I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
8
,
No
.
2
,
A
p
r
il
201
8
,
p
p
.
908
~
9
1
6
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
8
i
2
.
p
p
9
0
8
-
9
1
6
908
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
A Sys
te
m
a
tic
Rev
iew
of Ex
isting
Data
Mining
Appr
o
a
ches
Env
isio
ned f
o
r
K
no
w
ledg
e Disco
v
ery fro
m
Multi
m
e
dia
B
ena
k
a
Sa
ntho
s
ha
S
1
,
C
hitr
a
K
ira
n N
2
1
De
p
a
rtme
n
t
o
f
El
e
c
tro
n
ics
&
Co
m
m
u
n
ica
ti
o
n
En
g
in
e
e
rin
g
,
C
o
o
rg
In
stit
u
te o
f
T
e
c
h
n
o
lo
g
y
,
Ko
d
a
g
u
,
Ka
rn
a
tak
a
,
In
d
ia
2
De
p
a
rtme
n
t
o
f
El
e
c
tro
n
ics
&
Co
m
m
u
n
ica
ti
o
n
En
g
in
e
e
rin
g
,
S
a
i
V
i
d
y
a
In
stit
u
te o
f
T
e
c
h
n
o
lo
g
y
,
Be
n
g
a
lu
ru
,
Ka
rn
a
tak
a
,
In
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Sep
2
5
,
2
0
1
7
R
ev
i
s
ed
J
an
3
,
2
0
1
8
A
cc
ep
ted
J
an
1
7
,
2
0
1
8
T
h
e
e
x
ten
siv
e
u
se
o
f
m
u
lt
i
m
e
d
ia
tec
h
n
o
lo
g
ies
e
x
ten
d
e
d
t
h
e
a
p
p
li
c
a
b
il
it
y
o
f
in
f
o
rm
a
ti
o
n
tec
h
n
o
lo
g
y
to
a
l
a
rg
e
e
x
t
e
n
t
w
h
ich
re
su
lt
s
e
n
o
r
m
o
u
s
g
e
n
e
ra
ti
o
n
o
f
c
o
m
p
lex
m
u
lt
i
m
e
d
ia
c
o
n
ten
ts
o
v
e
r
th
e
in
tern
e
t.
T
h
e
re
f
o
re
th
e
n
u
m
b
e
r
o
f
m
u
lt
i
m
e
d
ia
c
o
n
ten
ts
a
v
a
il
a
b
le
t
o
th
e
u
se
r
is
a
lso
e
x
p
o
n
e
n
t
ially
in
c
re
a
sin
g
.
In
th
is
d
ig
it
a
l
e
ra
o
f
th
e
c
lo
u
d
-
e
n
a
b
led
In
ter
n
e
t
o
f
T
h
in
g
s
(Io
T
),
a
n
a
ly
sis
o
f
c
o
m
p
lex
v
id
e
o
a
n
d
ima
g
e
d
a
ta
p
la
y
s
a
c
ru
c
ial
ro
le.It
a
i
m
s
to
e
x
trac
t
m
e
a
n
i
n
g
f
u
l
in
f
o
r
m
a
ti
o
n
a
s
th
e
d
istri
b
u
te
d
sto
ra
g
e
s
a
n
d
p
ro
c
e
ss
in
g
e
le
m
e
n
ts
w
it
h
in
a
b
a
n
d
w
id
th
c
o
n
stra
i
n
t
n
e
t
w
o
rk
s
e
e
k
o
p
ti
m
a
l
so
lu
ti
o
n
s
to
i
n
c
re
a
se
th
e
th
ro
u
g
h
p
u
t
a
lo
n
g
w
it
h
a
n
o
p
ti
m
a
l
trad
e
-
o
ff
b
e
t
w
e
e
n
c
o
m
p
u
tatio
n
a
l
c
o
m
p
lex
it
y
a
n
d
p
o
w
e
r
c
o
n
su
m
p
ti
o
n
.
H
o
w
e
v
e
r,
d
u
e
to
c
o
m
p
lex
c
h
a
ra
c
teristics
o
f
v
isu
a
l
p
a
tt
e
rn
s
a
n
d
v
a
riatio
n
s
in
v
id
e
o
f
ra
m
e
s,
it
is
n
o
t
a
tri
v
ial
tas
k
to
d
isc
o
v
e
r
m
e
a
n
in
g
fu
l
in
f
o
rm
a
ti
o
n
a
n
d
c
o
rre
latio
n
.
He
n
c
e
,
d
a
ta
m
in
in
g
h
a
s
e
m
e
r
g
e
d
a
s
a
f
ield
w
h
ich
h
a
s
d
iv
e
rse
a
sp
e
c
ts
p
re
se
n
tl
y
in
e
x
trac
ti
n
g
m
e
a
n
in
g
f
u
l
h
id
d
e
n
p
a
tt
e
rn
s
f
ro
m
th
e
c
o
m
p
lex
i
m
a
g
e
a
n
d
v
id
e
o
d
a
ta
c
o
n
sid
e
rin
g
d
if
f
e
re
n
t
p
a
tt
e
rn
c
las
si
f
ica
ti
o
n
a
p
p
ro
a
c
h
.
T
h
e
stu
d
y
m
o
stl
y
in
v
e
stig
a
tes
th
e
e
x
isti
n
g
d
a
ta
-
m
i
n
in
g
to
o
ls
a
n
d
th
e
ir
p
e
rf
o
rm
a
n
c
e
m
e
tri
c
f
o
r
th
e
p
u
r
p
o
se
o
f
re
v
ie
w
in
g
th
is
re
se
a
rc
h
trac
k
.
It
a
lso
h
ig
h
li
g
h
ts
th
e
re
latio
n
sh
ip
b
e
tw
e
e
n
f
re
q
u
e
n
t
p
a
tt
e
rn
s
a
n
d
d
is
c
rim
in
a
ti
v
e
f
e
a
tu
re
s
a
ss
o
c
iate
d
w
i
th
a
v
id
e
o
o
b
jec
t.
F
in
a
ll
y
,
th
e
stu
d
y
a
d
d
re
ss
e
s
th
e
e
x
isti
n
g
re
se
a
rc
h
issu
e
s
to
stre
n
g
th
e
n
u
p
t
h
e
f
u
tu
re
d
irec
ti
o
n
o
f
re
se
a
rc
h
to
w
a
rd
s
v
id
e
o
a
n
a
l
y
ti
c
s
a
n
d
p
a
t
tern
re
c
o
g
n
it
io
n
.
K
ey
w
o
r
d
:
Data
m
i
n
i
n
g
I
n
ter
n
et
o
f
t
h
i
n
g
s
(
I
o
T
)
Mu
lti
m
ed
ia
Co
p
y
rig
h
t
©
2
0
1
8
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
B
en
ak
a
Sa
n
t
h
o
s
h
a
S
,
Dep
ar
t
m
en
t o
f
E
lectr
o
n
ics
&
C
o
m
m
u
n
ica
tio
n
E
n
g
i
n
ee
r
in
g
,
C
o
o
r
g
I
n
s
ti
tu
te
o
f
T
ec
h
n
o
lo
g
y
,
Ko
d
ag
u
,
Kar
n
ata
k
a,
I
n
d
ia
.
E
m
ail:
-
b
en
a
k
a.
r
esear
ch
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
Data
m
i
n
i
n
g
is
a
w
ell
-
k
n
o
w
n
p
r
o
ce
s
s
o
f
k
n
o
w
led
g
e
d
is
co
v
er
y
a
n
d
also
e
x
p
lo
r
in
g
s
i
g
n
i
f
ica
n
t
p
atter
n
s
f
r
o
m
a
m
as
s
iv
e
s
et
o
f
d
ata.
T
h
e
ex
te
n
s
iv
e
u
s
e
o
f
d
ata
m
in
i
n
g
i
n
t
h
e
f
ield
o
f
i
n
f
o
r
m
atio
n
tech
n
o
lo
g
y
m
ad
e
i
t
an
ac
tiv
e
r
esear
c
h
ar
ea
th
u
s
s
ev
er
al
co
m
m
er
cial
p
r
o
d
u
cts
an
d
r
esear
ch
p
r
o
to
ty
p
es
ar
e
w
it
n
es
s
ed
.
T
h
e
cu
r
r
en
t
r
esear
ch
tr
en
d
s
r
ev
ea
l
a
m
att
er
o
f
f
ac
t
th
at
m
o
r
e
e
m
p
h
asi
ze
h
as
b
ee
n
p
u
t
o
n
co
r
p
o
r
at
e
d
ata
-
t
y
p
icall
y
i
n
alp
h
an
u
m
er
ic
d
ata
-
b
a
s
e
w
h
er
e
v
er
y
les
s
f
o
cu
s
h
as
b
ee
n
la
i
d
to
w
ar
d
s
m
i
n
i
n
g
o
f
m
u
lt
i
m
e
d
ia
d
ata
[
1
]
Z
aia
n
e,
Han
,
&
Z
h
u
,
2
0
0
0
)
.
Mu
lti
m
e
d
ia
d
ata
m
in
i
n
g
till
d
ate
h
as b
ee
n
co
n
ce
p
tu
a
lized
f
o
r
d
if
f
er
e
n
t t
y
p
e
s
o
f
f
ile
s
s
u
ch
as
au
d
io
,
i
m
ag
e,
an
d
v
id
eo
.
I
n
th
e
r
ec
en
t
ti
m
es,
ac
ce
s
s
ib
il
it
y
to
a
h
u
g
e
a
m
o
u
n
t
o
f
v
id
eo
co
n
ten
ts
in
b
o
t
h
in
ter
n
e
t
an
d
tele
v
is
io
n
r
eq
u
ir
e
i
m
p
licit
k
n
o
w
led
g
e
ex
tr
ac
tio
n
,
an
d
it
h
a
s
b
ec
o
m
e
a
cr
u
cia
l
task
o
w
i
n
g
to
it
s
n
o
n
-
s
tr
u
ct
u
r
ed
n
atu
r
e.
Vid
eo
an
al
y
s
is
o
r
m
i
n
i
n
g
i
s
ev
e
n
m
o
r
e
co
m
p
licate
d
task
t
h
an
a
n
al
y
zi
n
g
s
till
i
m
ag
e
s
[
2
]
,
[
3
]
.
Ho
w
e
v
er
,
a
v
id
eo
o
b
j
ec
t
co
n
s
is
ts
o
f
a
co
llectio
n
o
f
a
ti
m
e
-
o
r
d
er
ed
s
eq
u
en
ce
o
f
i
m
a
g
es
w
h
er
e
th
e
s
u
b
j
ec
t
in
ea
ch
i
m
ag
e
is
s
tat
is
ticall
y
co
r
r
elate
d
w
i
th
o
th
er
.
T
h
e
v
id
eo
co
n
te
n
t
co
n
s
i
s
ts
o
f
b
o
th
te
m
p
o
r
al
an
d
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
-
8708
A
S
y
ste
ma
ti
c
Rev
iew o
f
Existin
g
Da
ta
M
in
in
g
Ap
p
ro
a
c
h
e
s E
n
v
isi
o
n
e
d
f
o
r Kn
o
wled
g
e
…
.
(
Ben
a
k
a
S
a
n
t
h
o
sh
a
S
)
909
tex
t
u
al
i
n
f
o
r
m
atio
n
w
h
ich
cr
e
ates
co
m
p
le
x
it
y
d
u
r
i
n
g
co
m
p
u
tatio
n
a
l
m
in
in
g
o
p
er
atio
n
s
.
A
v
id
eo
co
n
te
n
t
ca
n
b
e
class
i
f
ied
in
to
th
r
ee
d
if
f
er
e
n
t
ca
te
g
o
r
ies
w
h
ich
ar
e
n
a
m
el
y
(a
)
L
o
w
-
le
v
el
f
ea
tu
r
e
i
n
f
o
r
m
atio
n
,
(
b
)
S
y
n
tactic
I
n
f
o
r
m
a
tio
n
a
n
d
(
c
)
Se
m
a
n
tic
I
n
f
o
r
m
atio
n
.
T
h
e
lo
w
-
lev
e
l
f
ea
tu
r
e
i
n
f
o
r
m
atio
n
is
s
u
b
j
ec
ted
to
co
n
ce
r
n
ab
o
u
t
th
e
f
ea
t
u
r
es o
f
a
v
id
eo
o
b
j
ec
t su
ch
a
s
co
lo
r
,
tex
t
u
r
e,
s
h
ap
e
an
d
s
o
o
n
.
S
y
n
tact
ic
i
n
f
o
r
m
atio
n
r
ef
er
s
to
d
if
f
er
en
t
s
alien
t
o
b
j
ec
ts
th
eir
s
p
atial
-
te
m
p
o
r
al
co
o
r
d
in
ates
an
d
co
r
r
elatio
n
s
.
Se
m
a
n
tic
in
f
o
r
m
at
io
n
co
n
v
e
y
s
ab
o
u
t
t
h
e
s
eq
u
en
ce
o
f
th
e
v
id
eo
f
r
o
m
ev
er
y
a
s
p
ec
t
an
d
also
d
escr
ib
es
w
h
at
is
v
is
u
all
y
p
er
ce
iv
ed
b
y
th
e
v
ie
w
er
[
4
]
,
[
5
]
.
T
h
e
f
o
llo
w
i
n
g
F
ig
u
r
e
1
ex
h
ib
i
ts
a
b
asic stru
ct
u
r
e
o
f
v
id
eo
h
i
er
ar
ch
y
f
r
o
m
a
s
eg
m
e
n
tatio
n
v
ie
w
p
o
in
t.
V
i
d
e
o
S
c
e
n
e
s
V
i
d
e
o
G
r
o
u
p
S
h
o
t
s
K
e
y
F
r
a
m
e
s
Fig
u
r
e
1
.
Vid
eo
Hier
ar
ch
y
Vid
eo
s
eg
m
e
n
tatio
n
is
a
p
ar
t
o
f
v
id
eo
p
r
o
ce
s
s
in
g
w
h
ich
i
n
v
ar
iab
l
y
d
ec
o
m
p
o
s
e
s
v
id
eo
tr
ac
k
in
to
s
m
al
ler
u
n
its
.
Ho
w
e
v
er
,
th
e
v
is
u
al
l
y
b
ased
s
e
g
m
en
tat
io
n
id
en
ti
f
ies
s
h
o
t
b
o
u
n
d
ar
ies
w
h
er
e
th
e
m
o
tio
n
b
ased
s
eg
m
e
n
tatio
n
tr
ac
k
s
d
o
w
n
th
e
p
an
s
a
n
d
zo
o
m
s
.
Ho
w
e
v
er
,
th
i
s
m
a
n
u
s
cr
ip
t
in
ten
d
s
to
co
n
tr
ib
u
te
to
w
ar
d
s
in
v
e
s
ti
n
g
th
e
ex
i
s
ti
n
g
r
esear
c
h
tr
en
d
s
ab
o
u
t
th
e
ef
f
icie
n
t
k
n
o
w
led
g
e
d
is
co
v
er
y
f
r
o
m
m
u
lti
m
ed
ia
co
n
te
n
ts
to
m
ax
i
m
ize
t
h
e
ac
c
u
r
ac
y
o
f
a
n
a
l
y
ze
d
co
n
te
n
ts
.
T
h
e
p
ap
er
is
o
r
g
a
n
ized
w
i
th
a
p
atter
n
w
h
er
e
Sectio
n
1
.
1
.
tal
k
s
ab
o
u
t
th
e
b
ac
k
g
r
o
u
n
d
o
f
Vid
e
o
Data
Min
i
n
g
co
n
ce
p
t
w
h
er
e
Sectio
n
1
.
2
.
d
is
cu
s
s
e
s
t
h
e
k
e
y
r
esear
ch
p
r
o
b
lem
s
id
en
ti
f
ied
in
th
i
s
f
ield
.
Sec
tio
n
2
h
ig
h
li
g
h
t
s
th
e
ex
is
ti
n
g
v
id
eo
m
in
i
n
g
ap
p
r
o
ac
h
es
f
r
o
m
th
eo
r
etica
l
p
er
s
p
ec
tiv
es.
Sectio
n
3
tal
k
s
a
b
o
u
t th
e
e
x
is
t
in
g
r
esear
c
h
co
n
t
r
ib
u
tio
n
s
u
s
in
g
t
h
eir
ad
d
r
ess
ed
p
r
o
b
lem
s
,
ap
p
lied
tech
n
iq
u
es
an
d
t
h
e
p
er
f
o
r
m
a
n
ce
p
ar
am
e
ter
s
co
n
s
id
er
ed
.
Fin
all
y
,
t
h
e
s
tu
d
y
ex
tr
ac
t
r
esear
c
h
g
ap
in
Sectio
n
4
af
ter
r
ev
ie
w
i
n
g
th
e
co
n
v
e
n
ti
o
n
al
ap
p
r
o
ac
h
es
an
d
th
eir
co
n
tr
ib
u
tio
n
co
n
ce
r
n
ed
f
o
llo
w
e
d
b
y
co
n
cl
u
s
io
n
in
Sectio
n
5
.
1
.
1
.
B
a
ck
g
ro
un
d o
f
Video
Da
t
a
M
ini
ng
Vid
eo
d
ata
m
in
i
n
g
d
ea
ls
w
i
th
ex
tr
ac
ti
n
g
m
ea
n
i
n
g
f
u
l
i
n
f
o
r
m
atio
n
f
r
o
m
a
v
id
eo
d
ata
o
b
j
ec
t
s
eq
u
en
c
e
co
n
s
id
er
in
g
an
i
m
p
licit
k
n
o
w
l
ed
g
e
d
is
co
v
er
y
p
r
o
ce
s
s
.
Vis
u
a
l
in
ter
p
r
etatio
n
o
f
m
ea
n
i
n
g
f
u
l
p
atter
n
s
in
a
v
id
eo
f
r
a
m
e
s
eq
u
e
n
ce
is
q
u
ite
a
ch
allen
g
in
g
tas
k
as
v
id
e
o
o
b
j
ec
t
co
m
p
r
is
e
s
co
m
p
le
x
d
if
f
er
en
t
p
atter
n
s
o
f
s
e
m
i
-
s
tr
u
ct
u
r
ed
an
d
u
n
s
tr
u
ctu
r
ed
d
ata.
I
t
also
i
n
cl
u
d
es
p
atter
n
d
i
s
co
v
er
y
p
r
o
ce
s
s
w
h
ile
p
atter
n
s
ar
e
id
en
tifia
b
le
i
n
v
id
eo
d
atab
ases
[
6
]
.
Ho
w
e
v
er
,
p
atter
n
d
is
co
v
er
y
i
n
v
id
eo
d
atab
ases
p
er
f
o
r
m
ed
co
n
s
id
er
in
g
a
n
ex
te
n
s
io
n
o
f
s
till
i
m
ag
e
m
i
n
in
g
f
o
llo
w
ed
b
y
m
i
n
i
n
g
o
f
te
m
p
o
r
al
i
m
a
g
e
s
eq
u
en
ce
s
[
7
]
.
T
h
e
p
r
o
ce
s
s
also
n
o
t
o
n
l
y
m
ea
n
t
to
ex
tr
ac
t
co
n
te
n
t,
s
tr
u
ctu
r
e,
t
h
e
s
p
atial
o
r
te
m
p
o
r
al
co
r
r
elatio
n
b
et
w
ee
n
m
o
v
in
g
o
b
j
ec
ts
o
f
v
id
eo
co
n
ten
t
r
ath
er
it e
m
p
h
a
s
izes
m
o
r
e
o
n
ex
tr
ac
ti
n
g
p
atter
n
s
co
n
ce
r
n
in
g
o
b
j
ec
t a
ctiv
itie
s
an
d
e
v
en
t
s
f
r
o
m
a
v
a
s
t a
m
o
u
n
t o
f
v
id
eo
d
ata.
T
h
er
e
ex
is
t c
er
tain
d
is
s
i
m
ilar
ities
w
h
ich
m
a
k
es
v
id
eo
d
ata
m
i
n
in
g
a
u
n
iq
u
e
f
r
o
m
o
t
h
er
r
elate
d
ar
ea
s
.
a.
Vid
eo
Data
Min
in
g
V
s
.
Vid
eo
P
r
o
ce
s
s
in
g
:
T
h
e
r
elatio
n
s
h
ip
b
et
w
ee
n
v
id
eo
p
r
o
ce
s
s
in
g
an
d
Vid
eo
d
ata
an
al
y
tic
s
is
q
u
ite
s
u
b
j
ec
tiv
e
f
r
o
m
d
i
f
f
er
en
t
co
n
tex
ts
.
Vi
d
eo
d
ata
m
i
n
in
g
r
e
f
er
s
to
th
e
p
r
o
ce
s
s
o
f
ex
tr
ac
ti
n
g
m
ea
n
in
g
f
u
l
p
atter
n
s
f
r
o
m
a
v
id
eo
s
eq
u
en
ce
w
h
ile
v
id
eo
p
r
o
ce
s
s
in
g
f
o
cu
s
es
o
n
m
o
s
tl
y
f
ea
tu
r
e
e
x
tr
ac
tio
n
.
b.
Vid
eo
Dat
a
Mi
n
i
n
g
Vs
P
atter
n
R
ec
o
g
n
itio
n
:
B
o
th
t
h
e
ar
ea
s
ar
e
i
n
cli
n
ed
i
n
to
f
ea
t
u
r
e
e
x
t
r
ac
tio
n
s
tep
s
b
u
t
t
h
e
v
id
eo
d
ata
m
i
n
i
n
g
d
if
f
er
s
i
n
ter
m
s
o
f
p
atter
n
s
p
ec
if
icit
y
r
ec
o
g
n
itio
n
,
a
n
d
P
atter
n
r
ec
o
g
n
itio
n
d
ea
ls
w
it
h
cla
s
s
i
f
y
in
g
s
p
ec
ial
s
a
m
p
les
w
it
h
t
h
e
h
elp
o
f
e
x
i
s
ti
n
g
m
o
d
el
w
h
i
le
v
id
eo
m
i
n
in
g
in
d
u
l
g
in
g
in
to
a
s
t
u
d
y
w
h
ic
h
p
er
f
o
r
m
s
d
etec
tin
g
o
f
r
u
les
a
n
d
p
atter
n
s
ir
r
esp
ec
ti
v
e
o
f
a
n
y
v
id
eo
p
r
o
ce
s
s
in
g
o
p
er
atio
n
s
.
c.
Vid
eo
Data
Min
in
g
V
s
.
Vid
eo
I
n
f
o
r
m
atio
n
R
etr
ie
v
al:
T
h
e
d
i
f
f
er
en
ce
i
n
t
h
i
s
co
n
te
x
t
is
v
er
y
m
u
c
h
s
i
m
ilar
to
th
e
d
i
f
f
er
en
ce
t
h
at
ex
is
t
s
b
et
w
ee
n
t
h
e
tr
ad
itio
n
al
d
ata
b
ase
m
a
n
ag
e
m
e
n
t
s
y
s
te
m
s
a
n
d
th
e
d
ata
m
i
n
in
g
[
8
]
.
T
h
e
p
r
im
e
o
b
j
ec
tiv
e
o
f
v
id
eo
m
in
in
g
is
to
f
in
d
o
u
t
co
r
r
elatio
n
an
d
p
atter
n
s
w
h
ic
h
ar
e
y
et
to
u
n
d
er
s
ta
n
d
f
r
o
m
a
s
et
o
f
v
id
eo
d
ata
b
ases
.
Vid
eo
m
in
i
n
g
p
er
f
o
r
m
s
in
f
o
r
m
atio
n
r
etr
iev
al
f
r
o
m
t
h
e
v
id
eo
d
at
ab
ases
an
d
f
u
r
t
h
er
p
er
f
o
r
m
s
m
i
n
i
n
g
o
p
er
atio
n
s
to
r
ec
o
g
n
ize
t
h
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
2
,
A
p
r
il
201
8
:
9
0
8
–
9
1
6
910
p
atter
n
s
an
d
tr
en
d
s
w
h
er
e
v
id
eo
s
cr
ip
tin
g
p
la
y
s
a
s
i
g
n
i
f
ica
n
t
r
o
le.
T
h
e
f
o
llo
w
i
n
g
F
i
g
u
r
e
2
s
h
o
w
s
a
g
en
er
al
f
r
a
m
e
w
o
r
k
i
n
te
n
d
ed
to
r
ep
r
esen
t o
v
er
all
v
id
eo
d
ata
m
in
in
g
o
p
er
atio
n
s
.
V
i
d
e
o
f
r
a
m
e
S
e
q
u
e
n
c
e
S
t
a
g
e
:
1
G
r
o
u
p
i
n
g
F
r
a
m
e
s
t
o
S
e
g
m
e
n
t
s
S
t
a
g
e
:
2
F
e
a
t
u
r
e
E
x
t
r
a
c
t
i
o
n
S
t
a
g
e
:
3
I
n
d
e
x
i
n
g
a
n
d
C
l
u
s
t
e
r
i
n
g
S
t
a
g
e
:
4
V
i
d
e
o
d
a
t
a
m
i
n
i
n
g
S
t
a
g
e
:
5
V
i
d
e
o
d
a
t
a
c
o
m
p
r
e
s
s
i
o
n
D
A
T
A
B
A
S
E
M
e
t
a
D
a
t
a
a
n
d
K
n
o
w
l
e
d
g
e
B
a
s
e
M
e
t
a
D
a
t
a
a
n
d
K
n
o
w
l
e
d
g
e
B
a
s
e
Fig
u
r
e
2
.
An
Ov
er
v
ie
w
o
f
Vid
eo
Data
Min
in
g
T
h
e
ab
o
v
e
-
s
tated
f
i
g
u
r
e
ex
h
ib
it
th
a
t
h
o
w
m
u
l
ti
m
ed
ia
d
ata
m
i
n
in
g
r
ei
n
f
o
r
ce
s
f
ea
tu
r
e
ex
tr
ac
tio
n
a
n
d
clu
s
ter
i
n
g
p
r
o
ce
s
s
to
f
in
d
a
s
ig
n
i
f
ica
n
t
p
atter
n
f
r
o
m
m
u
lti
m
ed
ia
d
ata.
Sev
er
al
s
tu
d
ies
ar
e
f
o
u
n
d
to
talk
ab
o
u
t
th
e
ar
ch
itect
u
r
al
p
r
o
ce
s
s
f
o
r
m
u
lti
m
ed
ia
d
ata
m
in
i
n
g
w
h
ic
h
in
v
o
lv
e
s
th
r
ee
d
if
f
er
en
t
p
r
i
m
e
tas
k
s
.
First
l
y
it
f
o
cu
s
es
o
n
p
r
e
-
p
r
o
ce
s
s
i
n
g
o
f
v
id
eo
d
ata
o
b
j
ec
ts
w
h
ic
h
co
m
p
r
i
s
es
p
ix
el
s
,
k
e
y
f
r
a
m
es,
s
eg
m
e
n
t
s
,
s
ce
n
e,
etc.
s
ec
o
n
d
l
y
it
in
v
o
lv
e
s
e
x
tr
ac
tio
n
o
f
d
i
f
f
er
en
t
t
y
p
e
s
o
f
f
ea
t
u
r
es
s
u
c
h
a
s
p
h
y
s
ical,
m
o
tio
n
,
r
elat
io
n
f
ea
t
u
r
es
f
r
o
m
a
v
id
eo
o
b
j
ec
t
w
h
ic
h
h
elp
s
f
u
r
t
h
er
in
k
n
o
w
led
g
e
d
is
co
v
er
y
an
d
p
atter
n
r
ec
o
g
n
itio
n
[
9
]
,
[
1
0
]
.
1
.
2
.
K
ey
Resea
rc
h
P
ro
ble
m
s
in V
ideo
M
ini
ng
T
h
is
s
ec
tio
n
b
r
ie
f
l
y
tal
k
s
ab
o
u
t th
e
e
x
is
tin
g
r
esear
c
h
p
r
o
b
lem
s
a
s
s
o
ciate
d
w
it
h
t
h
e
o
p
er
ati
o
n
al
d
esi
g
n
asp
ec
ts
o
f
v
id
eo
m
i
n
i
n
g
alg
o
r
ith
m
s
a
n
d
t
h
e
p
r
o
b
le
m
e
n
co
u
n
ter
ed
i
n
c
u
r
r
en
t
r
esear
ch
tr
ac
k
as
w
ell.
Vid
eo
m
i
n
in
g
o
f
ten
r
ef
er
r
ed
as
an
e
m
er
g
i
n
g
f
ie
ld
o
f
v
id
eo
an
al
y
t
ics
w
h
ic
h
i
n
f
lu
e
n
ce
s
t
h
e
f
u
t
u
r
is
tic
d
ata
s
cien
ce
f
r
o
m
d
i
f
f
er
e
n
t
asp
ec
t
s
.
T
h
e
u
n
s
u
p
er
v
i
s
ed
lear
n
i
n
g
o
f
d
i
f
f
er
e
n
t
a
u
d
io
-
v
is
u
al
p
atter
n
s
m
ak
e
s
it
an
o
p
er
atio
n
all
y
ch
alle
n
g
i
n
g
p
r
o
ce
s
s
.
Vid
eo
d
ata
m
i
n
i
n
g
an
d
d
ata
m
a
n
ag
e
m
e
n
t
o
p
en
u
p
a
n
e
w
er
a
o
f
s
m
ar
t
ap
p
licatio
n
s
w
h
ic
h
in
cl
u
d
es
in
tel
lig
e
n
t
co
n
te
n
t
f
ilter
s
,
s
u
r
v
eilla
n
ce
,
p
er
s
o
n
al
v
id
eo
r
ec
o
m
m
e
n
d
atio
n
,
o
r
c
o
n
ten
t
-
b
ased
ad
v
er
tis
e
m
en
t.
T
h
e
co
r
e
ch
allen
g
e
s
ar
e
to
p
r
ed
ict
s
em
a
n
tic
f
ea
t
u
r
es
f
r
o
m
p
r
i
m
i
tiv
e
f
ea
tu
r
es.
T
h
er
e
s
h
o
u
ld
b
e
a
g
en
er
alize
d
f
r
a
m
e
w
o
r
k
w
h
i
ch
ca
n
h
av
e
h
i
g
h
e
f
f
icie
n
c
y
o
n
d
etec
tin
g
s
e
m
an
tic
f
ea
t
u
r
e
s
f
r
o
m
g
en
er
al
v
id
e
co
n
ten
t
s
an
d
f
u
r
t
h
er
ap
p
ly
t
h
a
t
to
an
y
t
y
p
e
s
o
f
v
id
eo
s
.
T
h
e
p
r
i
m
e
ch
alle
n
g
e
is
to
b
u
ild
u
p
a
f
r
a
m
e
w
o
r
k
w
h
ic
h
en
s
u
r
es e
f
f
ic
ien
t e
x
tr
ac
tio
n
o
f
m
u
ltip
le
s
e
m
a
n
tics
f
r
o
m
t
h
e
v
i
d
eo
s
w
it
h
th
e
u
s
e
o
f
p
r
i
m
iti
v
e
f
ea
t
u
r
es [
1
1
]
.
1
.
3
.
Co
nv
ent
io
na
l V
ideo
Da
t
a
M
i
nin
g
Appro
a
ches
T
h
e
cu
r
r
en
t
r
esear
ch
tr
en
d
s
ar
e
b
ein
g
w
it
n
e
s
s
ed
b
y
e
m
p
lo
y
i
n
g
v
ar
io
u
s
v
id
eo
d
ata
m
i
n
i
n
g
tech
n
iq
u
e
s
to
a
lar
g
e
ex
ten
t.
T
h
e
p
r
im
e
g
o
al
o
f
ev
er
y
m
in
i
n
g
tech
n
iq
u
e
is
to
ex
tr
ac
t
s
ig
n
i
f
ica
n
t
k
n
o
w
led
g
e
f
r
o
m
t
h
e
v
id
eo
d
atab
ases
v
er
y
e
f
f
ic
ie
n
tl
y
an
d
w
i
th
i
n
a
s
h
o
r
t
p
er
i
o
d
.
Ho
w
e
v
er
,
th
e
ex
tr
ac
ted
d
ata
s
h
o
u
ld
p
r
o
v
id
e
m
ax
i
m
u
m
ac
c
u
r
ac
y
i
n
a
n
al
y
z
ed
co
n
ten
t
s
.
Si
n
ce
m
a
n
y
y
ea
r
s
v
ar
io
u
s
v
id
eo
d
ata
m
i
n
i
n
g
a
p
p
r
o
ac
h
es
ar
e
b
ein
g
p
r
o
p
o
s
ed
w
h
ich
ca
n
b
e
cr
u
d
el
y
clas
s
i
f
ied
i
n
to
t
h
e
m
aj
o
r
f
iv
e
ca
teg
o
r
ies
s
u
c
h
a
s
1
)
Vid
eo
p
atter
n
m
i
n
in
g
[
1
2
]
,
2
)
Vid
eo
clu
s
ter
i
n
g
a
n
d
cla
s
s
if
i
ca
tio
n
,
3
)
Vid
eo
as
s
o
ciatio
n
m
i
n
i
n
g
,
4
)
Vid
eo
co
n
te
n
t
s
t
r
u
ctu
r
e
m
i
n
i
n
g
a
n
d
f
i
n
all
y
5
)
Vid
eo
m
o
tio
n
m
i
n
i
n
g
.
A
b
r
ief
d
is
cu
s
s
io
n
o
f
t
h
ese
v
id
eo
d
ata
m
in
i
n
g
ap
p
r
o
ac
h
es is
g
i
v
e
n
b
elo
w
:
1
.
4
.
Video
P
a
t
t
er
n
M
ini
ng
T
h
is
p
r
o
ce
s
s
ai
m
s
to
d
etec
t
v
a
r
io
u
s
s
p
atial
p
at
ter
n
s
m
o
d
eled
in
ad
v
a
n
ce
w
i
th
i
n
a
v
id
eo
o
b
ject.
W
h
at
a
s
et
o
f
s
eq
u
en
t
ial
c
h
ar
ac
ter
ized
ev
en
t
s
s
u
c
h
a
s
d
ialo
g
u
e
o
r
p
r
esen
tatio
n
i
m
ag
e
b
elo
n
g
s
to
a
m
ed
ical
v
id
e
o
i
s
r
ef
er
r
ed
in
th
is
co
n
te
x
t.
T
h
e
ex
is
t
in
g
v
id
eo
p
atter
n
m
i
n
i
n
g
tech
n
iq
u
es
ar
e
clu
s
ter
ed
in
t
o
tw
o
p
r
i
m
e
ca
teg
o
r
ie
s
s
u
c
h
as (
a
)
m
i
n
i
n
g
s
i
m
ila
r
m
o
t
io
n
p
atter
n
s
a
n
d
s
ec
o
n
d
l
y
(
b
)
m
i
n
in
g
s
i
m
ilar
o
b
j
ec
ts
[
1
3
]
.
1
.
5
.
Video
Clus
t
er
i
ng
a
nd
Cla
s
s
if
ica
t
io
n
I
t is a
p
r
o
ce
s
s
o
f
cl
u
s
ter
in
g
a
n
d
class
i
f
y
in
g
t
h
e
v
id
eo
u
n
it
s
c
o
n
ce
r
n
i
n
g
d
if
f
er
en
t
ca
teg
o
r
ies.
Ho
w
e
v
er
,
clu
s
ter
i
n
g
m
ea
n
t
f
o
r
p
er
f
o
r
m
i
n
g
a
n
u
n
s
u
p
er
v
i
s
ed
lear
n
i
n
g
t
o
d
is
co
v
er
ce
r
tain
s
i
g
n
if
ica
n
t
k
n
o
w
led
g
e
f
r
o
m
a
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
-
8708
A
S
y
ste
ma
ti
c
Rev
iew o
f
Existin
g
Da
ta
M
in
in
g
Ap
p
ro
a
c
h
e
s E
n
v
isi
o
n
e
d
f
o
r Kn
o
wled
g
e
…
.
(
Ben
a
k
a
S
a
n
t
h
o
sh
a
S
)
911
d
ataset.
T
h
e
ex
ten
s
i
v
e
cl
u
s
ter
in
g
p
r
o
ce
s
s
d
u
r
i
n
g
v
id
eo
m
i
n
in
g
in
c
lu
d
es
p
er
f
o
r
m
s
a
s
eq
u
en
tial
o
p
er
atio
n
s
o
f
(a
)
s
h
o
t
d
etec
tio
n
,
(
b
)
Key
f
r
a
m
e
ex
tr
ac
ti
o
n
,
(
c
)
f
ea
tu
r
e
ex
tr
ac
ti
o
n
an
d
s
i
m
ilar
it
y
m
ea
s
u
r
e
m
en
t,
(
d
)
clu
s
te
r
in
g
o
f
k
e
y
f
r
a
m
e
s
f
o
llo
w
ed
b
y
(
e
)
s
e
m
a
n
tic
in
ter
p
r
etatio
n
an
d
q
u
er
y
r
etr
iev
al.
T
h
e
clu
s
ter
i
n
g
h
as
a
p
o
ten
tial
o
f
ex
tr
ac
ti
n
g
m
ea
n
i
n
g
f
u
l
p
atter
n
s
f
r
o
m
a
v
id
eo
s
eq
u
e
n
ce
w
h
i
ch
h
av
e
a
s
i
g
n
if
ica
n
t
i
m
p
ac
t
o
n
d
if
f
er
en
t
asp
ec
t
s
s
u
c
h
as
in
d
e
x
i
n
g
,
s
u
r
v
e
illan
ce
,
ac
tiv
it
y
d
is
co
v
e
r
y
a
n
d
ev
en
t
r
ec
o
g
n
itio
n
[
1
4
]
.
A
p
r
o
ce
s
s
o
f
v
id
eo
clu
s
ter
i
n
g
i
s
d
ep
icted
b
elo
w
:
V
i
d
e
o
f
r
a
m
e
s
e
q
u
e
n
c
e
S
h
o
t
D
e
t
e
c
t
i
o
n
K
e
y
f
r
a
m
e
e
x
t
r
a
c
t
i
o
n
F
e
a
t
u
r
e
E
x
t
r
a
c
t
i
o
n
S
i
m
i
l
a
r
i
t
y
C
o
m
p
u
t
a
t
i
o
n
C
l
u
s
t
e
r
i
n
g
o
f
K
e
y
f
r
a
m
e
s
S
e
m
a
n
t
i
c
I
n
t
e
r
p
r
e
t
a
t
i
o
n
Q
u
e
r
y
a
n
d
R
e
t
r
i
e
v
a
l
Fig
u
r
e
3
.
Fu
n
c
tio
n
al
b
lo
ck
s
o
f
Vid
eo
C
lu
s
ter
i
n
g
P
r
o
ce
s
s
es
C
lu
s
ter
i
n
g
a
n
d
class
i
f
icatio
n
an
al
y
s
is
i
n
as
s
o
c
iatio
n
w
it
h
ab
o
v
e
h
ig
h
li
g
h
ted
(
F
ig
u
r
e
3
)
f
u
n
ctio
n
al
b
lo
ck
s
ex
cl
u
s
i
v
el
y
tr
ies
to
f
i
n
d
th
e
u
n
iq
u
e
p
atter
n
s
o
f
m
o
v
in
g
o
b
j
ec
t
in
a
v
id
eo
s
eq
u
en
ce
.
C
l
u
s
ter
i
n
g
o
f
s
i
m
ilar
s
h
o
ts
in
a
m
o
v
i
n
g
s
eq
u
en
ce
o
f
f
r
a
m
es
p
r
o
d
u
ce
s
m
o
r
e
p
r
ec
is
e
v
id
eo
o
b
j
ec
t
w
it
h
l
ess
r
ed
u
n
d
an
c
y
a
n
d
n
o
is
y
e
n
v
ir
o
n
m
e
n
t.
T
h
er
e
ar
e
s
e
v
er
al
cl
u
s
ter
i
n
g
alg
o
r
ith
m
s
,
w
h
ich
ar
e
ca
teg
o
r
ized
in
to
m
p
ar
titi
o
n
in
g
m
et
h
o
d
s
,
h
ier
ar
ch
ical
m
et
h
o
d
s
,
d
en
s
it
y
-
b
a
s
ed
m
et
h
o
d
s
,
an
d
g
r
id
b
ased
m
et
h
o
d
s
an
d
m
o
d
el
-
b
ased
m
et
h
o
d
s
.
1
.
6
.
Video
Ass
o
cia
t
io
n M
ini
ng
I
t
r
ef
er
s
to
th
e
o
p
er
atio
n
s
w
h
ich
i
n
v
o
lv
e
d
i
s
co
v
er
i
n
g
ass
o
c
iatio
n
s
th
a
t
e
x
is
t
a
m
o
n
g
v
id
eo
f
r
a
m
e
s
.
T
h
is
tech
n
iq
u
e
ex
tr
ac
ts
t
h
e
k
n
o
w
led
g
e
f
r
o
m
a
v
id
eo
s
eq
u
en
ce
b
y
i
n
co
r
p
o
r
atin
g
t
w
o
d
if
f
er
e
n
t
f
u
n
ctio
n
a
l
s
tag
e
s
w
h
er
e
i
n
Stag
e
-
I
t
h
e
v
i
d
eo
co
n
ten
t
is
s
eg
m
e
n
ted
i
n
to
ce
r
tain
u
n
it
s
w
h
er
e
a
n
a
n
al
y
s
is
f
u
r
t
h
er
ca
r
r
y
o
u
t
to
ex
tr
ac
t
s
i
g
n
if
ica
n
t
f
ea
t
u
r
e
s
o
r
d
ata
p
atter
n
s
.
I
n
Sta
g
e
-
I
I
th
e
o
p
er
atio
n
al
u
n
i
ts
p
er
f
o
r
m
co
n
s
ec
u
tiv
e
ass
o
ciatio
n
m
i
n
i
n
g
to
ex
tr
ac
t m
ea
n
in
g
f
u
l k
n
o
w
led
g
e
f
r
o
m
f
ea
tu
r
e
d
escr
ip
t
o
r
s
[
1
5
]
.
1
.
7
.
Video
Co
nte
nt
Str
uct
ure
M
i
nin
g
As
v
id
eo
co
n
te
n
ts
co
m
p
r
is
i
n
g
o
f
co
m
p
le
x
u
n
s
tr
u
c
tu
r
ed
d
ata
p
atter
n
s
th
u
s
it
m
a
k
es
ac
ce
s
s
to
v
id
eo
co
n
ten
t
i
n
a
d
atab
ase
co
m
p
u
ta
tio
n
all
y
c
h
alle
n
g
in
g
.
T
o
m
ak
e
th
is
o
p
er
atio
n
ef
f
icien
t
th
e
p
r
i
m
e
o
b
j
ec
tiv
e
is
to
co
n
v
er
t
u
n
s
tr
u
ct
u
r
ed
d
ata
p
atter
n
s
i
n
to
s
tr
u
c
t
u
r
ed
p
atter
n
s
(
Fig
u
r
e
1
)
u
s
in
g
v
id
eo
s
tr
u
ctu
r
e
m
i
n
i
n
g
.
T
h
e
s
y
n
t
h
etic
lev
e
l
co
m
p
o
s
itio
n
wh
ich
d
etec
t
s
s
tr
u
ct
u
r
ed
p
atter
n
s
f
r
o
m
a
v
id
eo
d
ata
m
ak
e
s
r
an
d
o
m
ac
ce
s
s
to
th
e
co
n
ten
t
s
f
a
s
ter
.
I
t a
ls
o
d
ef
i
n
es
th
e
f
u
n
d
a
m
e
n
ta
l lo
g
ic
s
tr
u
ctu
r
e
w
it
h
i
n
a
v
id
eo
o
b
j
ec
t c
o
n
s
titu
ti
n
g
a
r
elatio
n
s
h
ip
b
et
w
ee
n
s
e
m
a
n
tic
m
u
lt
i
-
m
o
d
a
lit
y
co
n
ce
p
ts
an
d
co
m
p
u
ter
lo
w
-
le
v
el
f
ea
t
u
r
es.
T
h
e
f
o
llo
w
i
n
g
F
i
g
u
r
e
4
s
h
o
w
s
h
o
w
s
e
m
an
t
ic
co
n
ce
p
ts
ar
e
d
etec
ted
u
tili
zi
n
g
a
m
u
lti
m
o
d
al
co
n
ten
t a
n
al
y
s
i
s
an
d
A
-
p
r
io
r
i a
lg
o
r
ith
m
.
V
i
d
e
o
S
t
r
e
a
m
F
e
a
t
u
r
e
E
x
t
r
a
c
t
a
n
d
N
o
r
m
a
l
i
z
a
t
i
o
n
T
r
a
i
n
i
n
g
D
a
t
a
D
i
s
c
r
e
t
i
z
a
t
i
o
n
T
e
s
t
i
n
g
D
a
t
a
C
o
r
r
e
l
a
t
i
o
n
P
e
r
c
e
n
t
a
g
e
I
n
f
o
r
m
a
t
i
o
n
V
i
d
e
o
S
t
r
e
a
m
D
e
t
e
c
t
e
d
C
o
n
c
e
p
t
s
C
l
a
s
s
i
f
i
c
a
t
i
o
n
F
ig
u
re
4
.
A
p
r
o
c
e
ss
o
f
se
m
a
n
ti
c
c
o
n
c
e
p
t
d
e
tec
ti
o
n
[
1
6
]
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
2
,
A
p
r
il
201
8
:
9
0
8
–
9
1
6
912
1
.
8
.
Video
M
o
t
io
n
M
ini
ng
I
t
is
o
n
e
o
f
th
e
m
o
s
t
p
r
o
m
i
n
en
t
m
i
n
in
g
tec
h
n
iq
u
es,
w
h
ic
h
ch
ar
ac
ter
izes
th
e
m
o
tio
n
o
f
o
b
j
e
cts
f
r
o
m
a
v
id
eo
s
eq
u
en
ce
.
T
h
e
m
o
tio
n
co
n
tain
s
te
m
p
o
r
al
in
f
o
r
m
at
io
n
,
an
d
it
is
q
u
ite
c
h
alle
n
g
in
g
to
r
etr
iev
e
v
is
u
a
l
in
f
o
r
m
atio
n
f
r
o
m
th
e
m
o
t
io
n
o
b
j
ec
t
r
eg
io
n
s
.
T
h
e
ca
m
er
a
m
o
tio
n
also
p
o
s
es
s
o
m
e
k
e
y
i
s
s
u
es
in
v
id
eo
m
o
tio
n
d
etec
tio
n
s
s
u
ch
a
s
ca
m
er
a
lo
c
ated
in
a
s
tatic
p
o
s
itio
n
w
h
ile
o
b
j
ec
ts
ar
e
m
o
v
i
n
g
,
th
e
ca
m
e
r
a
is
m
o
v
i
n
g
w
h
ile
o
b
j
ec
ts
ar
e
also
m
o
v
in
g
a
n
d
a
s
et
o
f
ca
m
er
a
s
r
ec
o
r
d
in
g
th
e
s
a
m
e
o
b
j
ec
ts
[
1
7
]
.
T
h
e
n
ex
t
s
ec
tio
n
f
u
r
t
h
er
h
ig
h
li
g
h
ts
a
s
u
m
m
ar
y
o
f
e
x
is
t
in
g
r
esear
ch
co
n
tr
ib
u
tio
n
s
u
s
i
n
g
th
eir
ad
d
r
ess
ed
p
r
o
b
lem
s
,
ap
p
lied
tech
n
iq
u
e
s
an
d
th
e
p
er
f
o
r
m
a
n
ce
p
ar
a
m
ete
r
s
co
n
s
id
er
ed
.
2.
RE
VI
E
W
O
F
L
I
T
E
RA
T
UR
E
S
A
p
ar
t
f
r
o
m
t
h
e
b
elo
w
m
e
n
ti
o
n
ed
s
tu
d
ies,
t
h
e
f
o
llo
w
in
g
ar
e
f
e
w
o
f
t
h
e
ex
is
ti
n
g
liter
a
tu
r
e
f
o
u
n
d
r
elev
an
t
to
th
e
v
id
eo
m
i
n
i
n
g
an
d
k
n
o
w
led
g
e
ex
tr
ac
tio
n
.
T
h
e
s
tu
d
y
o
f
W
lo
d
ar
cz
ak
[
3
6
]
r
ev
ie
w
s
f
e
w
o
f
th
e
s
ig
n
i
f
ica
n
t
d
ee
p
lear
n
in
g
m
et
h
o
d
s
,
r
ec
en
tl
y
ad
o
p
ted
in
to
t
h
e
s
co
p
e
o
f
m
u
lti
m
ed
ia
an
d
v
i
d
eo
d
ata
m
i
n
in
g
.
I
n
th
e
s
t
u
d
y
o
f
K
u
m
ar
[
3
7
]
,
a
m
e
th
o
d
th
at
e
f
f
icie
n
tl
y
clas
s
i
f
ies
an
i
m
al
o
b
j
ec
ts
f
r
o
m
i
m
a
g
es
h
as
b
ee
n
i
n
tr
o
d
u
ce
d
.
T
h
e
s
tu
d
y
al
s
o
e
m
p
h
asized
o
n
eli
m
i
n
ati
n
g
t
h
e
b
ac
k
g
r
o
u
n
d
o
b
j
ec
ts
f
r
o
m
a
g
i
v
e
n
i
m
a
g
e
b
y
u
s
i
n
g
a
g
r
ap
h
cu
t
b
ased
tech
n
iq
u
e.
F
u
r
th
er
,
t
h
e
s
eg
m
e
n
ted
an
i
m
al
i
m
a
g
es
ar
e
d
iv
id
ed
in
to
b
lo
ck
s
a
n
d
p
ass
ed
th
r
o
u
g
h
a
p
r
o
ce
s
s
in
g
w
h
er
e
co
lo
r
te
x
tu
r
e
m
o
m
e
n
t
s
ar
e
f
u
r
t
h
er
ex
tr
ac
t
ed
b
y
r
e
f
er
r
in
g
ea
c
h
b
lo
ck
s
e
g
m
e
n
t.
Fu
r
t
h
er
,
t
h
e
co
n
ce
p
ts
o
f
p
r
o
b
ab
ilis
tic
n
eu
r
al
n
et
w
o
r
k
s
an
d
K
-
n
ea
r
es
t
n
ei
g
h
b
o
r
s
ar
e
tak
en
i
n
to
co
n
s
id
er
atio
n
to
p
er
f
o
r
m
th
e
class
i
f
icatio
n
.
S
h
a
h
b
az
et
a
l.
[
3
8
]
co
n
ce
p
tu
alize
d
a
n
o
v
el
ap
p
r
o
ac
h
w
el
l
ca
p
ab
le
o
f
c
lass
i
f
y
in
g
s
atelli
te
i
m
a
g
es.
I
t
also
o
b
j
ec
tif
ies
d
i
f
f
er
en
t
u
n
iq
u
e
p
atter
n
s
f
r
o
m
s
atellite
i
m
ag
e
s
v
er
y
e
f
f
ic
ien
t
l
y
.
A
n
e
x
p
er
i
m
e
n
tal
test
b
ed
h
as
b
ee
n
cr
ea
ted
to
v
a
l
id
ate
th
e
p
er
f
o
r
m
an
ce
e
f
f
icie
n
c
y
o
f
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
w
h
er
e
th
e
m
et
h
o
d
h
as
b
ee
n
test
ed
w
it
h
a
s
et
o
f
4
2
s
atellite
i
m
a
g
es.
T
h
e
p
er
f
o
r
m
an
ce
an
al
y
s
is
f
u
r
th
er
s
h
o
ws
th
at
th
e
p
r
o
p
o
s
ed
tech
n
iq
u
e
o
u
tp
er
f
o
r
m
s
th
e
co
n
v
e
n
tio
n
al
v
id
eo
m
i
n
i
n
g
ap
p
r
o
ac
h
es
w
it
h
an
ac
cu
r
ac
y
o
f
8
0
%.
T
h
e
s
tu
d
y
o
f
B
en
o
it
[
3
9
]
m
o
s
tl
y
e
m
p
h
a
s
ized
o
n
i
m
p
r
o
v
i
n
g
th
e
p
er
f
o
r
m
a
n
ce
co
n
ce
p
t
d
etec
to
r
s
(
C
D)
an
d
f
u
r
t
h
er
co
n
ce
p
tu
alize
d
an
a
n
al
y
tical
m
o
d
el
b
y
i
m
p
r
o
v
is
i
n
g
th
e
q
u
alit
y
a
s
p
ec
ts
o
f
s
e
m
i
-
s
u
p
er
v
is
ed
lear
n
in
g
en
ab
led
r
ef
in
e
m
e
n
t
f
r
a
m
e
w
o
r
k
.
T
h
e
s
tu
d
y
e
m
p
lo
y
ed
a
s
el
f
-
tr
ain
i
n
g
p
ar
ad
ig
m
c
ap
ab
le
o
f
ex
p
an
d
in
g
t
h
e
tr
ain
i
n
g
d
ataset
w
it
h
a
u
to
m
atica
ll
y
lab
eled
d
ata.
T
h
e
n
u
m
er
ical
an
al
y
s
i
s
a
n
d
s
i
m
u
latio
n
al
s
o
ex
h
ib
ited
th
e
p
r
i
m
e
r
o
le
o
f
ex
tr
ac
ted
v
is
u
al
f
ea
t
u
r
es
an
d
tex
t
m
etad
ata
to
en
h
a
n
ce
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
co
n
ce
p
t
class
if
ier
s
co
n
ce
r
n
i
n
g
d
i
f
f
er
e
n
t
u
n
lab
ele
d
v
id
eo
o
b
j
ec
ts
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
t
h
e
p
r
o
p
o
s
ed
m
eth
o
d
h
as
b
ee
n
v
alid
ated
co
n
s
id
er
in
g
a
d
ataset
co
m
p
r
is
in
g
2
1
,
0
0
0
en
titi
es
w
h
ic
h
s
h
o
w
s
e
x
p
an
d
i
n
g
t
h
e
tr
ain
i
n
g
s
et
w
it
h
lab
eled
s
h
o
ts
an
d
C
D
s
i
g
n
i
f
i
ca
n
tl
y
ex
ce
ls
b
etter
o
u
tco
m
e
s
.
R
esear
ch
i
n
s
i
m
ilar
d
ir
ec
t
io
n
p
er
f
o
r
m
ed
b
y
C
ao
an
d
W
a
n
g
[
4
0
]
w
h
er
e
th
e
a
u
t
h
o
r
s
p
r
esen
ted
a
n
o
v
el
co
n
ce
p
t
w
h
ic
h
is
s
u
b
j
ec
ted
to
en
h
an
ce
th
e
p
er
f
o
r
m
an
c
e
o
f
i
m
a
g
e
m
in
in
g
m
o
d
eli
n
g
.
T
h
e
o
u
tco
m
es
o
b
t
ain
ed
af
ter
s
i
m
u
lati
n
g
th
e
p
r
o
p
o
s
ed
tech
n
iq
u
e
co
n
v
e
y
ed
its
s
u
p
er
io
r
ity
a
s
co
m
p
ar
ed
to
o
th
er
im
a
g
e
m
i
n
in
g
s
c
h
e
m
es.
I
n
t
h
e
s
tu
d
y
o
f
Dev
ase
n
a
an
d
He
m
ala
th
a
[
4
1
]
,
a
n
o
v
el
ap
p
r
o
ac
h
f
o
r
v
id
eo
an
al
y
tic
s
h
as
b
ee
n
in
tr
o
d
u
ce
d
.
T
h
e
tech
n
iq
u
e
i
n
teg
r
ated
w
it
h
a
u
n
iq
u
e
L
I
M
b
ased
clu
s
ter
in
g
p
ar
ad
ig
m
w
h
ic
h
u
s
es
s
el
f
-
o
r
g
an
izi
n
g
m
ap
s
to
d
i
s
co
v
er
u
n
iq
u
e
p
atter
n
s
f
r
o
m
v
id
eo
f
r
a
m
e
s
a
n
d
also
d
is
tin
g
u
is
h
n
o
v
elt
y
i
n
t
h
e
f
r
a
m
es
b
elo
n
g
s
to
a
v
id
eo
s
e
q
u
en
ce
.
T
h
e
d
esi
g
n
an
d
i
m
p
le
m
e
n
tatio
n
o
f
t
h
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
h
av
e
b
ee
n
ca
r
r
ied
o
u
t
i
n
a
n
u
m
er
ical
c
o
m
p
u
tatio
n
al
e
n
v
ir
o
n
m
en
t
w
h
er
e
a
s
et
o
f
s
a
m
p
le
v
id
eo
s
ar
e
co
n
s
id
er
ed
to
test
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
p
r
o
p
o
s
ed
m
o
d
el.
T
h
e
ex
p
er
im
e
n
ta
l
o
u
tco
m
e
s
f
u
r
th
er
ex
h
ib
ited
t
h
at
it
p
r
o
v
id
es
p
r
o
m
is
in
g
r
esu
lts
a
n
d
en
s
u
r
e
it
'
s
ad
ap
tab
ilit
y
to
o
b
j
ec
t
d
etec
tio
n
s
y
s
te
m
s
i
n
cl
u
d
in
g
r
e
m
o
te
v
id
eo
s
u
r
v
eilla
n
ce
in
d
ef
en
s
e
f
o
r
n
at
io
n
al
a
n
d
in
ter
n
a
tio
n
al
b
o
r
d
er
tr
ac
k
in
g
.
Sar
av
an
a
n
an
d
Sri
n
iv
a
s
an
[
4
2
]
h
av
e
id
en
ti
f
ied
th
e
n
ee
d
f
o
r
an
ef
f
icie
n
t
Vid
eo
Fra
m
e
B
ased
r
etr
iev
al
s
y
s
te
m
.
Vad
u
v
a
[
4
3
]
p
r
esen
te
d
th
e
i
m
a
g
e
i
n
f
o
r
m
atio
n
m
i
n
i
n
g
b
ased
o
n
a
co
m
m
u
n
icatio
n
ch
a
n
n
el
co
n
ce
p
t.
T
h
e
au
th
o
r
s
co
n
s
id
er
ed
th
is
a
p
p
r
o
ac
h
f
o
r
a
r
ea
l
m
ea
n
i
n
g
b
ased
s
e
m
a
n
tic
a
n
n
o
tatio
n
o
f
v
er
y
h
i
g
h
-
r
eso
l
u
tio
n
r
e
m
o
te
s
en
s
in
g
i
m
ag
e
s
.
T
h
e
s
ce
n
e
co
n
ten
t
is
d
escr
ib
ed
u
s
in
g
a
m
u
lti
-
le
v
el
h
ier
ar
c
h
ical
in
f
o
r
m
at
io
n
r
ep
r
esen
tatio
n
.
Featu
r
e
h
ier
ar
ch
ies
ar
e
d
is
co
v
er
ed
co
n
s
id
er
in
g
t
h
at
h
i
g
h
er
le
v
el
s
ar
e
f
o
r
m
ed
b
y
co
m
b
in
in
g
f
ea
t
u
r
es
f
r
o
m
th
e
lo
w
er
lev
e
l
.
Su
c
h
a
lev
el
to
le
v
el
m
ap
p
in
g
d
ef
i
n
es
o
u
r
m
et
h
o
d
o
lo
g
y
as
a
d
ee
p
lear
n
i
n
g
p
r
o
ce
s
s
.
W
an
g
[
4
4
]
h
av
e
th
e
f
r
a
m
e
w
o
r
k
o
f
Ma
p
R
ed
u
ce
i
s
ex
p
lo
r
ed
f
o
r
lar
g
e
-
s
ca
le
m
u
lti
m
e
d
ia
d
ata
m
i
n
i
n
g
.
Yan
g
et
al.
[
4
5
]
d
esig
n
f
o
llo
ws
a
m
o
d
el
-
v
ie
w
-
co
n
tr
o
ller
(
MV
C
)
p
atter
n
f
o
r
ap
p
ly
i
n
g
s
e
m
a
n
tics
.
T
ab
le
1
s
h
o
w
th
e
s
u
m
m
er
y
o
f
E
x
is
ti
n
g
St
u
d
i
es to
w
ar
d
s
d
ata
m
in
i
n
g
k
e
y
ac
tiv
itie
s
.
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
-
8708
A
S
y
ste
ma
ti
c
Rev
iew o
f
Existin
g
Da
ta
M
in
in
g
Ap
p
ro
a
c
h
e
s E
n
v
isi
o
n
e
d
f
o
r Kn
o
wled
g
e
…
.
(
Ben
a
k
a
S
a
n
t
h
o
sh
a
S
)
913
T
a
b
le
1.
S
u
m
m
e
r
y
o
f
Ex
i
stin
g
S
t
u
d
ies
t
o
w
a
rd
s d
a
ta m
in
in
g
k
e
y
a
c
t
iv
it
ies
A
u
t
h
o
r
s
Y
e
a
r
C
o
n
c
e
p
t
s o
r
p
r
o
b
l
e
m F
o
c
u
se
d
A
p
p
l
i
c
a
t
i
o
n
D
o
mai
n
T
i
e
n
e
t
a
l
.
[
1
8
]
,
H
u
a
n
g
y
e
t
a
l
.
[
1
9
]
C
h
a
u
d
h
a
r
y
e
t
a
l
.
[
2
0
]
2
0
0
8
S
y
mb
o
l
i
c
st
r
e
a
ms mi
n
i
n
g
,
v
i
s
u
a
l
f
e
a
t
u
r
e
mi
n
i
n
g
,
s
u
p
p
o
r
t
v
e
c
t
o
r
mac
h
i
n
e
S
p
o
r
t
s,
mo
v
i
e
,
s
u
r
v
e
i
l
l
a
n
c
e
A
n
j
u
l
a
n
a
n
d
C
a
n
a
g
a
r
a
j
a
h
[
2
1
]
,
G
a
i
d
o
n
e
t
a
l
.
[
2
2
]
2
0
0
9
O
b
j
e
c
t
M
i
n
i
n
g
,
S
u
p
p
o
r
t
V
e
c
t
o
r
M
a
c
h
i
n
e
M
o
v
i
e
/
T
v
sh
o
w
s,
sp
o
r
t
s,
v
i
d
e
o
G
i
l
b
e
r
t
e
t
a
l
.
[
2
3
]
,
H
a
r
i
k
r
i
sh
n
a
e
t
a
l
.
[
2
4
]
,
Ji
a
n
g
e
t
a
l
.
[
2
5
]
2
0
1
1
S
e
q
u
e
n
t
i
a
l
p
a
t
t
e
r
n
m
i
n
i
n
g
M
o
v
i
e
/
T
v
sh
o
w
s,
sp
o
r
t
s,
v
i
d
e
o
,
su
r
v
e
i
l
l
a
n
c
e
C
u
i
e
t
a
l
.
[
2
6
]
2
0
1
1
H
i
e
r
a
r
c
h
i
c
a
l
v
i
s
u
a
l
e
v
e
n
t
p
a
t
t
e
r
n
mi
n
i
n
g
S
u
r
v
e
i
l
l
a
n
c
e
G
o
w
si
k
h
a
a
e
t
a
l
.
[
2
7
]
2
0
1
2
T
e
mp
o
r
a
l
a
sso
c
i
a
t
i
o
n
r
u
l
e
mi
n
i
n
g
S
e
c
u
r
i
t
y
a
n
d
s
u
r
v
e
i
l
l
a
n
c
e
V
i
j
a
y
k
u
m
a
r
a
n
d
N
e
d
u
n
c
h
e
z
h
i
a
n
[
2
8
]
2
0
1
2
V
i
d
e
o
d
a
t
a
mo
d
e
l
,
V
i
d
e
o
d
a
t
a
mi
n
i
n
g
M
o
v
i
e
/
T
v
sh
o
w
s,
sp
o
r
t
s,
v
i
d
e
o
,
su
r
v
e
i
l
l
a
n
c
e
X
i
a
o
-
C
h
a
o
e
t
a
l
.
[
2
9
]
2
0
1
3
D
a
t
a
mi
n
i
n
g
b
a
se
d
R
a
d
a
r
T
r
a
c
k
si
mu
l
a
t
i
o
n
R
a
d
a
r
d
a
t
a
mi
n
i
n
g
Z
h
u
[
3
0
]
2
0
1
4
sl
a
c
k
b
a
se
d
me
a
s
u
r
e
(
S
B
M
)
mo
d
e
l
a
n
d
d
a
t
a
m
i
n
i
n
g
Ed
u
c
a
t
i
o
n
Ef
f
i
c
i
e
n
c
y
Z
h
o
u
e
t
a
l
.
[
3
1
]
2
0
1
4
C
o
mm
e
r
c
i
a
l
d
e
c
i
si
o
n
ma
k
i
n
g
a
n
d
D
a
t
a
mi
n
i
n
g
D
a
t
a
A
n
a
l
y
si
s
P
a
r
k
e
t
a
l
.
[
3
2
]
2
0
1
6
A
n
o
p
t
i
m
i
z
a
t
i
o
n
p
r
o
b
l
e
m
f
o
r
mu
l
a
t
e
d
t
o
c
l
u
st
e
r
b
e
h
a
v
i
o
r
p
a
t
t
e
r
n
s (
b
P
s)
V
i
d
e
o
se
g
me
n
t
a
t
i
o
n
i
n
t
h
e
f
i
e
l
d
o
f
S
e
c
u
r
i
t
y
a
n
d
s
u
r
v
e
i
l
l
a
n
c
e
P
a
l
a
z
z
o
e
t
a
l
.
[
3
3
]
2
0
1
6
V
i
d
e
o
o
b
j
e
c
t
se
g
me
n
t
a
t
i
o
n
V
i
d
e
o
g
a
me
d
e
v
e
l
o
p
me
n
t
,
su
r
v
e
i
l
l
a
n
c
e
H
i
n
a
mi
a
n
d
S
a
t
o
h
.
[
3
4
]
2
0
1
7
D
i
sco
v
e
r
i
n
g
k
n
o
w
l
e
d
g
e
i
n
T
V
r
a
t
i
n
g
d
a
t
a
T
v
sh
o
w
s,
w
a
t
e
r
a
n
d
sp
o
r
t
r
e
p
o
r
t
s
L
e
y
v
a
e
t
a
l
.
[
3
5
]
2
0
1
7
V
i
d
e
o
a
n
o
m
a
l
y
d
e
t
e
c
t
i
o
n
S
e
c
u
r
i
t
y
a
n
d
v
i
d
e
o
su
r
v
e
i
l
l
a
n
c
e
sy
st
e
ms
2
.
1
.
Sta
t
is
t
ics o
f
E
x
is
t
i
ng
Resea
rc
h T
re
nd
s
o
n Video
M
ini
ng
T
h
is
s
ec
tio
n
e
x
clu
s
i
v
el
y
h
ig
h
lig
h
ts
th
e
cu
r
r
e
n
t
r
esear
c
h
s
ta
tis
tics
ab
o
u
t
th
e
v
id
eo
d
ata
m
in
i
n
g
an
d
an
al
y
tics
.
T
h
er
e
ar
e
n
o
t
m
u
c
h
-
p
u
b
lis
h
ed
r
esear
ch
ar
ticle
s
f
o
u
n
d
in
I
E
E
E
Xp
lo
r
e
d
ig
it
al
lib
r
ar
y
w
i
th
i
n
a
ti
m
eli
n
e
o
f
2
0
0
8
-
2
0
1
7
.
T
h
e
s
tatis
tics
o
f
th
e
p
u
b
lis
h
ed
co
n
te
n
ts
t
i
ll d
ate
is
r
ep
r
esen
ted
in
F
ig
u
r
e
5
.
T
h
e
s
tu
d
y
also
e
m
p
h
as
ized
o
n
ex
p
lo
r
in
g
ex
is
ti
n
g
s
t
u
d
ies
f
o
cu
s
ed
o
n
s
tr
en
g
t
h
e
n
in
g
d
if
f
e
r
en
t
v
id
eo
d
ata
m
in
in
g
ap
p
licatio
n
s
.
T
h
e
d
ata
co
llected
ex
clu
s
iv
e
l
y
r
e
f
er
r
in
g
I
E
E
E
Xp
lo
r
e
d
ig
ital
li
b
r
ar
y
w
h
ic
h
s
h
o
w
s
v
er
y
f
e
w
er
w
o
r
k
s
h
a
v
e
b
ee
n
ca
r
r
ied
o
u
t
o
n
v
id
eo
d
ata
an
al
y
t
ics
co
n
s
id
er
in
g
d
if
f
er
en
t
ap
p
licatio
n
asp
ec
ts
s
u
c
h
as
s
ec
u
r
it
y
a
n
d
v
id
eo
s
u
r
v
ei
ll
an
ce
,
v
id
eo
g
a
m
e
ap
p
licatio
n
s
,
etc.
T
h
e
f
o
llo
w
in
g
Fi
g
u
r
e
6
s
h
o
w
s
a
n
o
v
er
v
ie
w
o
f
t
h
e
c
u
r
r
en
t
s
tati
s
tic
s
s
u
b
j
ec
ted
to
th
e
n
u
m
b
er
o
f
r
esear
ch
ar
ticle
s
p
u
b
lis
h
ed
o
n
v
id
e
o
d
ata
an
al
y
tic
s
f
o
r
d
if
f
er
e
n
t v
id
eo
ap
p
licatio
n
s
.
Fig
u
r
e
5
.
R
esear
ch
S
tatis
t
ics o
n
Vid
eo
Data
Min
i
n
g
(
I
E
E
E
Xp
lo
r
e)
Fig
u
r
e
6
Statis
tic
s
o
f
r
esear
c
h
o
n
Vid
eo
d
ata
m
i
n
i
n
g
ap
p
licatio
n
s
(
I
E
E
E
Xp
lo
r
e)
T
h
e
ab
o
v
e
f
ig
u
r
e
s
h
o
w
s
t
h
at
v
id
eo
d
ata
m
i
n
in
g
h
as
b
ee
n
m
o
s
t
ex
te
n
s
i
v
el
y
an
d
w
id
el
y
ad
o
p
ted
in
to
s
ec
u
r
it
y
an
d
s
u
r
v
eilla
n
ce
ap
p
licatio
n
s
s
in
ce
2
0
0
8
.
T
h
e
s
tatis
tics
al
s
o
co
n
v
e
y
s
th
at
v
er
y
l
ess
co
n
tr
ib
u
tio
n
to
m
o
v
ie/T
V
s
h
o
w
a
n
al
y
tics
.
T
h
e
r
esear
ch
g
ap
ex
tr
ac
te
d
b
r
ief
l
y
ill
u
s
tr
ated
b
elo
w
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
2
,
A
p
r
il
201
8
:
9
0
8
–
9
1
6
914
3.
RE
S
E
ARCH
G
AP
Af
ter
r
ev
ie
w
i
n
g
th
e
ab
o
v
e
s
ta
ted
ex
is
t
in
g
liter
at
u
r
e,
co
n
ce
r
n
ed
d
if
f
er
en
t
asp
ec
ts
o
f
v
id
eo
d
ata
m
i
n
i
n
g
,
th
e
s
tu
d
y
o
u
tli
n
e
s
m
o
s
t
s
ig
n
i
f
ica
n
t
is
s
u
es
n
ee
d
ed
to
b
e
ad
d
r
ess
ed
to
r
ein
f
o
r
ce
th
e
ex
i
s
tin
g
v
id
eo
m
i
n
i
n
g
tech
n
iq
u
es.
a.
Min
i
n
g
o
f
Se
m
an
tic
C
o
n
ce
p
ts
:
Ver
y
f
e
w
s
t
u
d
ies
ar
e
f
o
u
n
d
to
e
m
p
h
a
s
ize
o
n
m
i
n
in
g
o
f
s
e
m
an
tic
co
n
ce
p
ts
f
r
o
m
d
i
f
f
er
e
n
t
i
n
tell
ig
en
t
v
id
eo
ap
p
licatio
n
s
.
Mo
s
t
o
f
th
e
w
o
r
k
s
f
o
u
n
d
to
ca
r
r
y
o
u
t
th
eo
r
etica
l
d
is
c
u
s
s
io
n
o
n
p
r
ed
ictiv
e
s
e
m
an
t
ic
f
ea
tu
r
e
p
r
o
b
lem
s
w
h
er
ea
s
n
o
e
x
te
n
s
i
v
e
s
i
m
u
latio
n
to
d
eter
m
i
n
e
t
h
e
ef
f
ec
ti
v
en
e
s
s
o
f
v
id
eo
an
al
y
s
i
s
b
ased
o
n
p
r
im
iti
v
e
f
ea
tu
r
e
s
h
as b
ee
n
w
it
n
e
s
s
ed
.
b.
L
es
s
Fo
cu
s
to
w
ar
d
s
No
n
-
De
t
er
m
in
is
tic
A
p
p
r
o
ac
h
es:
Mo
s
t
o
f
th
e
ex
i
s
ti
n
g
s
tu
d
ie
s
f
o
cu
s
es
o
n
d
eter
m
in
i
s
tic
ap
p
r
o
ac
h
es
w
h
e
r
e
v
er
y
f
e
w
f
o
u
n
d
to
ap
p
ly
n
o
n
-
d
eter
m
i
n
i
s
tic
ap
p
r
o
ac
h
es
d
u
r
in
g
k
n
o
w
led
g
e
ex
tr
ac
tio
n
f
r
o
m
a
v
id
eo
o
b
j
ec
t.
c.
Fe
w
B
en
c
h
m
ar
k
i
n
g
:
Ver
y
l
ess
ef
f
ec
ti
v
e
s
t
u
d
ies
ar
e
f
o
u
n
d
till
d
ata
w
h
er
e
co
m
p
u
t
atio
n
al
co
m
p
le
x
it
y
an
d
b
en
c
h
m
ar
k
i
n
g
o
f
th
e
p
r
o
p
o
s
ed
s
o
lu
tio
n
to
w
a
r
d
s
v
id
eo
m
in
i
n
g
h
i
g
h
l
y
ig
n
o
r
ed
.
d.
No
-
Op
ti
m
izatio
n
:
Ver
y
f
e
w
er
s
tu
d
ie
s
co
n
s
id
er
ed
alg
o
r
it
h
m
o
p
tim
izatio
n
to
m
a
x
i
m
ize
t
h
e
s
y
s
te
m
th
r
o
u
g
h
p
u
t f
r
o
m
a
n
o
p
er
atio
n
a
l v
ie
w
p
o
in
t.
4.
CO
NCLU
SI
ON
T
h
e
p
r
o
p
o
s
ed
s
tu
d
y
i
n
te
n
d
s
to
p
er
f
o
r
m
an
d
i
n
-
d
ep
th
an
a
l
y
s
i
s
o
f
th
e
co
n
v
e
n
tio
n
al
v
id
eo
m
in
in
g
tech
n
iq
u
es a
n
d
th
e
ir
p
er
f
o
r
m
a
n
ce
ef
f
icie
n
c
y
.
T
h
e
s
tu
d
y
a
ls
o
h
ig
h
li
g
h
ts
a
co
m
p
r
e
h
en
s
i
v
e
o
v
er
v
ie
w
o
f
d
i
f
f
er
e
n
t
v
id
eo
m
in
i
n
g
tec
h
n
iq
u
e
s
a
n
d
th
eir
ad
ap
tab
ilit
y
in
to
d
i
f
f
er
en
t
s
y
s
te
m
s
f
o
r
e
f
f
icie
n
t
k
n
o
w
led
g
e
d
is
co
v
er
y
p
r
o
ce
s
s
.
T
h
e
in
v
est
ig
at
io
n
al
s
t
u
d
y
d
ep
icted
th
e
f
ac
t
t
h
at
t
h
e
ex
is
t
in
g
s
o
l
u
tio
n
ap
p
r
o
ac
h
es
l
ac
k
s
co
m
p
u
tatio
n
al
ef
f
icien
c
y
a
n
d
d
o
esn
't
ac
h
ie
v
e
an
o
p
ti
m
al
tr
ad
e
-
o
f
f
b
et
w
e
en
m
a
x
i
m
u
m
ac
c
u
r
ac
y
i
n
an
al
y
ze
d
co
n
ten
ts
an
d
o
p
er
atio
n
al
co
n
s
tr
ain
t
s
.
I
t
als
o
o
u
tlin
es
t
h
e
e
x
is
ti
n
g
r
esear
ch
is
s
u
e
s
w
h
ic
h
ar
e
n
ee
d
ed
t
o
b
e
m
i
n
i
m
ized
to
m
ak
e
t
h
i
s
r
esear
ch
tr
ac
k
m
o
r
e
ef
f
ec
tiv
e
a
n
d
o
p
er
ativ
e.
RE
F
E
R
E
NC
E
S
[1
]
F
.
A
n
wa
r,
I.
P
e
tro
u
n
ias
,
T
.
M
o
r
ris,
V
.
Ko
d
o
g
ian
n
is,
“
M
in
in
g
a
n
o
m
a
lo
u
s
e
v
e
n
ts
a
g
a
in
st
f
re
q
u
e
n
t
se
q
u
e
n
c
e
s
i
n
su
rv
e
il
lan
c
e
v
id
e
o
s f
ro
m
c
o
m
m
e
r
c
ial
e
n
v
iro
n
m
e
n
ts,
"
Exp
S
y
st A
p
p
l
3
9
:
4
5
1
1
–
4
5
3
1
,
2
0
1
2
[2
]
A
.
A
n
ju
lan
,
N.
Ca
n
a
g
a
ra
jah
,
“
A
u
n
ifi
e
d
f
ra
m
e
w
o
r
k
f
o
r
o
b
jec
t
re
tri
e
v
a
l
a
n
d
m
in
in
g
,
"
IEE
E
T
ra
n
s
Circ
S
y
st
Vi
d
e
o
T
e
c
h
n
o
l
1
9
(1
):
6
3
–
7
6
,
2
0
0
9
[3
]
A
.
A
h
m
e
d
,
“
V
id
e
o
re
p
re
se
n
tati
o
n
a
n
d
p
r
o
c
e
ss
in
g
f
o
r
m
u
lt
ime
d
iad
a
tam
in
in
g
,
"
S
e
ma
n
ti
c
min
in
g
tec
h
n
o
lo
g
ies
fo
rm
u
lt
ime
d
i
a
d
a
t
a
b
a
se
s
.
IG
I
P
re
ss
,
p
p
1
–
3
1
,
2
0
0
9
[4
]
A
.
A
n
ju
lan
,
N.
Ca
n
a
g
a
ra
jah
,
“
A
n
o
v
e
l
v
id
e
o
min
i
n
g
sy
ste
m,"
In
P
r
o
c
e
e
d
in
g
s
o
f
1
4
th
IEE
E
i
n
tern
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
im
a
g
e
p
ro
c
e
ss
in
g
,
S
a
n
A
n
to
n
i
o
,
T
e
x
a
s,
p
p
1
8
5
–
1
8
9
,
2
0
0
7
[5
]
H.
A
ra
d
h
y
e
,
G
.
T
o
d
e
rici,
J.
Ya
g
n
ik
,
“
Vi
d
e
o
2
T
e
x
t:
le
a
rn
i
n
g
t
o
a
n
n
o
ta
te
v
i
d
e
o
c
o
n
ten
t
”
,
I
n
P
r
o
c
e
e
d
in
g
s
o
f
IEE
E
in
tern
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
d
a
ta
m
in
in
g
w
o
rk
sh
o
p
s,
p
p
1
4
4
–
1
5
2
,
2
0
0
9
[6
]
C.
A
.
Bh
a
tt
,
M
.
S
.
Ka
n
k
a
n
h
a
ll
i,
“
M
u
lt
im
e
d
iad
a
tam
in
in
g
:
sta
te
o
f
t
h
e
a
rt
a
n
d
c
h
a
ll
e
n
g
e
s,"
M
u
lt
ime
d
ia
T
o
o
ls
Ap
p
l
5
1
:
3
5
–
7
6
,
2
0
1
1
[7
]
D.
Bre
z
e
a
le,
D.J.
Co
o
k
,
“
A
u
to
m
a
ti
c
v
id
e
o
c
las
sifi
c
a
ti
o
n
:
a
su
rv
e
y
o
f
th
e
li
ter
a
tu
re
,
"
IEE
E
T
ra
n
s
S
y
st
M
a
n
Cy
b
e
rn
Pa
rt C:
A
p
p
l
Rev
3
8
(3
):
4
1
6
–
4
3
0
,
2
0
0
8
[8
]
M
-
C.
T
ien
,
Y
-
T
.
W
a
n
g
,
C
-
W
.
Ch
o
u
,
K
-
Y.
Hs
ieh
,
W
-
T
.
Ch
u
,
J
.
L
.
W
u
,
“
Eve
n
t
d
e
tec
ti
o
n
i
n
ten
n
is
m
a
tch
e
s
b
a
se
d
o
n
v
id
e
o
d
a
ta
min
i
n
g
,
"
P
ro
c
ICM
E,
p
p
.
1
4
7
7
–
1
4
8
0
,
2
0
0
8
[9
]
P
.
Cu
i,
Z
-
Q.
L
iu
,
L
-
F
.
S
u
n
,
S
-
Q
.
Ya
n
g
,
“
Hie
r
a
rc
h
ica
l
v
isu
a
l
e
v
e
n
t
p
a
tt
e
rn
m
in
in
g
a
n
d
it
s
a
p
p
li
c
a
ti
o
n
s,"
J
D
a
t
a
M
in
in
g
Kn
o
wl
Disc
2
2
(3
):
4
6
7
–
4
9
2
,
2
0
1
1
[1
0
]
B
-
W
.
Ch
e
n
,
J
-
C.
W
a
n
g
,
F
.
W
a
n
g
,
“
A
n
o
v
e
l
v
id
e
o
su
m
m
a
riza
ti
o
n
b
a
se
d
o
n
m
in
in
g
th
e
sto
r
y
-
stru
c
tu
re
a
n
d
se
m
a
n
ti
c
re
latio
n
s am
o
n
g
c
o
n
c
e
p
t
e
n
t
it
ies
,
"
IEE
E
T
r
a
n
s M
u
l
ti
me
d
ia
1
1
(2
):
2
9
5
–
3
1
3
,
2
0
0
9
[1
1
]
X
.
Zh
u
,
X.
W
u
,
A
.
El
m
a
g
a
r
m
id
,
Z.
F
e
n
g
,
L
.
W
u
,
“
V
id
e
o
d
a
ta
m
i
n
in
g
:
se
m
a
n
ti
c
in
d
e
x
in
g
a
n
d
e
v
e
n
t
d
e
tec
ti
o
n
f
ro
m
th
e
a
ss
o
c
iatio
n
p
e
rsp
e
c
ti
v
e
,
"
IEE
E
T
ra
n
s Kn
o
wl
Da
t
a
E
n
g
1
7
(5
):
1
–
1
4
,
2
0
0
5
[1
2
]
J.
Ch
e
n
,
Z.
L
iT
,
D.
P
.
X
u
B
,
“
Co
mm
e
rc
ia
ld
e
tec
ti
o
n
b
y
min
i
n
g
ma
x
ima
l
re
p
e
a
te
d
se
q
u
e
n
c
e
in
a
u
d
io
stre
a
m,
"
P
r
o
c
e
e
d
in
g
s o
f
IEE
E,
2
0
1
1
[1
3
]
A
.
A
n
ju
lan
,
N.
Ca
n
a
g
a
ra
jah
,
“
A
n
o
v
e
l
v
id
e
o
min
i
n
g
sy
ste
m,"
In
P
r
o
c
e
e
d
in
g
s
o
f
1
4
th
IEE
E
i
n
tern
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
im
a
g
e
p
ro
c
e
ss
in
g
,
S
a
n
A
n
to
n
i
o
,
T
e
x
a
s,
p
p
1
8
5
–
1
8
9
,
2
0
0
7
[1
4
]
M
.
Ch
e
n
,
S
-
C.
Ch
e
n
,
M
-
L
.
S
h
y
u
,
“
Hie
ra
rc
h
ic
a
lt
e
mp
o
r
a
la
ss
o
c
ia
ti
o
n
min
i
n
g
fo
r
v
id
e
o
e
v
e
n
t
d
e
te
c
ti
o
n
i
n
v
id
e
o
d
a
t
a
b
a
se
s,"
In
T
h
e
se
c
o
n
d
IE
EE
in
tern
a
ti
o
n
a
l
w
o
rk
sh
o
p
o
n
m
u
lt
i
m
e
d
ia
d
a
tab
a
se
s
a
n
d
d
a
ta
m
a
n
a
g
e
m
e
n
t
(M
DD
M
'
0
7
),
i
n
c
o
n
j
u
n
c
ti
o
n
w
it
h
IEE
E
in
tern
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
d
a
ta en
g
in
e
e
rin
g
,
Ista
n
b
u
l,
T
u
rk
e
y
,
2
0
0
7
[1
5
]
S
-
C.
Ch
e
n
,
M
.
Ch
e
n
,
C.
Z
h
a
n
g
,
M
-
L
.
S
h
y
u
,
“
Exc
it
in
g
e
v
e
n
t
d
e
tec
ti
o
n
u
sin
g
mu
lt
i
-
lev
e
l
mu
lt
im
o
d
a
l
d
e
sc
rip
to
rs
a
n
d
d
a
t
a
c
la
ss
i
fi
c
a
ti
o
n
,
"
I
n
P
r
o
c
e
e
d
in
g
s o
f
e
ig
h
th
IEE
E
i
n
tern
a
ti
o
n
a
l
sy
m
p
o
siu
m
o
n
m
u
lt
ime
d
i
a
,
p
p
1
9
3
–
2
0
0
,
2
0
0
6
[1
6
]
S
-
C.
Ch
e
n
,
M
-
L
.
S
h
y
u
,
Z.
C.
L
u
o
L
,
M
.
Ch
e
n
,
“
De
tec
ti
o
n
o
f
so
c
c
e
r
g
o
a
l
s
h
o
ts
u
sin
g
j
o
i
n
t
mu
l
ti
me
d
i
a
fea
t
u
re
s
a
n
d
c
la
ss
ifi
c
a
ti
o
n
ru
les
,
"
In
P
r
o
c
e
e
d
i
n
g
s
o
f
in
tern
a
ti
o
n
a
l
w
o
rk
sh
o
p
o
n
m
u
lt
ime
d
ia
d
a
ta
m
in
in
g
(M
DM/
KD
D’2
0
0
3
)
,
USA
,
p
p
3
6
–
4
4
,
2
0
0
3
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
-
8708
A
S
y
ste
ma
ti
c
Rev
iew o
f
Existin
g
Da
ta
M
in
in
g
Ap
p
ro
a
c
h
e
s E
n
v
isi
o
n
e
d
f
o
r Kn
o
wled
g
e
…
.
(
Ben
a
k
a
S
a
n
t
h
o
sh
a
S
)
915
[1
7
]
S
-
C.
Ch
e
n
M
-
L
.
S
h
y
u
,
C.
Zh
a
n
g
,
J.
S
tri
c
k
ro
tt
,
“
M
u
lt
ime
d
ia
d
a
t
a
min
in
g
fo
r
tra
f
fi
c
v
id
e
o
se
q
u
e
n
c
e
s,"
In
:
P
r
o
c
e
e
d
in
g
s
se
c
o
n
d
in
tern
a
ti
o
n
a
l
w
o
rk
sh
o
p
o
n
m
u
lt
ime
d
ia
d
a
ta
m
in
in
g
M
DM/
KD
D’2
0
0
1
in
c
o
n
j
u
n
c
ti
o
n
w
it
h
A
CM
S
IGK
DD
se
v
e
n
th
in
tern
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
k
n
o
w
le
d
g
e
d
isc
o
v
e
r
y
a
n
d
d
a
ta m
in
in
g
,
p
p
7
8
–
8
6
,
2
0
0
1
[1
8
]
M
-
C.
T
ien
,
Y
-
T
.
W
a
n
g
,
C
-
W
.
Ch
o
u
,
K
-
Y.
Hs
ieh
,
W
-
T
.
Ch
u
,
J.L
.
W
u
,
“
Eve
n
t
d
e
tec
ti
o
n
i
n
te
n
n
is
m
a
tch
e
s
b
a
se
d
o
n
v
id
e
o
d
a
ta
min
i
n
g
,
"
P
ro
c
ICM
E
2
0
0
8
:1
4
7
7
–
1
4
8
0
,
2
0
0
8
[1
9
]
H
-
Y.
Hu
a
n
g
y
,
W
-
S
.
S
h
ih
,
W
-
H.
Hs
u
,
“
A
f
il
m
c
las
si
f
ier
b
a
s
e
d
o
n
lo
w
-
lev
e
l
v
isu
a
l
f
e
a
tu
re
s
,”
J
M
u
lt
ime
d
ia
3
(3
)
:
26
–
3
3
,
2
0
0
8
[2
0
]
A
.
Ch
o
u
d
h
a
ry
,
S
.
Ch
a
u
d
h
u
ry
,
S
.
Ba
sn
e
rjee
,
“
A
fra
me
wo
rk
fo
r
a
n
a
lys
is
o
f
su
rv
e
il
la
n
c
e
v
id
e
o
s,
”
I
n
P
r
o
c
e
e
d
in
g
s
o
f
six
th
In
d
ian
c
o
n
f
e
re
n
c
e
o
n
c
o
m
p
u
ter v
isio
n
,
g
ra
p
h
ics
&
im
a
g
e
p
ro
c
e
ss
in
g
,
p
p
3
4
4
–
3
5
0
,
2
0
0
8
[2
1
]
A
.
A
n
ju
lan
,
N.
Ca
n
a
g
a
ra
jah
,
“
A
u
n
if
ied
f
ra
m
e
w
o
rk
f
o
r
o
b
jec
t
re
tr
iev
a
l
a
n
d
m
in
in
g
,
”
IEE
E
T
r
a
n
s
Circ
S
y
st
Vi
d
e
o
T
e
c
h
n
o
l
1
9
(1
):
6
3
–
7
6
,
2
0
0
9
[2
2
]
A
.
G
a
id
o
n
,
M
.
M
a
rsz
a
lek
,
C.
S
c
h
m
id
,
“
M
in
in
g
v
isu
a
l
a
c
ti
o
n
s
fr
o
m
mo
v
ies
,
”
In
P
ro
c
e
e
d
i
n
g
s
o
f
th
e
Brit
ish
m
a
c
h
in
e
c
o
n
f
e
re
n
c
e
.
BM
V
A
P
re
ss
,
p
p
1
2
5
.
1
–
1
2
5
.
1
1
,
2
0
0
9
[2
3
]
A
.
G
il
b
e
rt,
J.
Ill
i
n
g
w
o
rth
,
R.
Bo
w
d
e
n
,
“
Ac
ti
o
n
re
c
o
g
n
it
i
o
n
u
sin
g
m
in
e
d
h
iera
rc
h
ica
l
c
o
m
p
o
u
n
d
f
e
a
tu
re
s,
”
IEE
E
T
ra
n
s P
a
tt
e
rn
An
a
l
M
a
c
h
I
n
tell
3
3
(5
):
8
8
3
–
8
9
7
,
2
0
1
1
[2
4
]
N.
Ha
rik
rish
n
a
,
S
.
S
a
t
h
e
e
sh
,
D.
S
riram
,
K.S
.
Eas
w
a
ra
k
u
m
a
r,
“
T
e
m
p
o
ra
l
c
las
sif
ica
ti
o
n
o
f
e
v
e
n
ts
in
c
rick
e
t
v
id
e
o
s.
In
:
P
r
o
c
e
e
d
in
g
s
o
f
se
v
e
n
tee
n
th
n
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
c
o
m
m
u
n
ica
t
io
n
s
NCC
2
0
1
1
”
,
In
d
ian
I
n
stit
u
te
o
f
S
c
ien
c
e
,
Ba
n
g
a
lo
re
,
2
0
1
1
[2
5
]
F
.
Jia
n
g
,
J.
Yu
a
n
,
S
.
A
.
T
sa
f
tari
s,
A
.
K.
Ka
tsa
g
g
e
lo
s
,
“
A
n
o
m
a
lo
u
s
v
id
e
o
e
v
e
n
t
d
e
tec
ti
o
n
u
sin
g
sp
a
ti
o
tem
p
o
ra
l
c
o
n
tex
t”,
In
t
J
C
o
mp
u
t
Vi
s Im
a
g
e
Un
d
e
rs
t
1
1
5
:
3
2
3
–
3
3
3
,
2
0
1
1
[2
6
]
P
.
Cu
i,
Z
-
Q.
L
iu
,
L
-
F
.
S
u
n
,
S
-
Q
.
Ya
n
g
,
“
Hie
r
a
rc
h
ica
l
v
isu
a
l
e
v
e
n
t
p
a
tt
e
rn
m
in
in
g
a
n
d
it
s
a
p
p
li
c
a
ti
o
n
s”
,
J
Da
t
a
M
in
in
g
Kn
o
wl
Disc
2
2
(3
):
4
6
7
–
49
2
,
2
0
1
1
[2
7
]
D.
G
o
w
sik
h
a
a
,
A
.
S
.
M
a
n
ju
n
a
th
,
“
S
u
sp
icio
u
s
h
u
m
a
n
a
c
ti
v
it
y
d
e
tec
ti
o
n
f
ro
m
su
rv
e
il
lan
c
e
v
id
e
o
s”
,
In
t
J
In
t
Distrib
Co
mp
u
t
S
y
st
2
(2
):
1
4
1
–
1
4
9
,
2
0
1
2
[2
8
]
W
U
X
iao
-
c
h
a
o
,
W
a
n
g
L
ian
-
d
o
n
g
,
YA
N
L
iao
-
li
a
o
,
X
UE
F
a
n
g
-
x
ia,
"
S
im
u
latio
n
o
f
Ra
d
a
r
T
ra
c
k
Ba
se
d
o
n
Da
t
a
M
in
i
n
g
T
e
c
h
n
iq
u
e
s"
,
T
EL
KOM
NIKA
(
T
e
lec
o
mm
u
n
ica
ti
o
n
Co
mp
u
ti
n
g
El
e
c
tro
n
ics
a
n
d
Co
n
t
ro
l)
,
V
o
l.
1
1
,
p
p
.
3
7
8
0
~
3
7
8
8
,
No
.
7
,
J
u
ly
2
0
1
3
.
[2
9
]
[2
9
]
X
iao
d
o
n
g
Zh
u
"
On
Da
ta M
i
n
in
g
T
e
c
h
n
o
l
o
g
y
to
th
e
Qu
a
n
ti
tati
v
e
Eff
i
c
ien
c
y
A
ss
e
ss
m
e
n
t
u
sin
g
S
BM
M
o
d
e
l:
A
n
Em
p
iri
c
a
l
S
tu
d
y
o
n
Ed
u
c
a
ti
o
n
Eff
icie
n
c
y
in
Jia
n
g
x
i
P
ro
v
in
c
e
"
,
T
EL
KOM
NIKA
(
T
e
le
c
o
mm
u
n
ica
t
io
n
Co
m
p
u
ti
n
g
El
e
c
tro
n
ics
a
n
d
C
o
n
tro
l)
Vo
l.
1
2
,
No
.
3
,
M
a
rc
h
2
0
1
4
,
p
p
.
1
9
3
3
~
1
9
3
8
,
M
a
rc
h
2
0
1
4
.
[3
0
]
Qish
e
n
Zh
o
u
,
Zu
o
w
e
i
Yin
,
Qiu
Y
in
g
,
W
a
n
g
S
h
a
n
h
u
i"
In
telli
g
e
n
t
Da
ta M
in
in
g
a
n
d
De
c
isio
n
S
y
ste
m
f
o
r
C
o
m
m
e
r
c
ial
De
c
isio
n
M
a
k
in
g
"
T
EL
KOM
NI
KA
(
T
e
lec
o
mm
u
n
ica
ti
o
n
C
o
mp
u
ti
n
g
El
e
c
tro
n
ics
a
n
d
C
o
n
tro
l)
V
o
l
.
1
2
,
No
.
1
,
p
p
.
7
9
2
~
8
0
1
,
Ja
n
u
a
ry
2
0
1
4
.
[3
1
]
V
.
V
ij
a
y
a
k
u
m
a
r
R.
Ne
d
u
n
c
h
e
z
h
i
a
n
,
“
A
stu
d
y
o
n
v
id
e
o
d
a
ta
m
in
in
g
”
,
In
t
J
M
u
lt
ime
d
In
fo
Retr
,
v
o
l.
1
,
p
p
.
1
5
3
-
1
7
2
,
2
0
1
2
[3
2
]
S.
P
a
rk
,
H.
P
a
rk
a
n
d
C.
D.
Yo
o
,
"
Co
m
p
lex
V
id
e
o
S
c
e
n
e
A
n
a
l
y
sis
Us
in
g
Ke
rn
e
li
z
e
d
-
Co
ll
a
b
o
ra
ti
v
e
Be
h
a
v
io
r
P
a
tt
e
r
n
L
e
a
rn
in
g
Ba
s
e
d
o
n
Hie
ra
rc
h
ica
l
Re
p
re
se
n
tativ
e
Ob
jec
t
Be
h
a
v
io
rs,"
in
IEE
E
T
r
a
n
sa
c
ti
o
n
s
o
n
Circ
u
it
s
a
n
d
S
y
ste
ms
fo
r V
i
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
2
7
,
n
o
.
6
,
p
p
.
1
2
7
5
-
1
2
8
9
,
J
u
n
e
2
0
1
7
.
[3
3
]
C.
S
p
a
m
p
in
a
to
,
S
.
P
a
laz
z
o
a
n
d
D.
G
io
rd
a
n
o
,
"
Ga
m
i
fy
in
g
V
id
e
o
Ob
jec
t
S
e
g
m
e
n
tatio
n
,
"
in
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Pa
tt
e
rn
A
n
a
lys
is
a
n
d
M
a
c
h
i
n
e
In
t
e
ll
ig
e
n
c
e
,
v
o
l.
3
9
,
n
o
.
1
0
,
p
p
.
1
9
4
2
-
1
9
5
8
,
Oc
t.
1
2
0
1
7
.
[3
4
]
R.
Hin
a
m
i
a
n
d
S
.
S
a
to
h
,
"
A
u
d
ien
c
e
Be
h
a
v
io
r
M
in
in
g
:
I
n
teg
ra
ti
n
g
T
V
Ra
ti
n
g
s
w
it
h
M
u
lt
im
e
d
ia
Co
n
ten
t,
"
i
n
IEE
E
M
u
lt
iM
e
d
ia
,
v
o
l.
2
4
,
n
o
.
2
,
p
p
.
4
4
-
5
4
,
A
p
r.
-
Ju
n
e
2
0
1
7
.
[3
5
]
R.
L
e
y
v
a
,
V
.
S
a
n
c
h
e
z
a
n
d
C.
T
.
L
i,
"
V
id
e
o
A
n
o
m
a
l
y
De
te
c
ti
o
n
W
it
h
Co
m
p
a
c
t
F
e
a
tu
re
S
e
ts
f
o
r
On
li
n
e
P
e
rf
o
rm
a
n
c
e
,
"
in
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Im
a
g
e
Pr
o
c
e
ss
in
g
,
v
o
l
.
2
6
,
n
o
.
7
,
p
p
.
3
4
6
3
-
3
4
7
8
,
Ju
ly
2
0
1
7
.
[3
6
]
W
.
P
e
ter,
J.
S
o
a
r,
M
.
A
ll
y
,
“
M
u
lt
ime
d
ia
D
a
ta
M
i
n
in
g
u
si
n
g
De
e
p
L
e
a
rn
in
g
”
,
IE
EE
In
tern
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
Dig
it
a
l
In
f
o
rm
a
ti
o
n
P
r
o
c
e
ss
in
g
a
n
d
Co
m
m
u
n
ica
ti
o
n
s
,
2
0
1
5
[3
7
]
Y.H.S
.
Ku
m
a
r,
M
.
N,
Ch
e
th
a
n
H
K,
“
An
ima
l
Cla
ss
if
ica
ti
o
n
S
y
ste
m:
A
Bl
o
c
k
Ba
se
d
Ap
p
ro
a
c
h
”
,
El
se
v
ier
-
In
tern
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
A
d
v
a
n
c
e
d
Co
m
p
u
ti
n
g
T
e
c
h
n
o
lo
g
ies
a
n
d
A
p
p
li
c
a
ti
o
n
s,
2
0
1
5
[3
8
]
M
.
S
h
a
h
b
a
z
,
A
.
G
u
e
rg
a
c
h
i,
A
.
No
re
e
n
,
a
n
d
M
.
S
h
a
h
e
e
n
,
“
A
D
a
ta
M
in
in
g
A
p
p
ro
a
c
h
to
Re
c
o
g
n
ize
Ob
jec
ts
in
S
a
telli
te Im
a
g
e
s to
P
re
d
ict
Na
tu
ra
l
Re
so
u
rc
e
s”
,
S
p
r
n
g
e
r
-
IA
EN
G
T
r
a
n
sa
c
ti
o
n
s o
n
E
n
g
in
e
e
ri
n
g
T
e
c
h
n
o
lo
g
ies
,
2
0
1
3
[3
9
]
H.
Be
n
o
it
,
“
M
u
l
ti
m
e
d
ia Co
n
ten
t
Un
d
e
rsta
n
d
i
n
g
:
Brin
g
i
n
g
Co
n
tex
t
to
Co
n
ten
t”,
HA
L
-
Eu
re
c
o
m
,
2
0
1
2
[4
0
]
X
.
Ca
o
a
n
d
S
.
W
a
n
g
,
“
Re
se
a
rc
h
a
b
o
u
t
Im
a
g
e
M
in
in
g
T
e
c
h
n
iq
u
e
,
S
p
rin
g
e
r”
,
2
0
1
2
[4
1
]
C.
L
.
De
v
a
s
e
n
a
a
,
M
.
He
m
a
l
a
th
a
,
“
Vi
d
e
o
M
in
i
n
g
u
sin
g
L
IM
Ba
se
d
Clu
ste
rin
g
a
n
d
S
e
lf
Or
g
a
n
izin
g
M
a
p
s”
,
El
se
v
ier
-
In
tern
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Co
m
m
u
n
ica
ti
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
S
y
ste
m
D
e
sig
n
,
2
0
1
2
[4
2
]
D.
S
a
ra
v
a
n
a
n
,
S
.
S
rin
iv
a
sa
n
,
“
Da
ta
M
in
in
g
F
ra
m
e
w
o
rk
f
o
r
V
id
e
o
Da
ta”
,
IEE
E
Re
c
e
n
t
A
d
v
a
n
c
e
s
in
S
p
a
c
e
T
e
c
h
n
o
lo
g
y
S
e
rv
i
c
e
s an
d
Cli
m
a
te
Ch
a
n
g
e
,
2
0
1
0
[4
3
]
C.
V
a
d
u
v
a
,
I.
G
a
v
a
t,
M
.
Da
tcu
,
“
De
e
p
L
e
a
rn
in
g
in
V
e
ry
Hig
h
Re
so
lu
ti
o
n
Re
m
o
te
S
e
n
sin
g
Im
a
g
e
In
f
o
r
m
a
ti
o
n
M
in
i
n
g
”
,
Co
m
m
u
n
ica
ti
o
n
C
o
n
c
e
p
t,
EUS
I
P
CO,
2
0
1
2
[4
4
]
H.
W
a
n
g
,
Y.
S
h
e
n
,
L
.
W
a
n
g
,
“
L
a
rg
e
-
S
c
a
le
M
u
lt
ime
d
i
a
Da
t
a
M
in
in
g
Us
in
g
M
a
p
Re
d
u
c
e
Fra
me
wo
rk
”
,
IEE
E
4
t
h
In
tern
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Cl
o
u
d
Co
m
p
u
ti
n
g
T
e
c
h
n
o
lo
g
y
a
n
d
S
c
i
e
n
c
e
,
2
0
1
2
[4
5
]
Y.
Ya
n
g
,
H
-
Y.
Ha
,
F
.
C.
F
leite
s,
“
A
M
u
lt
im
e
d
ia
S
e
m
a
n
ti
c
Re
tr
iev
a
l
M
o
b
il
e
S
y
ste
m
Ba
s
e
d
o
n
HCFG
s
”
,
IEE
E
Co
m
p
u
ter S
o
c
iety
,
2
0
1
4
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
2
,
A
p
r
il
201
8
:
9
0
8
–
9
1
6
916
BIO
G
RA
PHI
ES
O
F
AU
TH
O
R
S
B
e
n
a
k
a
S
a
n
t
h
o
s
h
a
S
,
As
sista
n
t
P
r
o
f
e
ss
o
r,
De
p
a
rt
m
e
n
t
o
f
El
e
c
tro
n
ics
a
n
d
Co
m
m
u
n
ica
ti
o
n
,
Co
o
rg
In
stit
u
te
o
f
T
e
c
h
n
o
lo
g
y
,
P
o
n
n
a
m
p
e
t
.
I
h
a
v
e
d
o
n
e
BE
De
g
re
e
i
n
El
e
c
tro
n
ics
a
n
d
Co
m
m
u
n
ica
ti
o
n
f
ro
m
V
T
U
in
2
0
1
0
.
I
h
a
v
e
M
.
T
e
c
h
De
g
re
e
in
Dig
it
a
l
El
e
c
tro
n
ics
a
n
d
C
o
m
m
u
n
ica
ti
o
n
S
y
ste
m
s
f
ro
m
V
T
U
in
2
0
1
2
.
I
h
a
v
e
5
y
e
a
rs
o
f
e
x
p
e
rien
c
e
in
tea
c
h
in
g
.
I
a
m
p
e
r
u
sin
g
P
h
D
i
n
V
T
U.
M
y
A
r
e
a
o
f
In
tere
st i
s Im
a
g
e
P
ro
c
e
ss
in
g
,
S
ig
n
a
l
P
r
o
c
e
ss
in
g
,
a
n
d
M
u
lt
im
e
d
ia.
N.
Ch
itra
K
ira
n
,
P
ro
f
e
ss
o
r
a
n
d
He
a
d
,
De
p
a
rtm
e
n
t
o
f
El
e
c
tro
n
ics
&
Co
m
m
u
n
ica
ti
o
n
En
g
i
n
e
e
rin
g
.
S
a
i
V
id
y
a
In
stit
u
te
o
f
T
e
c
h
n
o
l
o
g
y
,
B
e
n
g
a
lu
ru
,
S
h
e
h
a
s
d
o
n
e
P
h
D
in
El
e
c
tro
n
ics
&
Co
m
m
u
n
ica
ti
o
n
En
g
in
e
e
rig
.
Evaluation Warning : The document was created with Spire.PDF for Python.