ases
P
h
a
se
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
s
c
o
r
e
(
%)
Tr
a
c
k
i
n
g
a
c
c
u
r
a
c
y
(
M
O
TA
,
%)
Tr
a
c
k
i
n
g
p
r
e
c
i
si
o
n
(
M
O
TP,
%)
I
D
S
F
r
a
g
me
n
t
a
t
i
o
n
s
(F
G
)
R
u
n
t
i
me
(FPS)
Tr
a
i
n
i
n
g
9
1
.
2
±
0
.
8
8
2
.
3
±
1
.
1
8
6
.
6
±
0
.
9
6
7
.
4
±
1
.
3
8
1
.
2
±
0
.
7
6
4
2
±
28
1
7
0
5
±
45
3
2
.
4
±
1
.
1
Te
st
i
n
g
9
0
.
5
±
1
.
3
8
1
.
0
±
1
.
5
8
6
.
1
±
1
.
2
6
6
.
5
±
1
.
7
8
0
.
3
±
1
.
0
6
6
9
±
31
1
7
4
5
±
52
3
1
.
8
±
1
.
2
T
a
b
l
e
5
a
d
d
i
ti
o
n
a
l
l
y
d
e
m
o
n
s
t
r
a
t
es
t
h
a
t
t
h
e
s
u
g
g
e
s
t
e
d
f
u
z
z
y
g
e
n
e
t
i
c
d
ee
p
-
c
o
u
p
l
e
d
f
r
a
m
e
w
o
r
k
o
u
t
p
e
r
f
o
r
m
s
b
o
t
h
c
o
n
v
e
n
t
i
o
n
a
l
a
n
d
h
y
b
r
i
d
t
r
a
c
k
e
r
s
.
O
n
a
v
e
r
ag
e
,
t
h
i
s
le
a
d
s
t
o
3
-
5
%
h
i
g
h
e
r
F
1
-
s
c
o
r
e
a
n
d
2
-
3
%
h
i
g
h
e
r
m
e
a
n
a
v
e
r
a
g
e
t
r
a
c
k
i
n
g
p
r
e
c
i
s
i
o
n
(
MO
T
A
)
s
c
o
r
es
,
wh
i
l
e
m
a
i
n
ta
i
n
i
n
g
a
r
at
e
>
3
0
FP
S
.
D
e
e
p
SO
R
T
is
d
e
p
e
n
d
e
n
t
o
n
l
y
o
n
m
o
t
i
o
n
a
n
d
r
e
-
I
D
w
e
i
g
h
t
i
n
g
,
w
h
i
c
h
a
r
e
f
i
x
e
d
.
I
n
c
o
n
t
r
a
s
t
,
t
h
e
f
u
z
z
y
-
g
e
n
e
t
i
c
l
a
y
e
r
d
y
n
a
m
i
c
a
l
l
y
c
h
a
n
g
e
s
ass
o
c
i
a
tio
n
c
o
n
f
i
d
e
n
c
e
a
l
o
n
g
d
y
n
a
m
i
c
u
s
e
r
s
'
n
o
ti
o
n
o
f
c
o
n
f
i
d
e
n
c
e
(
t
r
a
c
k
e
r
s
'
r
at
i
o
s
o
f
r
e
s
p
o
n
s
es
)
.
T
h
e
f
u
z
z
y
g
e
n
e
t
ic
_
d
e
e
p
c
o
u
p
l
i
n
g
'
s
g
o
a
l
i
s
t
o
r
e
d
u
c
e
I
D
s
w
i
tc
h
e
s
a
n
d
FG
w
h
e
n
o
c
c
l
u
s
i
o
n
s
o
c
c
u
r
.
A
s
c
o
m
p
a
r
e
d
t
o
R
e
t
i
n
aF
a
c
e+
K
a
lm
a
n
[
2
3
]
,
t
h
e
p
r
o
p
o
s
e
d
m
e
t
h
o
d
'
s
d
e
e
p
c
o
u
p
l
e
d
t
r
a
c
k
e
r
d
ev
e
l
o
p
s
m
o
r
e
s
t
a
b
l
e
t
r
a
c
k
i
n
g
p
e
r
f
o
r
m
a
n
c
e
b
e
c
a
u
s
e
i
t
c
o
u
p
l
e
s
t
h
e
d
e
e
p
R
es
N
et
e
m
b
e
d
d
i
n
g
s
w
i
t
h
f
u
z
z
y
o
p
t
im
i
z
a
t
i
o
n
t
o
en
s
u
r
e
c
o
n
t
e
x
t
-
a
w
a
r
e
as
s
o
ci
a
ti
o
n
s
.
T
a
b
l
e
6
r
e
v
e
a
ls
t
h
at
t
h
e
f
u
z
z
y
g
e
n
e
t
i
c
d
e
e
p
-
c
o
u
p
l
e
d
f
r
a
m
ewo
r
k
p
r
o
p
o
s
e
d
i
n
t
h
i
s
p
a
p
e
r
o
u
t
p
e
r
f
o
r
m
s
r
e
c
e
n
t
M
F
T
te
c
h
n
i
q
u
e
s
wi
t
h
th
e
h
i
g
h
e
s
t
F
1
-
s
c
o
r
e
(
8
6
.
1
%
)
,
a
n
d
a
c
o
m
p
e
t
i
t
i
v
e
M
OT
A
o
f
6
6
.
5
%
,
e
x
c
e
e
d
i
n
g
r
e
s
u
l
ts
o
f
d
e
e
p
m
e
t
r
ic
-
l
e
a
r
n
i
n
g
b
a
s
e
li
n
e
s
(
s
ia
m
e
s
e
,
t
r
i
p
l
e
t
,
an
d
S
y
m
T
r
i
p
l
e
t
)
a
n
d
o
p
t
i
m
i
z
a
tio
n
-
b
a
s
e
d
m
o
d
e
l
i
n
g
r
e
s
u
l
ts
s
u
c
h
a
s
AD
M
M
[
2
4
]
a
n
d
i
t
e
r
a
t
i
v
e
H
a
n
k
e
l
t
o
t
al
l
e
as
t
s
q
u
a
r
e
s
(
I
H
T
L
S
)
[
2
5
]
.
T
h
e
l
o
w
f
r
e
q
u
e
n
c
y
o
f
f
a
l
s
e
a
l
a
r
m
s
(
FA
F
=
0
.
1
5
)
a
l
s
o
i
n
d
i
ca
t
e
s
r
e
li
a
b
l
e
t
r
a
c
k
c
o
n
t
i
n
u
it
i
es
i
n
c
h
a
o
t
i
c
c
o
n
d
i
t
i
o
n
s
.
W
h
i
l
e
als
o
b
u
i
l
t
o
n
t
h
e
s
a
m
e
d
a
t
a
s
et
,
t
h
e
p
r
o
b
a
b
i
l
is
t
i
c
i
n
te
g
r
a
t
e
d
t
r
a
c
k
i
n
g
a
n
d
d
e
t
ec
t
i
o
n
f
r
am
e
w
o
r
k
(
P
I
T
AD
F
)
[
2
6
]
r
e
p
o
r
te
d
F
1
-
s
c
o
r
e
=
8
5
.
3
%
a
n
d
M
O
T
A
=
6
9
.
2
%
.
T
h
e
p
r
o
p
o
s
e
d
a
p
p
r
o
a
c
h
a
c
h
i
e
v
e
d
a
h
ig
h
e
r
p
r
e
c
i
s
i
o
n
(
9
0
.
5
%
)
w
it
h
s
i
g
n
i
f
i
c
a
n
t
l
y
f
e
w
e
r
I
D
S
=
6
6
9
a
n
d
FG
=
1
7
4
5
,
w
h
i
c
h
i
s
a
r
e
s
u
l
t
o
f
t
h
e
s
t
a
b
i
l
it
y
p
r
o
v
i
d
e
d
b
y
t
h
e
f
u
z
z
y
r
u
l
e
–
g
u
i
d
e
d
G
A
o
p
t
i
m
i
z
e
r
.
T
h
e
d
e
e
p
c
o
u
p
l
e
d
R
e
s
N
et
e
m
b
e
d
d
i
n
g
s
a
u
g
m
e
n
t
t
h
i
s
b
e
h
a
v
i
o
r
b
y
m
a
i
n
t
a
i
n
i
n
g
i
d
e
n
t
it
y
c
o
h
e
r
e
n
c
y
o
v
e
r
l
o
w
-
r
e
s
o
l
u
t
i
o
n
a
n
d
o
c
c
l
u
d
e
d
f
a
c
e
s
.
T
h
is
is
a
ls
o
s
u
p
p
o
r
t
e
d
b
y
o
u
r
MO
T
P
s
c
o
r
e
o
f
8
0
.
3
%
,
a
s
s
p
at
i
a
l
c
o
n
s
is
t
e
n
c
y
is
r
et
a
i
n
e
d
t
h
r
o
u
g
h
P
D
A
F
a
n
d
C
J
L
F
,
w
h
ic
h
r
e
d
u
c
e
d
d
r
i
f
t
i
n
t
h
e
t
r
aj
e
c
t
o
r
y
.
O
v
e
r
a
l
l
,
t
h
es
e
q
u
a
n
t
it
a
t
i
v
e
m
e
t
r
i
c
s
d
e
m
o
n
s
t
r
a
te
t
h
a
t
t
h
e
p
r
o
p
o
s
e
d
f
r
a
m
ew
o
r
k
co
n
s
i
s
t
e
n
t
l
y
p
r
o
v
i
d
e
s
h
i
g
h
a
c
c
u
r
a
c
y
a
n
d
t
e
m
p
o
r
a
l
c
o
n
s
i
s
t
e
n
c
y
;
t
h
e
r
e
f
o
r
e
,
i
t
m
a
y
b
e
i
m
p
l
e
m
e
n
t
e
d
i
n
r
e
al
-
t
i
m
e
s
u
r
v
ei
l
l
a
n
c
e
-
d
r
i
v
e
n
a
p
p
li
c
a
ti
o
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
o
b
u
s
t m
u
lti
-
fa
ce
s
r
ec
o
g
n
itio
n
a
n
d
tr
a
ck
in
g
via
f
u
z
z
y
g
en
etic
a
lg
o
r
ith
ms a
n
d
…
(
A
d
il A
b
d
u
lh
u
r
A
b
u
s
h
a
n
a
)
215
T
ab
le
5
.
C
o
m
p
a
r
ativ
e
ev
alu
ati
o
n
o
f
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
ag
ain
s
t b
aselin
e
m
o
d
els ac
r
o
s
s
m
u
ltip
le
d
atasets
M
e
t
h
o
d
D
a
t
a
s
e
t
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
s
c
o
r
e
(
%)
M
O
TA
(
%)
M
O
TP
(
%)
I
D
S
F
r
a
g
R
u
n
t
i
me
(FPS)
D
e
e
p
S
O
R
T
[
3
]
W
I
D
ER
-
F
a
c
e
8
8
.
1
±
1
.
7
7
4
.
6
±
2
.
1
8
0
.
7
±
1
.
8
6
1
.
5
±
2
.
3
7
9
.
2
±
1
.
1
1
0
2
1
1
8
3
0
2
8
.
4
±
0
.
9
Y
TF
8
7
.
4
±
1
.
9
7
5
.
8
±
2
.
0
8
1
.
2
±
1
.
7
6
0
.
8
±
2
.
4
7
8
.
6
±
1
.
3
9
8
6
1
7
7
5
2
7
.
9
±
1
.
0
I
JB
-
S
8
6
.
8
±
2
.
2
7
3
.
3
±
2
.
4
7
9
.
4
±
1
.
9
5
9
.
6
±
2
.
6
7
7
.
8
±
1
.
4
1
1
2
5
1
9
2
1
2
6
.
8
±
1
.
1
R
e
t
i
n
a
F
a
c
e
+
K
a
l
m
a
n
[
2
3
]
W
I
D
ER
-
F
a
c
e
9
0
.
4
±
1
.
3
8
0
.
5
±
1
.
6
8
5
.
1
±
1
.
4
6
5
.
9
±
2
.
0
8
1
.
6
±
1
.
0
7
4
4
1
6
1
0
2
5
.
6
±
1
.
2
Y
TF
8
9
.
9
±
1
.
4
8
0
.
7
±
1
.
5
8
5
.
0
±
1
.
3
6
5
.
4
±
2
.
1
8
1
.
2
±
1
.
1
7
5
9
1
6
3
4
2
5
.
0
±
1
.
1
I
JB
-
S
8
9
.
5
±
1
.
5
7
9
.
8
±
1
.
7
8
4
.
3
±
1
.
5
6
4
.
8
±
2
.
2
8
0
.
9
±
1
.
1
7
7
1
1
6
5
9
2
4
.
5
±
1
.
0
P
r
o
p
o
se
d
(
f
u
z
z
y
g
e
n
e
t
i
c
+
d
e
e
p
c
o
u
p
l
e
d
)
W
I
D
ER
-
F
a
c
e
9
1
.
0
±
1
.
1
8
2
.
4
±
1
.
3
8
6
.
5
±
1
.
2
6
8
.
2
±
1
.
8
8
2
.
5
±
0
.
9
6
9
8
1
5
6
2
3
1
.
4
±
1
.
0
Y
TF
9
0
.
5
±
1
.
2
8
1
.
3
±
1
.
5
8
6
.
1
±
1
.
2
6
6
.
5
±
1
.
7
8
0
.
3
±
1
.
0
6
6
9
1
7
4
5
3
1
.
8
±
1
.
2
I
JB
-
S
9
0
.
2
±
1
.
3
8
0
.
9
±
1
.
4
8
5
.
8
±
1
.
2
6
6
.
1
±
1
.
9
7
9
.
9
±
1
.
1
7
0
1
1
7
9
2
3
0
.
9
±
1
.
1
T
ab
le
6
.
Qu
a
n
titativ
e
co
m
p
ar
i
s
o
n
s
with
th
e
s
tate
-
of
-
th
e
-
ar
t t
r
ac
k
in
g
m
et
h
o
d
s
o
n
th
e
v
id
eo
d
ataset
M
e
t
h
o
d
R
e
c
a
l
l
P
r
e
c
i
s
i
o
n
F1
-
sc
o
r
e
F
A
F
MT
I
D
S
F
G
M
O
TA
M
O
TP
A
D
M
M
[
2
4
]
7
5
.
5
6
1
.
8
6
8
.
0
0
.
5
0
23
2
3
8
2
2
9
5
9
5
1
.
7
6
3
.
7
I
H
TLS
[
2
5
]
7
5
.
5
6
8
.
0
7
1
.
6
0
.
4
1
23
2
0
1
3
2
8
8
0
5
6
.
2
6
3
.
7
Pre
-
t
r
a
i
n
e
d
[
1
1
]
6
0
.
1
8
8
.
8
7
1
.
7
0
.
1
7
5
9
3
1
2
1
4
0
5
1
.
5
7
9
.
5
M
u
l
t
i
-
t
a
r
g
e
t
l
e
a
r
n
i
n
g
a
n
d
d
e
t
e
c
t
i
o
n
(
mTLD
)
[
1
1
]
6
9
.
1
8
8
.
1
7
1
.
4
0
.
2
1
14
1
9
1
4
2
7
8
6
5
7
.
7
8
0
.
1
P
I
TA
D
F
[
2
6
]
8
1
.
7
9
0
.
2
8
5
.
3
0
.
2
7
32
6
2
4
1
6
4
5
6
9
.
2
8
6
.
0
O
u
r
8
1
.
0
9
0
.
5
8
6
.
1
0
.
1
5
30
6
6
9
1
7
4
5
6
6
.
5
8
0
.
3
0
T
ab
le
7
s
h
o
ws
th
at
th
e
p
r
o
p
o
s
ed
f
u
zz
y
g
e
n
etic+
d
ee
p
c
o
u
p
l
ed
R
esNet
f
r
am
ewo
r
k
is
s
u
p
e
r
io
r
to
th
e
h
y
b
r
id
tr
ac
k
er
s
o
f
C
NN+
p
ar
ticle
s
war
m
o
p
tim
izatio
n
(
PS
O)
an
d
YOL
O+
Kalm
an
f
ilte
r
:
+3
.
9
%
F1
-
s
co
r
e
im
p
r
o
v
em
e
n
t,
+
5
.
2
%
MO
T
A
im
p
r
o
v
e
m
en
t,
a
n
d
h
as
t
h
e
lo
west
I
D
s
witch
es
an
d
FG
r
ates.
T
h
ese
im
p
r
o
v
em
e
n
ts
ar
e
en
ab
led
b
y
th
e
f
u
zz
y
r
u
le
–
g
u
id
ed
GA,
wh
ich
ca
n
ad
ap
tiv
el
y
r
ef
in
e
ass
o
ciatio
n
h
y
p
o
th
eses
f
o
r
o
b
jects
u
n
d
e
r
o
cc
lu
s
io
n
an
d
u
n
ce
r
tain
ty
o
f
o
b
ject
m
o
tio
n
,
u
n
lik
e
th
e
f
ix
ed
-
p
ar
a
m
eter
PS
O/Kalm
an
m
eth
o
d
s
.
T
h
e
co
u
p
led
R
esNet
em
b
ed
d
in
g
s
also
im
p
r
o
v
e
th
e
co
n
s
is
ten
cy
o
f
tem
p
o
r
ally
s
im
ilar
ap
p
ea
r
an
ce
s
,
wh
ich
in
cr
ea
s
es r
o
b
u
s
tn
ess
to
lo
w
r
eso
lu
tio
n
an
d
p
o
s
e
v
ar
iat
io
n
.
T
ab
le
7
.
Qu
a
n
titativ
e
co
m
p
ar
i
s
o
n
with
r
ec
en
t h
y
b
r
id
tr
ac
k
in
g
m
o
d
els
M
e
t
h
o
d
R
e
c
a
l
l
(
%)
P
r
e
c
i
s
i
o
n
(
%)
F1
-
sc
o
r
e
(
%)
F
A
F
MT
I
D
S
F
r
a
g
M
O
TA
(
%)
M
O
TP
(
%)
C
N
N
+
P
S
O
7
7
.
2
8
7
.
9
8
2
.
2
0
.
2
3
22
1
0
3
5
2
0
2
8
6
1
.
3
7
8
.
2
Y
O
LO
+
K
a
l
m
a
n
f
i
l
t
e
r
7
9
.
6
8
8
.
4
8
3
.
7
0
.
2
0
25
9
1
2
1
8
6
5
6
3
.
5
7
9
.
4
D
e
e
p
S
O
R
T
(
C
N
N
+
K
a
l
ma
n
+
R
e
I
D
)
8
0
.
1
8
9
.
2
8
4
.
4
0
.
1
8
28
7
5
4
1
7
5
0
6
5
.
0
8
0
.
0
P
r
o
p
o
se
d
(
F
u
z
z
y
g
e
n
e
t
i
c
+
d
e
e
p
c
o
u
p
l
e
d
R
e
sN
e
t
)
8
1
.
0
9
0
.
5
8
6
.
1
0
.
1
5
30
6
6
9
1
7
4
5
6
6
.
5
8
0
.
3
W
h
ile
th
e
f
r
am
ewo
r
k
d
em
o
n
s
tr
ates
co
n
s
id
er
ab
le
ac
cu
r
ac
y
an
d
r
o
b
u
s
tn
ess
,
th
er
e
a
r
e
li
m
itatio
n
s
.
C
o
m
p
u
tatio
n
al
co
s
t
in
cr
ea
s
e
s
with
s
ce
n
e
d
en
s
ity
,
in
d
ica
tin
g
a
p
r
ac
tical
n
ee
d
f
o
r
lig
h
ter
-
weig
h
t
C
NN
b
ac
k
b
o
n
es
f
o
r
lar
g
e
-
s
ca
le
an
d
/o
r
em
b
ed
d
ed
d
e
p
lo
y
m
e
n
t.
Per
f
o
r
m
a
n
ce
co
u
ld
s
u
f
f
er
u
n
d
er
h
ea
v
y
o
cc
l
u
s
io
n
o
r
p
o
o
r
illu
m
in
atio
n
co
n
d
itio
n
s
d
u
e
to
r
elian
ce
o
n
v
is
u
al
cu
es
as
a
s
in
g
le
s
o
u
r
ce
o
f
in
f
o
r
m
a
tio
n
.
T
h
e
f
u
zz
y
r
u
le
b
ase
r
eq
u
ir
es
in
itial
m
an
u
al
tu
n
in
g
,
p
o
in
tin
g
to
th
e
p
o
s
s
ib
ilit
y
o
f
s
elf
-
ad
ap
tiv
e
o
r
r
ein
f
o
r
ce
m
e
n
t
-
d
r
iv
e
n
s
y
n
th
esizin
g
co
n
f
ig
u
r
atio
n
s
.
Fin
ally
,
im
p
r
o
v
e
d
cr
o
s
s
-
d
o
m
ain
g
en
er
aliza
tio
n
an
d
ad
v
e
r
s
ar
ial
r
o
b
u
s
tn
ess
ar
e
s
till
r
eq
u
ir
ed
f
o
r
wid
er
d
ep
l
o
y
m
en
t
s
ce
n
ar
io
s
,
esp
ec
ially
if
p
r
iv
ac
y
co
n
s
id
er
atio
n
s
ar
e
al
s
o
r
esp
ec
ted
in
t
h
e
s
o
lu
tio
n
d
esig
n
.
Fu
tu
r
e
o
p
ti
o
n
s
m
ay
b
e
im
p
r
o
v
e
d
wit
h
h
a
r
d
war
e
-
lev
el
o
p
tim
izatio
n
a
n
d
lo
wer
-
co
s
t
im
p
lem
en
tatio
n
m
eth
o
d
s
.
Field
p
r
o
g
r
a
m
m
ab
le
g
ate
a
r
r
ay
(
F
PGA)
b
ased
n
e
u
r
al
n
etwo
r
k
i
m
p
lem
en
tatio
n
s
ca
n
p
r
o
v
id
e
ca
p
a
b
ilit
ies
th
at
ac
h
iev
e
r
ea
l
-
tim
e
p
r
o
ce
s
s
in
g
wit
h
lo
w
laten
c
y
at
lo
w
im
p
le
m
en
tatio
n
c
o
s
t
[
2
7
]
.
Ad
d
itio
n
ally
,
R
asp
b
er
r
y
Pi
p
l
atf
o
r
m
s
ef
f
ec
tiv
ely
d
e
p
lo
y
i
n
tel
lig
en
t
s
y
s
tem
s
in
lo
w
-
p
o
wer
en
v
ir
o
n
m
en
ts
s
u
c
h
as c
lass
r
o
o
m
s
o
r
s
m
all
s
u
r
v
eillan
ce
s
tu
d
ies
[
2
8
]
.
4.
CO
NCLU
SI
O
N
T
h
is
s
t
u
d
y
p
r
o
p
o
s
e
d
a
h
y
b
r
i
d
m
u
l
ti
-
f
ac
e
r
ec
o
g
n
iti
o
n
a
n
d
tr
a
c
k
i
n
g
f
r
a
m
e
wo
r
k
t
h
a
t
i
n
co
r
p
o
r
a
tes
d
e
ep
c
o
u
p
le
d
R
esN
et
v
is
u
al
f
ea
t
u
r
es
al
o
n
g
s
i
d
e
f
u
z
z
y
g
e
n
et
ic
o
p
ti
m
i
za
t
io
n
,
tac
k
l
in
g
p
e
r
s
i
s
ten
t
c
h
all
e
n
g
es
o
f
p
a
r
ti
al
o
cc
l
u
s
i
o
n
,
ill
u
m
i
n
at
io
n
ch
a
n
g
es,
a
n
d
b
a
ck
g
r
o
u
n
d
c
lu
tte
r
f
o
u
n
d
i
n
r
e
al
-
w
o
r
ld
s
u
r
v
e
illa
n
c
e
s
etti
n
g
s
.
B
y
lin
k
i
n
g
d
ee
p
f
ea
tu
r
e
em
b
e
d
d
in
g
s
t
o
a
d
a
p
ti
v
e
f
u
zz
y
–
g
e
n
e
t
ic
r
u
l
es,
t
h
e
f
r
a
m
ew
o
r
k
d
y
n
am
i
ca
l
ly
co
m
b
i
n
es
m
o
ti
o
n
,
a
p
p
ea
r
a
n
ce
,
an
d
c
o
n
f
id
e
n
ce
c
u
es
t
o
im
p
r
o
v
e
d
a
ta
ass
o
cia
ti
o
n
w
h
il
e
ac
co
u
n
ti
n
g
f
o
r
u
n
ce
r
ta
in
ty
i
n
v
is
u
al
c
o
n
d
iti
o
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
1
,
Ma
r
c
h
2
0
2
6
:
208
-
2
1
7
216
ACK
NO
WL
E
DG
M
E
N
T
S
T
h
e
au
th
o
r
s
ar
e
g
r
ate
f
u
l f
o
r
th
e
in
s
titu
tio
n
al
r
eso
u
r
ce
s
th
at
f
a
cilitated
th
is
r
esear
ch
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
A
d
i
l
A
b
d
u
l
h
u
r
A
b
u
s
h
a
n
a
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Y
o
u
s
i
f
Sa
m
e
r
M
u
d
h
a
f
a
r
✓
✓
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
Data
av
ailab
ilit
y
is
n
o
t a
p
p
lica
b
le
to
th
is
p
ap
er
as n
o
n
ew
d
at
a
wer
e
cr
ea
ted
in
t
h
is
s
tu
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
W
.
L
u
o
,
J.
X
i
n
g
,
A
.
M
i
l
a
n
,
X
.
Z
h
a
n
g
,
W
.
L
i
u
,
a
n
d
T.
K
.
K
i
m
,
“
M
u
l
t
i
p
l
e
o
b
j
e
c
t
t
r
a
c
k
i
n
g
:
a
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
,
”
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
2
9
3
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
r
t
i
n
t
.
2
0
2
0
.
1
0
3
4
4
8
.
[
2
]
G
.
C
i
a
p
a
r
r
o
n
e
,
F
.
L
.
S
á
n
c
h
e
z
,
S
.
Ta
b
i
k
,
L
.
Tr
o
i
a
n
o
,
R
.
Ta
g
l
i
a
f
e
r
r
i
,
a
n
d
F
.
H
e
r
r
e
r
a
,
“
D
e
e
p
l
e
a
r
n
i
n
g
i
n
v
i
d
e
o
m
u
l
t
i
-
o
b
j
e
c
t
t
r
a
c
k
i
n
g
:
a
su
r
v
e
y
,
”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
3
8
1
,
p
p
.
6
1
–
8
8
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m.2
0
1
9
.
1
1
.
0
2
3
.
[3
]
N
.
W
o
j
k
e
,
A
.
B
e
w
l
e
y
,
a
n
d
D
.
P
a
u
l
u
s,
“
S
i
m
p
l
e
o
n
l
i
n
e
a
n
d
r
e
a
l
t
i
me
t
r
a
c
k
i
n
g
w
i
t
h
a
d
e
e
p
a
ss
o
c
i
a
t
i
o
n
me
t
r
i
c
,
”
i
n
2
0
1
7
I
E
EE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
I
m
a
g
e
P
ro
c
e
ssi
n
g
(
I
C
I
P)
,
2
0
1
7
,
p
p
.
3
6
4
5
–
3
6
4
9
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
P
.
2
0
1
7
.
8
2
9
6
9
6
2
.
[
4
]
P
.
B
e
r
g
ma
n
n
,
T
.
M
e
i
n
h
a
r
d
t
,
a
n
d
L
.
L
.
-
Ta
i
x
e
,
“
Tr
a
c
k
i
n
g
w
i
t
h
o
u
t
b
e
l
l
s a
n
d
w
h
i
s
t
l
e
s
,
”
i
n
2
0
1
9
I
EE
E/
C
VF
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
s
i
o
n
(
I
C
C
V)
,
2
0
1
9
,
p
p
.
9
4
1
–
9
5
1
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
V
.
2
0
1
9
.
0
0
1
0
3
.
[
5
]
S
.
S
u
n
,
N
.
A
k
h
t
a
r
,
H
.
S
o
n
g
,
A
.
S
.
M
i
a
n
,
a
n
d
M
.
S
h
a
h
,
“
D
e
e
p
a
f
f
i
n
i
t
y
n
e
t
w
o
r
k
f
o
r
m
u
l
t
i
p
l
e
o
b
j
e
c
t
t
r
a
c
k
i
n
g
,
”
I
E
EE
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
.
4
3
,
n
o
.
1
,
p
p
.
1
0
4
–
1
1
9
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
TPA
M
I
.
2
0
1
9
.
2
9
2
9
5
2
0
.
[
6
]
M
.
W
a
n
g
a
n
d
W
.
D
e
n
g
,
“
D
e
e
p
f
a
c
e
r
e
c
o
g
n
i
t
i
o
n
:
a
s
u
r
v
e
y
,
”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
4
2
9
,
p
p
.
2
1
5
–
2
4
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m.
2
0
2
0
.
1
0
.
0
8
1
.
[
7
]
P
.
D
e
n
d
o
r
f
e
r
e
t
a
l
.
,
“
M
O
T2
0
:
a
b
e
n
c
h
mark
f
o
r
m
u
l
t
i
o
b
j
e
c
t
t
r
a
c
k
i
n
g
i
n
c
r
o
w
d
e
d
s
c
e
n
e
s,
”
2
0
2
0
,
a
rX
i
v
:
2
0
0
3
.
0
9
0
0
3
.
[
8
]
Y
.
Zh
a
n
g
e
t
a
l
.
,
“
B
y
t
e
Tr
a
c
k
:
mu
l
t
i
-
o
b
j
e
c
t
t
r
a
c
k
i
n
g
b
y
a
s
so
c
i
a
t
i
n
g
e
v
e
r
y
d
e
t
e
c
t
i
o
n
b
o
x
,
”
i
n
C
o
m
p
u
t
e
r
Vi
si
o
n
–
E
C
C
V
2
0
2
2
,
C
h
a
m
,
S
w
i
t
z
e
r
l
a
n
d
:
S
p
r
i
n
g
e
r
,
2
0
2
2
,
p
p
.
1
–
21
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
1
-
2
0
0
4
7
-
2
_
1
.
[
9
]
S
.
M
i
r
j
a
l
i
l
i
,
J.
S
.
D
o
n
g
,
A
.
S
.
S
a
d
i
q
,
a
n
d
H
.
F
a
r
i
s
,
“
G
e
n
e
t
i
c
a
l
g
o
r
i
t
h
m
:
t
h
e
o
r
y
,
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
,
a
n
d
a
p
p
l
i
c
a
t
i
o
n
i
n
i
ma
g
e
r
e
c
o
n
st
r
u
c
t
i
o
n
,
”
i
n
S
t
u
d
i
e
s i
n
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
,
v
o
l
.
8
1
1
,
2
0
2
0
,
p
p
.
6
9
–
85
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
0
-
1
2
1
2
7
-
3
_
5
.
[
1
0
]
Y
.
B
.
‐
S
h
a
l
o
m,
X
.
‐
R
o
n
.
Li
,
a
n
d
T.
K
i
r
u
b
a
r
a
j
a
n
,
Es
t
i
m
a
t
i
o
n
w
i
t
h
a
p
p
l
i
c
a
t
i
o
n
s
t
o
t
r
a
c
k
i
n
g
a
n
d
n
a
v
i
g
a
t
i
o
n
,
3
r
d
e
d
.
N
e
w
Y
o
r
k
:
J
o
h
n
W
i
l
e
y
& So
n
s
,
2
0
0
2
,
d
o
i
:
1
0
.
1
0
0
2
/
0
4
7
1
2
2
1
2
7
9
.
[
1
1
]
S
.
Z
h
a
n
g
e
t
a
l
.
,
“
Tr
a
c
k
i
n
g
p
e
r
so
n
s
-
of
-
i
n
t
e
r
e
st
v
i
a
a
d
a
p
t
i
v
e
d
i
scri
mi
n
a
t
i
v
e
f
e
a
t
u
r
e
s
,
”
i
n
C
o
m
p
u
t
e
r
V
i
si
o
n
–
EC
C
V
2
0
1
6
,
C
h
a
m
,
S
w
i
t
z
e
r
l
a
n
d
:
S
p
r
i
n
g
e
r
,
2
0
1
6
,
p
p
.
4
1
5
–
4
3
3
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
3
1
9
-
4
6
4
5
4
-
1
_
2
6
.
[
1
2
]
S
.
W
.
A
r
a
c
h
c
h
i
l
a
g
e
a
n
d
E.
I
z
q
u
i
e
r
d
o
,
“
A
d
a
p
t
i
v
e
a
g
g
r
e
g
a
t
e
d
t
r
a
c
k
l
e
t
l
i
n
k
i
n
g
f
o
r
m
u
l
t
i
-
f
a
c
e
t
r
a
c
k
i
n
g
,
”
i
n
2
0
2
0
I
E
EE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
m
a
g
e
Pro
c
e
s
s
i
n
g
(
I
C
I
P)
,
2
0
2
0
,
p
p
.
1
3
6
6
–
1
3
7
0
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
P
4
0
7
7
8
.
2
0
2
0
.
9
1
9
0
8
2
3
.
[
1
3
]
G
.
B
a
r
q
u
e
r
o
,
C
.
F
e
r
n
a
n
d
e
z
,
a
n
d
I
.
H
u
p
o
n
t
,
“
Lo
n
g
-
t
e
r
m
f
a
c
e
t
r
a
c
k
i
n
g
f
o
r
c
r
o
w
d
e
d
v
i
d
e
o
-
su
r
v
e
i
l
l
a
n
c
e
s
c
e
n
a
r
i
o
s
,
”
i
n
I
J
C
B
2
0
2
0
-
I
EEE/
I
AP
R
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
i
n
t
C
o
n
f
e
ren
c
e
o
n
B
i
o
m
e
t
r
i
c
s
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
I
JC
B
4
8
5
4
8
.
2
0
2
0
.
9
3
0
4
8
9
2
.
[
1
4
]
S
.
Z
h
a
n
g
e
t
a
l
.
,
“
Tr
a
c
k
i
n
g
p
e
r
s
o
n
s
-
of
-
i
n
t
e
r
e
st
v
i
a
u
n
s
u
p
e
r
v
i
se
d
r
e
p
r
e
se
n
t
a
t
i
o
n
a
d
a
p
t
a
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
C
o
m
p
u
t
e
r
Vi
si
o
n
,
v
o
l
.
1
2
8
,
n
o
.
1
,
p
p
.
9
6
–
1
2
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
2
6
3
-
019
-
0
1
2
1
2
-
1.
[
1
5
]
G
.
B
a
r
q
u
e
r
o
,
I
.
H
u
p
o
n
t
,
a
n
d
C
.
F
.
Te
n
a
,
“
R
a
n
k
-
b
a
se
d
v
e
r
i
f
i
c
a
t
i
o
n
f
o
r
l
o
n
g
-
t
e
r
m
f
a
c
e
t
r
a
c
k
i
n
g
i
n
c
r
o
w
d
e
d
s
c
e
n
e
s,
”
I
E
E
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
B
i
o
m
e
t
r
i
c
s,
Be
h
a
v
i
o
r,
a
n
d
I
d
e
n
t
i
t
y
S
c
i
e
n
c
e
,
v
o
l
.
3
,
n
o
.
4
,
p
p
.
4
9
5
–
5
0
5
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
TB
I
O
M
.
2
0
2
1
.
3
0
9
9
5
6
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
o
b
u
s
t m
u
lti
-
fa
ce
s
r
ec
o
g
n
itio
n
a
n
d
tr
a
ck
in
g
via
f
u
z
z
y
g
en
etic
a
lg
o
r
ith
ms a
n
d
…
(
A
d
il A
b
d
u
lh
u
r
A
b
u
s
h
a
n
a
)
217
[
1
6
]
J.
W
a
n
g
a
n
d
J.
L
a
n
g
,
“
V
i
s
u
a
l
mu
l
t
i
-
f
a
c
e
t
r
a
c
k
i
n
g
a
p
p
l
i
e
d
t
o
c
o
u
n
c
i
l
p
r
o
c
e
e
d
i
n
g
s
,
”
I
EEE
I
n
st
r
u
m
e
n
t
a
t
i
o
n
&
Me
a
su
r
e
m
e
n
t
Ma
g
a
zi
n
e
,
v
o
l
.
2
4
,
n
o
.
3
,
p
p
.
7
8
–
8
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
M
I
M
.
2
0
2
1
.
9
4
3
6
0
8
9
.
[
1
7
]
T.
M
.
Tr
a
n
e
t
a
l
.
,
“
R
e
S
O
R
T
:
a
n
I
D
-
r
e
c
o
v
e
r
y
m
u
l
t
i
-
f
a
c
e
t
r
a
c
k
i
n
g
m
e
t
h
o
d
f
o
r
s
u
r
v
e
i
l
l
a
n
c
e
c
a
m
e
r
a
s,
”
i
n
2
0
2
1
1
6
t
h
I
EEE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Au
t
o
m
a
t
i
c
Fa
c
e
a
n
d
G
e
st
u
re
Re
c
o
g
n
i
t
i
o
n
(
FG
2
0
2
1
)
,
2
0
2
1
,
p
p
.
1
–
8
,
d
o
i
:
1
0
.
1
1
0
9
/
F
G
5
2
6
3
5
.
2
0
2
1
.
9
6
6
6
9
4
1
.
[
1
8
]
G
.
R
e
n
,
X
.
Lu
,
a
n
d
Y
.
L
i
,
“
A
c
r
o
ss
-
c
a
mer
a
m
u
l
t
i
-
f
a
c
e
t
r
a
c
k
i
n
g
s
y
s
t
e
m
b
a
sed
o
n
d
o
u
b
l
e
t
r
i
p
l
e
t
n
e
t
w
o
r
k
s,
”
I
EE
E
Ac
c
e
ss
,
v
o
l
.
9
,
p
p
.
4
3
7
5
9
–
4
3
7
7
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
1
.
3
0
6
1
5
7
2
.
[
1
9
]
Z.
W
e
n
g
,
H
.
Zh
u
a
n
g
,
H
.
L
i
,
B
.
R
a
m
a
l
i
n
g
a
m,
R
.
E.
M
o
h
a
n
,
a
n
d
Z.
Li
n
,
“
O
n
l
i
n
e
m
u
l
t
i
-
f
a
c
e
t
r
a
c
k
i
n
g
w
i
t
h
mu
l
t
i
-
mo
d
a
l
i
t
y
c
a
sc
a
d
e
d
mat
c
h
i
n
g
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
i
rc
u
i
t
s
a
n
d
S
y
st
e
m
s
f
o
r
Vi
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
3
3
,
n
o
.
6
,
p
p
.
2
7
3
8
–
2
7
5
2
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
TC
S
V
T
.
2
0
2
2
.
3
2
2
4
6
9
9
.
[
2
0
]
J.
K
i
m,
C
.
Y
.
J
u
,
G
.
W
.
K
i
m,
a
n
d
D
.
H
.
Le
e
,
“
B
o
T
-
F
a
c
e
S
O
R
T:
b
a
g
-
of
-
t
r
i
c
k
s
f
o
r
r
o
b
u
s
t
m
u
l
t
i
-
f
a
c
e
t
r
a
c
k
i
n
g
i
n
u
n
c
o
n
st
r
a
i
n
e
d
v
i
d
e
o
s,
”
i
n
C
o
m
p
u
t
e
r V
i
si
o
n
–
A
C
C
V
2
0
2
4
,
S
i
n
g
a
p
o
r
e
:
S
p
r
i
n
g
e
r
,
2
0
2
5
,
p
p
.
2
7
8
–
294
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
981
-
96
-
0
9
0
1
-
7
_
1
7
.
[
2
1
]
R
.
J
o
c
h
l
a
n
d
A
.
U
h
l
,
“
F
a
c
e
Q
S
O
R
T:
c
a
n
a
c
o
m
b
i
n
a
t
i
o
n
o
f
t
w
o
b
i
o
m
e
t
r
i
c
f
e
a
t
u
r
e
s
a
c
h
i
e
v
e
c
o
mp
e
t
i
t
i
v
e
p
e
r
f
o
r
m
a
n
c
e
?
,
”
i
n
2
0
2
5
2
5
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
D
i
g
i
t
a
l
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
(
D
S
P)
,
2
0
2
5
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
D
S
P
6
5
4
0
9
.
2
0
2
5
.
1
1
0
7
5
1
5
1
.
[
2
2
]
L.
L
i
a
n
d
X
.
Z
h
a
n
,
“
A
n
o
v
e
l
d
a
t
a
a
ss
o
c
i
a
t
i
o
n
a
l
g
o
r
i
t
h
m
b
a
se
d
o
n
f
u
z
z
y
l
o
g
i
c
f
o
r
v
i
su
a
l
o
b
j
e
c
t
t
r
a
c
k
i
n
g
,
”
i
n
2
0
1
9
1
2
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
g
ress
o
n
I
m
a
g
e
a
n
d
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
,
B
i
o
Me
d
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
a
n
d
I
n
f
o
rm
a
t
i
c
s
,
2
0
1
9
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
C
I
S
P
-
B
M
EI
4
8
8
4
5
.
2
0
1
9
.
8
9
6
5
7
6
9
.
[
2
3
]
C
.
G
a
o
,
Q
.
Li
u
,
Q
.
X
u
,
L.
W
a
n
g
,
J
.
Li
u
,
a
n
d
C
.
Zo
u
,
“
S
k
e
t
C
h
y
C
o
c
o
:
i
ma
g
e
g
e
n
e
r
a
t
i
o
n
f
r
o
m
f
r
e
e
h
a
n
d
s
c
e
n
e
s
k
e
t
c
h
e
s,
”
i
n
2
0
2
0
I
EEE/
C
VF
C
o
n
f
e
r
e
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
rn
Re
c
o
g
n
i
t
i
o
n
(
C
VPR)
,
2
0
2
0
,
p
p
.
5
1
7
3
–
5
1
8
2
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
4
2
6
0
0
.
2
0
2
0
.
0
0
5
2
2
.
[
2
4
]
M
.
A
y
a
z
o
g
l
u
,
M
.
S
z
n
a
i
e
r
,
a
n
d
O
.
I
.
C
a
m
p
s,
“
F
a
st
a
l
g
o
r
i
t
h
ms
f
o
r
st
r
u
c
t
u
r
e
d
r
o
b
u
st
p
r
i
n
c
i
p
a
l
c
o
m
p
o
n
e
n
t
a
n
a
l
y
si
s,”
i
n
2
0
1
2
I
EEE
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
Pa
t
t
e
r
n
Re
c
o
g
n
i
t
i
o
n
,
2
0
1
2
,
p
p
.
1
7
0
4
–
1
7
1
1
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
.
2
0
1
2
.
6
2
4
7
8
6
5
.
[
2
5
]
C
.
D
i
c
l
e
,
O
.
I
.
C
a
m
p
s,
a
n
d
M
.
S
z
n
a
i
e
r
,
“
T
h
e
w
a
y
t
h
e
y
mo
v
e
:
t
r
a
c
k
i
n
g
m
u
l
t
i
p
l
e
t
a
r
g
e
t
s
w
i
t
h
si
m
i
l
a
r
a
p
p
e
a
r
a
n
c
e
,
”
i
n
2
0
1
3
I
E
E
E
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
p
u
t
e
r V
i
s
i
o
n
,
2
0
1
3
,
p
p
.
2
3
0
4
–
2
3
1
1
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
V
.
2
0
1
3
.
2
8
6
.
[
2
6
]
C.
-
C
.
Li
n
a
n
d
Y
.
H
u
n
g
,
“
A
p
r
i
o
r
-
l
e
ss
met
h
o
d
f
o
r
m
u
l
t
i
-
f
a
c
e
t
r
a
c
k
i
n
g
i
n
u
n
c
o
n
s
t
r
a
i
n
e
d
v
i
d
e
o
s,
”
i
n
2
0
1
8
I
EE
E/
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
Re
c
o
g
n
i
t
i
o
n
,
2
0
1
8
,
p
p
.
5
3
8
–
5
4
7
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
.
2
0
1
8
.
0
0
0
6
3
.
[
2
7
]
S
.
H
.
A
b
d
u
l
n
a
b
i
,
Y
.
S
.
M
u
d
h
a
f
a
r
,
A
.
A
.
K
a
d
h
i
m,
M
.
B
.
M
a
h
d
i
,
a
n
d
H
.
H
.
S
o
j
a
r
,
“
N
e
u
r
a
l
n
e
t
w
o
r
k
-
b
a
se
d
s
y
s
t
e
m
i
d
e
n
t
i
f
i
c
a
t
i
o
n
:
a
c
o
m
p
r
e
h
e
n
si
v
e
F
P
G
A
d
e
s
i
g
n
a
n
d
i
mp
l
e
me
n
t
a
t
i
o
n
,
”
i
n
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
A
rt
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
M
e
c
h
a
t
r
o
n
i
c
s
S
y
s
t
e
m
,
AI
M
S
2
0
2
4
,
2
0
2
4
,
p
p
.
1
–
7
,
d
o
i
:
1
0
.
1
1
0
9
/
A
I
M
S
6
1
8
1
2
.
2
0
2
4
.
1
0
5
1
2
5
3
1
.
[
2
8
]
A
.
M
.
A
.
A
l
-
M
u
q
a
r
m,
Y
.
M
u
d
h
a
f
a
r
,
A
.
M
.
S
h
a
k
i
r
,
M
.
K
a
z
e
m,
R
.
A
.
-
Y
a
h
i
y
a
,
a
n
d
B
.
S
.
A
.
Za
h
r
a
,
“
Lo
w
-
c
o
st
s
mart
l
e
a
r
n
i
n
g
w
i
t
h
mo
o
d
l
e
-
b
a
se
d
R
a
s
p
b
e
r
r
y
P
i
4
f
o
r
u
n
i
v
e
r
si
t
y
st
u
d
e
n
t
s,
”
i
n
6
t
h
I
r
a
q
i
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
E
n
g
i
n
e
e
ri
n
g
T
e
c
h
n
o
l
o
g
y
a
n
d
i
t
s
Ap
p
l
i
c
a
t
i
o
n
s,
I
I
C
ETA
2
0
2
3
,
2
0
2
3
,
p
p
.
6
0
3
–
6
0
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
I
C
ETA
5
7
6
1
3
.
2
0
2
3
.
1
0
3
5
1
2
6
6
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Adi
l
Abd
u
l
h
u
r
Abu
sh
a
n
a
is
a
lec
tu
re
r
a
t
th
e
De
p
a
rtme
n
t
o
f
C
o
m
p
u
ter
S
c
ien
c
e
,
Un
iv
e
rsity
o
f
K
u
fa
.
He
re
c
e
iv
e
d
h
is
B.
S
c
.
d
e
g
re
e
i
n
M
a
th
e
m
a
ti
c
s
fro
m
S
a
lah
a
d
d
i
n
U
n
iv
e
rsit
y
,
Ira
q
,
i
n
1
9
8
9
,
a
n
d
h
is
M
.
S
c
.
d
e
g
re
e
fro
m
t
h
e
De
p
a
rtme
n
t
o
f
In
fo
rm
a
ti
o
n
Tec
h
n
o
lo
g
y
,
Un
iv
e
rsity
o
f
Uta
ra
M
a
lay
sia
,
in
2
0
0
9
,
re
sp
e
c
ti
v
e
l
y
.
He
is
c
u
rre
n
tl
y
p
u
rsu
i
n
g
a
P
h
.
D.
in
t
h
e
De
p
a
rtme
n
t
o
f
In
f
o
rm
a
ti
o
n
S
y
st
e
m
s
a
t
Un
iv
e
rsiti
Tek
n
o
l
o
g
i
M
a
lay
sia
(UTM
).
His
re
se
a
rc
h
in
tere
sts
in
c
l
u
d
e
p
a
tt
e
rn
re
c
o
g
n
i
ti
o
n
a
n
d
c
o
m
p
u
ter
v
isio
n
,
with
a
p
p
li
c
a
ti
o
n
s
t
o
b
i
o
m
e
tri
c
s.
He
wa
s
o
n
e
o
f
th
e
p
a
rti
c
ip
a
n
ts
wh
o
re
c
e
iv
e
d
a
p
a
p
e
r
a
wa
rd
a
t
t
h
e
IEE
E
c
o
n
fe
re
n
c
e
in
2
0
1
5
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
d
e
lsh
a
n
a
3
0
0
0
@
g
m
a
il
.
c
o
m
o
r
a
d
e
l.
a
ln
a
sra
wi@u
o
k
u
fa
.
e
d
u
.
iq
.
Yo
u
sif
S
a
m
e
r
Mu
d
h
a
fa
r
is
a
lec
tu
re
r
a
t
th
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
Un
iv
e
rsity
o
f
Ku
fa
.
H
e
e
a
rn
e
d
h
is
B.
S
c
.
in
Co
m
p
u
ter T
e
c
h
n
i
q
u
e
s
E
n
g
i
n
e
e
rin
g
fro
m
th
e
Isla
m
ic
Un
iv
e
rsity
i
n
Na
jaf
i
n
2
0
1
8
.
He
c
o
m
p
lete
d
h
is
M
.
S
c
.
in
Co
m
p
u
te
r
S
c
ien
c
e
E
n
g
i
n
e
e
rin
g
a
t
th
e
Un
iv
e
rsity
o
f
De
b
re
c
e
n
i
n
2
0
2
2
,
g
ra
d
u
a
ti
n
g
with
h
o
n
o
rs
a
n
d
re
c
e
iv
in
g
t
h
e
o
u
tsta
n
d
i
n
g
stu
d
e
n
t
c
e
rti
fica
te.
He
wo
rk
s
a
t
th
e
Un
iv
e
rsity
o
f
Ku
fa
,
F
a
c
u
lt
y
o
f
E
d
u
c
a
ti
o
n
,
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
c
o
m
p
u
ter
n
e
tw
o
rk
s
,
Io
T,
a
n
d
AI.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
y
o
u
sif
.
m
u
d
h
a
fa
r@iu
n
a
jaf.
e
d
u
.
i
q
.
Evaluation Warning : The document was created with Spire.PDF for Python.