I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
15
,
No
.
4
,
A
u
g
u
s
t
20
26
,
p
p
.
3
0
2
6
~
3
0
3
5
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
4
.
p
p
3
0
2
6
-
3
0
3
5
3026
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Lo
ng
sho
rt
-
term
memo
ry
bas
ed activ
ity det
ection us
ing
skeleton
joints da
ta:
a sy
stema
tic
revi
ew
B
a
k
k
a
l
a
Sa
ntha
K
um
a
r
,
R
.
Sh
a
nk
a
r
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
r
i
n
g
,
K
o
n
e
r
u
La
k
s
h
ma
i
a
h
E
d
u
c
a
t
i
o
n
F
o
u
n
d
a
t
i
o
n
,
V
a
d
d
e
sw
a
r
a
m,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
13
,
2
0
2
4
R
ev
is
ed
J
u
n
9
,
2
0
2
6
Acc
ep
ted
J
u
l
9
,
2
0
2
6
In
t
o
d
a
y
'
s
se
c
u
rit
y
a
n
d
s
u
rv
e
il
lan
c
e
a
p
p
li
c
a
ti
o
n
s,
re
c
o
g
n
izin
g
a
b
n
o
rm
a
l
a
c
ti
v
it
y
is
c
rit
ica
l
c
o
m
p
o
n
e
n
t
o
f
id
e
n
ti
fy
in
g
p
o
ss
i
b
le
h
a
z
a
rd
o
u
s
o
r
u
n
u
su
a
l
h
u
m
a
n
b
e
h
a
v
i
o
rs.
Th
e
re
is
n
e
e
d
fo
r
n
e
w
tec
h
n
o
lo
g
ies
th
a
t
c
a
n
d
e
tec
t
a
b
n
o
rm
a
l
h
u
m
a
n
b
e
h
a
v
io
rs
p
r
e
c
ise
ly
.
Th
e
p
re
se
n
t
re
v
iew
i
n
v
e
stig
a
tes
v
a
rio
u
s
a
sp
e
c
ts
o
f
d
e
tec
ti
o
n
p
ro
c
e
ss
,
fo
c
u
sin
g
o
n
sk
e
let
o
n
j
o
in
ts
in
p
u
t
d
a
ta.
It
th
e
n
e
x
p
l
o
re
s
a
d
o
p
ti
o
n
o
f
d
e
e
p
lea
rn
i
n
g
(DL)
a
rc
h
it
e
c
tu
re
s,
s
u
c
h
a
s
l
o
n
g
sh
o
rt
-
term
m
e
m
o
ry
(
L
S
TM
s
)
a
n
d
t
ra
n
sfo
rm
e
rs,
t
o
imp
r
o
v
e
a
c
c
u
ra
c
y
a
n
d
ro
b
u
stn
e
ss
o
f
re
c
o
g
n
it
io
n
m
o
d
e
ls.
Th
e
re
v
iew
e
x
p
l
o
re
s
sy
n
e
rg
ist
ic
in
teg
ra
ti
o
n
o
f
L
S
TM
s
a
n
d
t
ra
n
sf
o
rm
e
rs
to
imp
r
o
v
e
re
c
o
g
n
i
ti
o
n
o
f
u
n
u
su
a
l
a
c
ti
v
it
y
.
B
y
in
teg
ra
ti
n
g
LS
TM
s'
p
ro
c
e
ss
in
g
c
a
p
a
b
il
it
ies
a
n
d
a
tt
e
n
ti
o
n
m
e
c
h
a
n
ism
s
o
f
t
ra
n
sfo
rm
e
rs,
e
n
h
a
n
c
e
d
m
o
d
e
ls
c
a
n
a
c
c
u
ra
tel
y
i
d
e
n
ti
f
y
in
tri
c
a
te
p
a
tt
e
r
n
s
o
f
a
c
ti
v
it
y
.
De
sp
it
e
t
h
e
a
d
v
a
n
c
e
m
e
n
ts
t
h
a
t
h
a
v
e
b
e
e
n
m
a
d
e
in
th
e
field
,
t
h
e
c
h
a
ll
e
n
g
e
s
t
h
a
t
re
m
a
in
a
re
st
il
l
re
late
d
t
o
re
c
o
g
n
it
i
o
n
o
f
u
n
u
s
u
a
l
a
c
ti
v
it
ies
.
Th
e
se
i
n
c
lu
d
e
lac
k
o
f
sc
a
lab
il
it
y
f
o
r
lar
g
e
d
a
ta
se
ts,
n
e
e
d
fo
r
m
o
d
e
ls t
h
a
t
c
a
n
re
c
o
g
n
ize
c
o
m
p
lex
b
e
h
a
v
io
rs ac
ro
ss
d
iv
e
rse
a
p
p
li
c
a
ti
o
n
s,
a
n
d
n
e
e
d
t
o
e
n
su
re
t
h
a
t
d
e
tec
ti
o
n
is
p
e
rfo
rm
e
d
in
a
l
o
w
-
late
n
c
y
m
a
n
n
e
r.
Th
e
p
a
p
e
r
e
x
p
l
o
re
s
fu
t
u
re
d
irec
ti
o
n
s
o
f
d
e
v
e
lo
p
in
g
L
S
TM
-
b
a
se
d
m
o
d
e
l
s
th
a
t
c
a
n
re
c
o
g
n
ize
u
n
u
s
u
a
l
a
c
ti
v
it
y
u
sin
g
sk
e
leto
n
j
o
in
t
d
a
ta
in
a
c
l
o
u
d
-
b
a
se
d
e
n
v
iro
n
m
e
n
t.
T
h
e
re
v
iew
e
m
p
h
a
siz
e
s
th
e
p
o
ten
ti
a
l
o
f
su
c
h
so
l
u
ti
o
n
s
th
a
t
c
a
n
tak
e
a
d
v
a
n
tag
e
o
f
th
e
p
r
o
c
e
ss
in
g
p
o
we
r
o
f
g
ra
p
h
ics
p
ro
c
e
ss
in
g
u
n
i
ts
(
G
P
Us
)
a
n
d
ten
so
r
p
r
o
c
e
ss
in
g
u
n
it
s
(
T
P
Us
)
a
n
d
e
n
a
b
le
re
a
l
-
ti
m
e
re
c
o
g
n
it
i
o
n
o
f
a
c
ti
v
it
y
i
n
lar
g
e
d
a
tas
e
ts.
K
ey
w
o
r
d
s
:
Ab
n
o
r
m
al
ac
tiv
ity
r
ec
o
g
n
itio
n
Hu
m
an
ac
tiv
ity
r
ec
o
g
n
itio
n
L
o
n
g
-
s
h
o
r
t te
r
m
m
e
m
o
r
y
R
ea
l
-
tim
e
ac
tiv
ity
r
ec
o
g
n
itio
n
Sk
eleto
n
jo
in
ts
d
ata
an
aly
s
is
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
:
B
ak
k
ala
San
th
a
Ku
m
ar
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
K
o
n
er
u
L
ak
s
h
m
aiah
E
d
u
ca
tio
n
Fo
u
n
d
atio
n
Vad
d
eswar
am
,
An
d
h
r
a
Pra
d
es
h
,
I
n
d
ia
E
m
ail: san
th
ak
u
m
ar
.
b
ak
k
ala
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
I
n
r
ec
en
t
tim
es,
th
e
d
o
m
ain
o
f
h
u
m
an
ac
tiv
ity
r
ec
o
g
n
iti
o
n
(
HAR)
h
as
ex
p
er
ien
ce
d
r
em
ar
k
ab
le
p
r
o
g
r
ess
,
esp
ec
ially
th
r
o
u
g
h
t
h
e
in
co
r
p
o
r
atio
n
o
f
d
ee
p
lear
n
in
g
(
DL
)
tech
n
iq
u
es.
Am
o
n
g
th
ese
tech
n
iq
u
es,
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
S
T
M)
n
etwo
r
k
s
[
1
]
–
[
3
]
h
av
e
s
u
r
f
ac
ed
as
an
ef
f
ec
tiv
e
in
s
tr
u
m
en
t
f
o
r
ca
p
tu
r
in
g
tem
p
o
r
al
r
elatio
n
s
h
ip
s
in
s
eq
u
en
tial in
f
o
r
m
atio
n
.
T
h
is
s
y
s
tem
atic
r
ev
iew
f
o
cu
s
es o
n
t
h
e
a
p
p
licatio
n
o
f
L
STM
-
b
ased
ac
tiv
ity
d
etec
tio
n
tec
h
n
iq
u
es
th
at
le
v
er
ag
e
s
k
eleto
n
jo
in
t
d
ata,
wh
ic
h
p
r
o
v
id
es
a
co
m
p
ac
t
a
n
d
in
f
o
r
m
ativ
e
r
e
p
r
esen
tatio
n
o
f
h
u
m
an
p
o
s
tu
r
e
an
d
m
o
v
em
en
t
.
T
h
e
u
s
e
o
f
s
k
eleto
n
jo
in
ts
,
d
er
iv
ed
f
r
o
m
m
o
tio
n
ca
p
tu
r
e
s
y
s
tem
s
o
r
d
ep
t
h
ca
m
er
as,
allo
ws
f
o
r
r
o
b
u
s
t
ac
tiv
ity
r
ec
o
g
n
itio
n
,
ev
en
in
c
h
allen
g
in
g
e
n
v
ir
o
n
m
en
ts
.
Nev
er
th
eless
,
ev
en
with
th
e
e
n
co
u
r
a
g
in
g
f
in
d
i
n
g
s
h
i
g
h
lig
h
t
ed
in
th
e
liter
atu
r
e,
n
u
m
er
o
u
s
o
b
s
tacle
s
co
n
tin
u
e
to
ex
is
t,
s
u
ch
as
d
is
cr
ep
a
n
cie
s
in
jo
in
t
co
n
f
ig
u
r
atio
n
s
,
o
b
s
tr
u
ctio
n
s
,
an
d
th
e
in
tr
icac
y
o
f
h
u
m
an
ac
tiv
ities
.
Ad
d
itio
n
ally
,
co
m
p
u
tatio
n
al
f
ac
to
r
s
lik
e
m
o
d
el
ef
f
icien
cy
,
r
ea
l
-
tim
e
p
r
o
ce
s
s
in
g
ab
ilit
ies,
an
d
s
ca
lab
ilit
y
ar
e
cr
u
cial
f
o
r
p
r
ac
tical
im
p
lem
e
n
tatio
n
s
in
ev
er
-
ch
an
g
in
g
en
v
ir
o
n
m
e
n
ts
.
T
h
is
r
ev
iew
in
ten
d
s
to
co
n
s
o
lid
ate
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Lo
n
g
s
h
o
r
t
-
term me
mo
r
y
b
a
s
ed
a
ctivity
d
etec
tio
n
u
s
in
g
s
ke
leto
n
…
(
B
a
kk
a
la
S
a
n
th
a
K
u
ma
r
)
3027
cu
r
r
en
t
r
esear
ch
o
n
L
STM
-
b
ased
ac
tiv
ity
d
etec
tio
n
u
tili
zin
g
s
k
eleto
n
jo
in
ts
,
u
n
d
er
s
co
r
e
th
e
f
u
n
d
am
e
n
tal
ch
allen
g
es
en
co
u
n
te
r
ed
b
y
r
esear
ch
er
s
,
an
d
ex
am
in
e
c
o
m
p
u
tatio
n
al
ap
p
r
o
ac
h
es
th
at
co
u
ld
im
p
r
o
v
e
p
er
f
o
r
m
an
ce
a
n
d
ap
p
licab
ilit
y
in
r
ea
l
-
w
o
r
ld
c
o
n
tex
ts
.
B
y
tack
lin
g
th
ese
elem
e
n
ts
,
th
e
r
ev
ie
w
aid
s
in
f
o
s
ter
in
g
a
m
o
r
e
co
m
p
r
eh
en
s
iv
e
u
n
d
er
s
tan
d
in
g
o
f
cu
ttin
g
-
ed
g
e
d
ev
elo
p
m
en
ts
an
d
lay
s
th
e
g
r
o
u
n
d
wo
r
k
f
o
r
f
u
tu
r
e
r
esear
ch
tr
ajec
to
r
ies in
th
is
ad
v
an
cin
g
f
ield
.
R
ec
en
t
p
r
o
g
r
ess
in
s
k
eleto
n
-
b
ased
HAR
h
as
b
ee
n
b
ased
o
n
eith
er
r
ec
u
r
r
en
t
n
etwo
r
k
s
,
e
.
g
.
L
STM
n
etwo
r
k
s
,
o
r
atten
tio
n
-
b
ased
t
r
an
s
f
o
r
m
er
m
o
d
els
[
4
]
–
[
7
]
to
ca
p
tu
r
e
tem
p
o
r
al
d
ep
e
n
d
en
ce
in
jo
in
t
-
lev
el
m
o
tio
n
d
ata.
Alth
o
u
g
h
L
STM
n
etwo
r
k
s
ar
e
b
est
ap
p
lied
in
le
ar
n
in
g
s
h
o
r
t
a
n
d
m
ed
iu
m
-
ter
m
tim
e
d
y
n
am
ics
in
s
k
eleto
n
s
eq
u
en
ce
s
,
th
e
y
ten
d
to
f
ail
to
ca
p
tu
r
e
l
o
n
g
-
r
an
g
e
t
em
p
o
r
al
d
y
n
am
ics
a
n
d
m
u
lti
-
jo
in
t
in
ter
ac
tio
n
s
in
ac
tiv
ities
o
v
er
lo
n
g
tim
e
s
ca
les.
T
r
an
s
f
o
r
m
e
r
a
r
ch
itectu
r
es,
o
n
th
e
s
id
e,
a
r
e
ef
f
ec
tiv
e
in
b
e
in
g
ab
le
to
c
o
m
p
u
te
g
lo
b
al
tem
p
o
r
al
m
o
d
ellin
g
v
ia
s
elf
-
atten
tio
n
.
B
u
t
t
r
an
s
f
o
r
m
er
s
co
n
s
u
m
e
lar
g
e
-
s
ca
le
d
ata
an
d
h
ig
h
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
,
wh
i
ch
r
estrict
th
eir
ap
p
licatio
n
i
n
r
ea
l
-
tim
e
o
r
r
eso
u
r
ce
-
co
n
s
tr
ain
ed
s
ettin
g
s
.
T
h
e
n
o
v
elty
o
f
th
is
r
ev
iew
is
in
th
e
s
y
s
tem
atic
s
tu
d
y
o
f
h
y
b
r
id
L
STM
-
t
r
an
s
f
o
r
m
er
m
o
d
els.
T
h
e
s
y
n
er
g
is
tic
co
m
b
in
atio
n
p
r
o
v
i
d
es
a
b
alan
ce
d
m
o
d
ellin
g
p
ar
a
d
ig
m
,
wh
i
ch
o
v
er
co
m
es
th
e
s
in
g
u
lar
ar
ch
itectu
r
e
lim
itatio
n
s
wh
en
u
s
ed
in
d
iv
i
d
u
ally
.
Mo
r
eo
v
e
r
,
th
e
in
ter
s
ec
tio
n
o
f
L
STM
-
t
r
an
s
f
o
r
m
e
r
m
o
d
els
with
clo
u
d
-
b
ase
d
co
m
p
u
tin
g
ar
ea
s
o
p
e
n
s
a
n
o
v
el
d
im
en
s
io
n
o
f
r
esear
ch
th
at
h
as
n
o
t
b
ee
n
a
d
eq
u
ately
e
x
am
in
ed
in
ea
r
lier
r
ev
iews.
C
lo
u
d
in
f
r
astru
ctu
r
es
[
8
]
ca
n
b
e
s
ca
led
to
s
u
p
p
o
r
t
tr
ain
in
g
,
d
is
tr
ib
u
ted
in
f
er
e
n
ce
[
9
]
,
a
n
d
r
ea
l
-
tim
e
d
ep
lo
y
m
e
n
t
o
f
c
o
m
p
u
tatio
n
ally
in
ten
s
iv
e
h
y
b
r
id
m
o
d
el
s
,
wh
ich
a
r
e
esp
ec
ially
u
s
ef
u
l
to
lar
g
e
-
s
ca
le
s
k
eleto
n
-
b
ased
HAR
s
y
s
tem
s
,
in
clu
d
in
g
s
m
ar
t
s
u
r
v
eilla
n
ce
,
h
ea
lth
ca
r
e
m
o
n
ito
r
in
g
,
an
d
h
u
m
an
-
r
o
b
o
t
in
ter
ac
tio
n
(
HR
I
)
.
T
h
e
r
ev
iew
also
in
d
icate
s
th
at
clo
u
d
-
en
a
b
led
L
STM
-
t
r
an
s
f
o
r
m
e
r
p
ip
elin
es a
r
e
n
o
v
el,
wh
e
r
e
in
d
u
s
tr
ial
s
o
p
h
is
ticatio
n
s
at
e
d
g
e
p
a
r
ts
co
n
s
is
t
o
f
lig
h
tweig
h
t
s
k
eleto
n
e
x
tr
ac
tio
n
,
an
d
d
e
ep
tem
p
o
r
al
m
o
d
els
an
d
atten
tio
n
-
in
f
u
s
ed
in
f
er
e
n
ce
ar
e
r
u
n
o
n
cl
o
u
d
s
er
v
er
s
.
T
h
is
s
tu
d
y
s
tan
d
s
o
u
t
as
o
n
e
o
f
th
e
f
ir
s
t
r
ev
iew
p
ap
er
s
to
f
ill
t
h
e
s
eq
u
e
n
ce
m
o
d
ellin
g
in
n
o
v
atio
n
(
L
STM
-
t
r
a
n
s
f
o
r
m
er
f
u
s
io
n
)
a
n
d
p
r
ac
tical
clo
u
d
-
b
ased
HAR
d
ep
lo
y
m
e
n
t,
c
r
ea
tin
g
g
ap
s
i
n
th
e
r
esear
ch
an
d
id
e
n
tify
in
g
o
p
p
o
r
tu
n
ities
to
war
d
s
s
ca
lab
le
an
d
s
m
ar
t
ac
tiv
ity
r
ec
o
g
n
itio
n
s
y
s
tem
s
.
T
o
co
n
tin
u
e
en
h
a
n
cin
g
th
e
s
tr
u
ctu
r
al
p
u
r
ity
o
f
HAR
r
ese
ar
ch
o
f
s
k
eleta
l
ty
p
e
f
o
r
m
,
t
h
is
r
ev
iew
s
u
g
g
ests
a
g
r
a
p
h
ical
ta
x
o
n
o
m
y
,
wh
ich
p
r
ev
is
u
alize
s
a
m
eth
o
d
ically
ca
teg
o
r
ized
s
et
o
f
m
o
d
el
i
n
f
r
astru
ctu
r
e
an
d
d
ep
lo
y
m
en
t
s
tr
ateg
ies
u
tili
ze
d
b
y
L
STM
-
an
d
t
r
an
s
f
o
r
m
er
-
b
ased
f
r
am
ewo
r
k
s
.
I
n
co
n
tr
ast
to
o
th
er
p
r
ev
io
u
s
s
u
r
v
ey
s
,
wh
ic
h
b
o
il
r
elev
an
t
s
tu
d
ies
d
o
wn
to
co
m
p
ar
is
o
n
s
o
n
a
tab
u
la
r
ca
s
e
o
r
n
ar
r
ativ
e
an
aly
s
is
,
th
e
p
r
o
p
o
s
ed
tax
o
n
o
m
y
will
p
r
o
v
id
e
a
m
u
lti
-
d
im
e
n
s
io
n
al
g
r
ap
h
ical
s
tr
u
ctu
r
e
w
h
ich
r
an
k
s
ex
is
tin
g
s
tu
d
ies
o
n
3
o
r
th
o
g
o
n
al
s
ca
les
.
T
h
ese
s
ca
le
s
in
clu
d
e:
i)
m
o
d
el
ar
ch
itectu
r
e,
wh
ich
is
h
o
u
s
ed
in
s
tan
d
alo
n
e
L
STM
m
o
d
els,
p
u
r
e
t
r
an
s
f
o
r
m
er
-
b
ased
m
o
d
els
a
n
d
m
ix
e
d
L
STM
-
t
r
an
s
f
o
r
m
er
p
ip
elin
e
s
;
ii)
tem
p
o
r
al
an
d
s
p
atial
m
o
d
elin
g
s
tr
ateg
ies,
in
clu
d
in
g
jo
in
t
-
lev
el
en
c
o
d
in
g
,
s
p
atio
-
tem
p
o
r
al
g
r
ap
h
r
ep
r
esen
tatio
n
s
an
d
atten
tio
n
-
b
ased
in
ter
-
jo
in
t
d
e
p
en
d
en
cy
lear
n
i
n
g
;
an
d
iii)
d
e
p
lo
y
m
en
t
s
ch
em
es,
wh
ich
in
c
lu
d
e
ed
g
e
-
b
ased
in
s
u
ch
v
is
u
al
ab
s
tr
ac
tio
n
allo
ws
th
e
r
ea
d
er
s
to
ea
s
ily
r
ec
o
g
n
ize
ar
ch
itectu
r
al
tr
en
d
,
p
r
ev
ailin
g
m
o
d
elin
g
p
r
ef
er
en
ce
s
,
a
n
d
u
n
ch
ar
te
d
o
p
t
io
n
s
in
th
e
liter
atu
r
e.
T
h
e
tax
o
n
o
m
y
also
f
ac
t
o
r
s
in
d
ep
lo
y
m
en
t
-
c
o
n
s
cio
u
s
d
esig
n
f
ac
to
r
s
,
p
o
in
tin
g
to
th
e
ef
f
ec
ts
o
f
task
-
o
r
ien
ted
co
n
s
tr
ain
ts
o
n
co
m
p
u
tatio
n
al
co
s
ts
,
laten
cy
to
le
r
an
ce
,
an
d
s
ca
lab
ilit
y
co
n
ce
r
n
s
in
th
e
o
p
e
r
atio
n
al
tip
o
f
th
e
iceb
er
g
o
n
th
e
ar
ch
itectu
r
e.
T
h
e
m
ap
p
in
g
o
f
h
y
b
r
id
L
STM
-
t
r
an
s
f
o
r
m
er
m
o
d
el
s
to
clo
u
d
-
ce
n
ter
ed
an
d
ed
g
e
-
clo
u
d
c
o
llab
o
r
ativ
e
m
o
d
els
allo
ws
ex
p
lo
r
atio
n
o
f
an
em
er
g
in
g
d
ir
ec
tio
n
in
r
esear
ch
with
th
e
in
v
o
lv
em
e
n
t
o
f
s
atis
f
y
in
g
tim
elin
ess
m
o
d
elin
g
,
in
tim
ac
y
,
an
d
d
ep
lo
y
m
en
t
ef
f
icien
cy
.
I
t
is
esp
ec
ially
th
e
co
n
tr
ib
u
tio
n
th
at
is
v
er
y
n
e
w
to
a
s
y
s
tem
atic
r
ev
iew
,
a
s
it
b
r
in
g
s
th
e
alg
o
r
ith
m
ic
i
n
n
o
v
atio
n
an
d
th
e
s
y
s
tem
-
lev
el
d
ep
lo
y
m
en
t
s
tr
a
teg
ies
in
to
co
n
tact,
p
r
o
v
id
in
g
r
esear
ch
er
s
an
d
p
r
ac
titi
o
n
er
s
with
a
s
tr
u
ctu
r
ed
v
iewp
o
in
t o
n
th
e
cu
r
r
en
t w
o
r
k
an
d
th
e
f
u
tu
r
e
o
p
p
o
r
tu
n
ities
o
f
r
esear
ch
.
T
h
e
s
u
g
g
ested
v
is
u
al
tax
o
n
o
m
y
as
s
h
o
wn
in
Fig
u
r
e
1
ca
n
b
e
u
s
ed
as a
s
u
m
m
ar
izatio
n
to
o
l,
as we
ll a
s
a
d
esig
n
tem
p
late
f
o
r
n
e
x
t
-
g
en
er
atio
n
,
s
ca
lab
le
s
k
eleto
n
-
b
ased
HAR
s
y
s
tem
s
.
T
h
e
s
tu
d
y
will
ex
p
lo
r
e
in
n
o
v
ativ
e
a
p
p
r
o
ac
h
es
th
at
lev
er
ag
e
DL
tech
n
i
q
u
es
to
im
p
r
o
v
e
ac
c
u
r
ac
y
a
n
d
r
o
b
u
s
tn
ess
in
ac
tiv
ity
r
ec
o
g
n
itio
n
s
y
s
tem
s
.
Fu
r
th
er
m
o
r
e,
th
e
in
teg
r
atio
n
o
f
m
u
lt
im
o
d
al
d
ata
s
o
u
r
ce
s
[
1
0
]
,
[
1
1
]
will
b
e
ex
am
in
ed
,
wh
ich
ca
n
p
r
o
v
id
e
r
ich
e
r
c
o
n
tex
tu
al
in
f
o
r
m
atio
n
a
n
d
p
o
t
en
tially
lead
t
o
m
o
r
e
n
u
a
n
ce
d
in
ter
p
r
etatio
n
s
o
f
h
u
m
an
ac
tiv
ities
.
Fu
r
th
er
m
o
r
e
,
tr
an
s
f
er
lear
n
in
g
will
b
e
ex
am
in
e
d
f
o
r
its
co
n
t
r
ib
u
tio
n
t
o
th
e
a
d
ju
s
tm
en
t
o
f
m
o
d
els
f
o
r
v
ar
io
u
s
en
v
ir
o
n
m
e
n
ts
,
allo
win
g
s
y
s
tem
s
to
ac
h
iev
e
b
etter
g
en
er
aliza
tio
n
ac
r
o
s
s
a
r
an
g
e
o
f
co
n
tex
ts
an
d
p
o
p
u
latio
n
s
[
1
2
]
.
T
h
e
eth
ica
l
co
n
s
id
er
atio
n
s
in
v
o
lv
ed
in
im
p
lem
en
tin
g
th
ese
tech
n
o
l
o
g
ies
in
p
r
ac
tical
ap
p
licatio
n
s
will
also
b
e
co
n
s
id
er
ed
,
en
s
u
r
in
g
th
at
p
r
iv
ac
y
a
n
d
s
ec
u
r
ity
is
s
u
es
ar
e
th
o
r
o
u
g
h
ly
ad
d
r
ess
ed
wh
ile
o
p
tim
izin
g
t
h
e
ad
v
an
tag
es
o
f
im
p
r
o
v
ed
ac
tiv
ity
r
ec
o
g
n
itio
n
.
Mo
r
eo
v
e
r
,
th
e
s
ig
n
if
ica
n
ce
o
f
co
n
tin
u
o
u
s
lear
n
in
g
an
d
ad
ap
tatio
n
wit
h
in
th
ese
m
o
d
els
will
b
e
h
ig
h
lig
h
ted
,
s
tr
ess
in
g
th
e
n
ec
ess
ity
f
o
r
r
e
g
u
lar
ass
es
s
m
en
t a
n
d
en
h
a
n
ce
m
en
t t
o
k
ee
p
u
p
with
c
h
an
g
in
g
u
s
er
d
em
an
d
s
an
d
tech
n
o
lo
g
ical
p
r
o
g
r
ess
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
26
:
3
0
2
6
-
3
0
3
5
3028
2.
RE
VI
E
W
M
E
T
H
O
D
T
h
e
en
tire
r
ev
iew
p
r
o
ce
s
s
is
ca
r
r
ied
o
u
t
b
y
f
r
am
in
g
t
h
e
r
ese
ar
ch
q
u
esti
o
n
s
th
at
ca
n
c
o
n
tr
i
b
u
te
to
th
e
u
n
d
er
s
tan
d
i
n
g
o
f
a
b
n
o
r
m
al
ac
tiv
ity
r
ec
o
g
n
itio
n
u
s
in
g
s
k
elet
o
n
jo
in
ts
,
h
o
w
t
o
en
h
an
ce
t
h
e
d
etec
tio
n
ac
cu
r
ac
y
,
th
e
r
o
le
o
f
L
STM
-
b
ased
m
o
d
els
in
an
aly
zin
g
th
e
s
k
ele
to
n
jo
i
n
ts
,
ch
allen
g
es
to
o
v
er
co
m
e
,
an
d
f
u
t
u
r
e
d
ir
ec
tio
n
s
th
at
ca
n
g
u
i
d
e
in
a
n
aly
zin
g
t
h
e
an
aly
s
is
o
f
s
k
eleto
n
jo
in
ts
d
ata
a
n
d
p
er
f
o
r
m
t
h
e
r
ec
o
g
n
itio
n
wit
h
h
ig
h
ef
f
icien
cy
.
T
h
e
m
ajo
r
co
n
tr
ib
u
tio
n
o
f
th
e
r
ev
iew
is
to
a
d
d
r
ess
th
e
f
o
llo
win
g
f
r
am
e
d
r
e
s
ea
r
ch
q
u
esti
o
n
s
.
−
R
Q1
:
h
o
w
d
o
s
k
eleto
n
j
o
in
ts
co
n
tr
ib
u
te
to
th
e
d
etec
tio
n
o
f
n
o
r
m
al
an
d
ab
n
o
r
m
al
h
u
m
an
b
e
h
av
io
r
?
−
R
Q2
:
h
o
w
d
o
th
e
tr
a
n
s
f
o
r
m
ati
o
n
m
o
d
els co
n
tr
ib
u
te
to
ac
tiv
it
y
d
etec
tio
n
?
−
R
Q3
:
h
o
w
a
r
e
L
STM
-
b
ased
m
o
d
els
ef
f
icien
t
in
an
aly
zin
g
th
e
s
k
eleto
n
j
o
in
ts
d
ata
an
d
co
n
d
u
ctin
g
th
e
d
etec
tio
n
well?
−
R
Q4
:
wh
at
ar
e
th
e
d
if
f
er
en
t
ch
allen
g
es
in
d
etec
tin
g
d
if
f
e
r
en
t
ac
tiv
ities
u
s
in
g
th
e
cu
r
r
en
t
r
ec
o
g
n
itio
n
f
r
am
ewo
r
k
s
an
d
f
u
tu
r
e
s
co
p
e?
G
o
o
g
l
e
S
ea
r
c
h
e
n
g
i
n
e
i
s
u
s
e
d
to
s
e
a
r
c
h
f
o
r
r
e
l
e
v
a
n
t
a
r
t
i
cl
e
s
p
u
b
l
i
s
h
e
d
i
n
v
a
r
i
o
u
s
r
e
p
u
t
e
d
j
o
u
r
n
a
l
s
.
T
h
e
s
e
a
r
c
h
e
n
g
i
n
e
p
e
r
f
o
r
m
e
d
t
h
e
s
e
a
r
c
h
i
n
m
a
n
y
r
e
p
u
t
e
d
d
at
a
b
ase
s
li
k
e
S
p
r
i
n
g
e
r
,
I
E
E
E
X
p
l
o
r
e
,
a
n
d
S
c
ie
n
c
e
Di
r
e
c
t
.
T
h
e
s
e
a
r
c
h
p
r
o
c
es
s
is
c
o
n
d
u
c
te
d
u
s
i
n
g
k
e
y
w
o
r
d
s
c
o
m
b
i
n
e
d
w
i
t
h
t
h
e
A
ND
B
o
o
l
e
a
n
o
p
e
r
a
to
r
.
A
l
l
t
h
e
d
a
t
a
b
as
e
s
w
e
r
e
e
x
p
l
o
r
e
d
t
o
i
d
e
n
t
i
f
y
t
h
e
a
r
t
i
c
l
es
t
o
b
e
c
o
n
s
i
d
e
r
e
d
f
o
r
r
e
v
i
e
w
s
y
s
t
e
m
a
ti
c
a
ll
y
a
n
d
p
r
o
v
i
d
e
a
n
o
p
p
o
r
t
u
n
i
t
y
t
o
i
m
p
r
o
v
e
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
t
h
e
m
o
d
e
l
s
a
n
d
p
r
e
d
i
ct
t
h
e
a
ct
i
v
it
i
e
s
m
o
s
t
e
f
f
ic
i
e
n
tl
y
.
F
o
u
r
k
e
y
w
o
r
d
s
we
r
e
s
e
l
e
ct
e
d
b
a
s
e
d
o
n
t
h
e
r
e
s
e
a
r
c
h
q
u
e
s
t
i
o
n
s
.
T
h
e
s
e
s
e
a
r
c
h
w
o
r
d
s
w
e
r
e
c
o
m
b
i
n
e
d
d
u
r
i
n
g
t
h
e
a
d
v
a
n
c
e
d
s
e
a
r
c
h
p
r
o
c
e
s
s
t
o
i
d
e
n
t
i
f
y
m
o
r
e
r
e
l
e
v
a
n
t
a
r
t
i
c
l
es
.
T
a
b
l
e
1
s
h
o
ws
t
h
e
li
s
t
o
f
k
e
y
wo
r
d
s
u
s
e
d
d
u
r
i
n
g
t
h
e
r
e
v
i
e
w
p
r
o
c
e
s
s
.
T
ab
le
1
.
Key
wo
r
d
s
u
s
ed
d
u
r
in
g
th
e
s
ea
r
ch
D
a
t
a
b
a
s
e
I
n
i
t
i
a
l
s
e
a
r
c
h
w
o
r
d
s
A
d
v
a
n
c
e
d
s
e
a
r
c
h
w
o
r
d
s
I
EEE
X
p
l
o
r
e
LSTM
mo
d
e
l
s
LSTM
mo
d
e
l
s [A
N
D
]
t
r
a
n
sf
o
r
mers
S
p
r
i
n
g
e
r
Tr
a
n
sf
o
r
mer
m
o
d
e
ls
LSTM
mo
d
e
l
s [A
N
D
]
h
u
m
a
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
S
c
i
e
n
c
e
D
i
r
e
c
t
H
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
HAR
[AND]
sk
e
l
e
t
o
n
j
o
i
n
t
s
d
a
t
a
O
t
h
e
r
s
S
k
e
l
e
t
o
n
j
o
i
n
t
s
d
a
t
a
LSTM
m
o
d
e
l
s
[
A
N
D
]
t
r
a
n
sf
o
r
m
e
r
s
[
A
N
D
]
HAR
[AND]
sk
e
l
e
t
o
n
j
o
i
n
t
s
d
a
t
a
a
n
a
l
y
s
i
s
A
s
y
s
tem
atic
s
cr
ee
n
in
g
p
r
o
ce
d
u
r
e
was e
m
p
lo
y
ed
in
ac
co
r
d
an
ce
with
estab
li
s
h
ed
s
y
s
tem
atic
liter
atu
r
e
r
ev
iew
an
d
PR
I
SMA
r
ep
o
r
tin
g
g
u
id
elin
es
[
1
3
]
,
[
1
4
]
.
St
u
d
ies
p
u
b
lis
h
ed
b
etwe
en
2
0
1
9
an
d
2
0
2
4
wer
e
co
n
s
id
er
ed
,
f
o
c
u
s
in
g
on
HAR
,
u
s
in
g
s
k
eleto
n
jo
in
t
d
ata,
an
d
m
en
tio
n
in
g
L
STM
,
t
r
an
s
f
o
r
m
er
,
o
r
r
elate
d
DL
ar
ch
itectu
r
es.
A
f
ew
p
a
p
er
s
wer
e
ex
clu
d
e
d
f
r
o
m
th
e
r
ev
ie
w
th
at
d
id
n
o
t
ap
p
ly
th
e
p
r
o
p
o
s
ed
tech
n
i
q
u
es
to
s
k
eleto
n
-
b
ased
ac
tiv
ity
d
etec
tio
n
.
T
h
e
r
e
m
ain
in
g
a
r
ticles
wer
e
s
cr
ee
n
ed
th
r
o
u
g
h
id
en
ti
f
icatio
n
,
d
u
p
licate
r
em
o
v
al,
r
elev
an
ce
ass
ess
m
e
n
t,
elig
ib
ilit
y
c
h
ec
k
in
g
,
an
d
f
u
ll
-
tex
t
ev
al
u
atio
n
b
ef
o
r
e
f
i
n
al
in
clu
s
io
n
in
th
e
r
ev
iew.
Fig
u
r
e
1
s
u
m
m
ar
izes
all
th
e
s
tag
es u
s
in
g
a
s
y
s
tem
atic
ar
ticle
-
s
elec
tio
n
f
lo
w.
Fig
u
r
e
1
.
Sy
s
tem
atic
ar
ticle
s
elec
tio
n
p
r
o
ce
s
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Lo
n
g
s
h
o
r
t
-
term me
mo
r
y
b
a
s
ed
a
ctivity
d
etec
tio
n
u
s
in
g
s
ke
leto
n
…
(
B
a
kk
a
la
S
a
n
th
a
K
u
ma
r
)
3029
T
h
e
in
itial
s
ea
r
ch
g
e
n
er
ated
1
2
6
a
r
ticles
,
b
u
t
o
n
ly
8
5
wer
e
c
o
n
s
id
er
ed
b
ased
o
n
th
e
r
elev
a
n
ce
an
d
th
e
y
ea
r
o
f
p
u
b
licatio
n
.
T
h
e
f
u
r
th
er
s
ea
r
ch
p
r
o
ce
s
s
is
co
n
d
u
cte
d
b
y
f
in
e
-
tu
n
in
g
t
h
e
s
ea
r
ch
k
ey
wo
r
d
s
u
s
in
g
th
e
co
m
m
o
n
k
ey
wo
r
d
'
h
u
m
an
ac
t
iv
ities
r
ec
o
g
n
itio
n
'
.
Fin
ally
,
5
2
ar
ticles
wer
e
co
n
s
id
er
e
d
f
o
r
th
e
f
in
al
r
ev
iew
p
r
o
ce
s
s
.
Fig
u
r
e
1
s
h
o
ws th
e
p
r
o
ce
s
s
f
lo
w
d
iag
r
am
s
u
m
m
a
r
izin
g
all
th
e
s
tep
s
.
3.
DIS
CU
SS
I
O
N
S
T
h
e
s
ec
tio
n
p
r
esen
ts
th
e
f
in
d
i
n
g
s
o
f
th
e
r
ev
iewe
d
ar
ticles
in
r
elatio
n
to
th
e
f
o
u
r
r
esear
ch
q
u
esti
o
n
s
f
o
r
m
u
lated
f
o
r
t
h
is
s
tu
d
y
.
T
h
e
d
is
cu
s
s
io
n
is
o
r
g
a
n
ized
a
cc
o
r
d
in
g
to
ea
ch
r
esear
ch
q
u
esti
o
n
to
en
s
u
r
e
a
s
y
s
tem
atic
in
ter
p
r
etatio
n
o
f
th
e
liter
atu
r
e.
T
h
e
s
u
b
s
ec
tio
n
s
h
ig
h
lig
h
t
th
e
m
ajo
r
a
p
p
r
o
ac
h
es,
f
in
d
in
g
s
,
ch
allen
g
es,
an
d
r
esear
ch
o
p
p
o
r
tu
n
ities
id
en
tifie
d
ac
r
o
s
s
th
e
s
elec
ted
s
tu
d
ies.
3
.
1
.
RQ
1
:
un
dersta
nd
ing
a
b
no
rm
a
l a
ct
i
v
it
y
re
c
o
g
nitio
n:
t
he
ro
le
o
f
s
k
ele
t
o
n j
o
ints
Fo
r
th
is
r
esear
ch
q
u
esti
o
n
,
3
5
p
ap
er
s
wer
e
co
n
s
id
er
e
d
b
y
f
ilter
in
g
th
em
with
th
e
r
o
le
o
f
s
k
eleto
n
jo
in
ts
in
u
n
d
er
s
tan
d
in
g
an
d
r
ec
o
g
n
izin
g
ab
n
o
r
m
al
b
e
h
av
io
r
.
At
last
,
9
p
ap
er
s
wer
e
co
n
s
id
er
ed
f
o
r
an
aly
s
is
.
T
h
is
r
ev
iew
aim
s
to
ex
p
lo
r
e
v
ar
io
u
s
m
eth
o
d
o
l
o
g
ies
em
p
lo
y
ed
in
d
etec
tin
g
an
d
in
t
er
p
r
etin
g
ab
n
o
r
m
al
ac
tiv
ities
th
r
o
u
g
h
th
e
an
aly
s
is
o
f
s
k
eleta
l
jo
in
t
m
o
v
em
en
ts
,
h
ig
h
lig
h
tin
g
a
d
v
an
ce
m
e
n
ts
in
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es
an
d
th
eir
ap
p
licatio
n
s
in
r
ea
l
-
tim
e
m
o
n
ito
r
in
g
s
y
s
tem
s
.
Fu
r
th
er
m
o
r
e,
th
e
in
teg
r
atio
n
o
f
d
e
p
th
s
en
s
o
r
s
an
d
co
m
p
u
ter
v
is
io
n
h
as
s
ig
n
if
ican
tly
e
n
h
an
ce
d
th
e
ac
cu
r
ac
y
o
f
th
ese
s
y
s
tem
s
,
en
ab
lin
g
m
o
r
e
n
u
an
ce
d
in
ter
p
r
etatio
n
s
o
f
h
u
m
an
b
eh
a
v
io
r
in
d
iv
er
s
e
en
v
ir
o
n
m
en
ts
.
Ad
d
itio
n
ally
,
th
e
u
s
e
o
f
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NNs)
h
as
p
r
o
v
en
ef
f
ec
tiv
e
in
ex
tr
ac
tin
g
f
ea
tu
r
es
f
r
o
m
s
k
eleta
l
d
ata,
al
lo
win
g
f
o
r
im
p
r
o
v
ed
class
if
icatio
n
o
f
ac
tiv
ities
an
d
t
h
e
id
e
n
tific
atio
n
o
f
s
u
b
tle
d
ev
iatio
n
s
th
at
m
ay
in
d
icate
ab
n
o
r
m
al
b
e
h
av
io
r
[
1
5
]
,
[
1
6
]
.
I
n
ad
d
itio
n
to
th
e
f
o
u
n
d
atio
n
al
r
o
le
o
f
s
k
eleto
n
jo
in
ts
in
r
ec
o
g
n
izin
g
ab
n
o
r
m
a
l
ac
tiv
ities
,
r
ec
en
t
ad
v
an
ce
m
e
n
ts
h
av
e
h
ig
h
lig
h
ted
th
e
im
p
o
r
tan
ce
o
f
in
teg
r
atin
g
m
u
ltimo
d
al
d
ata
f
o
r
en
h
a
n
ce
d
p
e
r
f
o
r
m
an
ce
[
1
7
]
.
B
y
em
p
lo
y
in
g
C
NNs
alo
n
g
s
id
e
atten
tio
n
-
b
ased
L
STM
m
o
d
els,
r
esear
ch
er
s
ca
n
ef
f
ec
tiv
ely
m
er
g
e
d
ee
p
s
p
atial
in
f
o
r
m
atio
n
with
tem
p
o
r
al
d
y
n
am
ics,
th
er
eb
y
im
p
r
o
v
in
g
th
e
e
x
tr
ac
tio
n
o
f
k
e
y
b
eh
a
v
io
r
s
f
r
o
m
co
m
p
lex
d
at
asets
[
1
8
]
.
I
n
ad
d
itio
n
t
o
th
e
eth
ical
co
n
s
id
er
atio
n
s
s
u
r
r
o
u
n
d
in
g
a
b
n
o
r
m
al
ac
tiv
ity
r
ec
o
g
n
itio
n
s
y
s
tem
s
,
it
i
s
im
p
er
ativ
e
to
ad
d
r
ess
th
e
tech
n
ical
ch
allen
g
es
th
at
p
er
s
is
t
in
ac
h
iev
in
g
o
p
tim
al
p
e
r
f
o
r
m
a
n
ce
.
R
ec
en
t
s
tu
d
ies
h
av
e
s
h
o
wn
th
at
em
p
lo
y
in
g
r
ec
u
r
r
en
t
n
eu
r
al
n
etwo
r
k
s
(
R
NNs)
ca
n
s
ig
n
if
ican
tly
en
h
an
ce
th
e
an
aly
s
is
o
f
tem
p
o
r
al
d
ep
e
n
d
en
cies
with
in
s
k
eleta
l
d
ata,
th
er
eb
y
im
p
r
o
v
in
g
ac
cu
r
ac
y
in
r
ec
o
g
n
izin
g
b
o
th
n
o
r
m
al
an
d
an
o
m
alo
u
s
ac
tiv
ities
[
1
9
]
,
[
2
0
]
.
As
th
e
lan
d
s
ca
p
e
o
f
ab
n
o
r
m
al
ac
tiv
ity
r
ec
o
g
n
itio
n
ev
o
lv
es,
th
e
p
o
ten
tial
f
o
r
in
teg
r
atin
g
ad
v
an
ce
d
m
ac
h
in
e
lear
n
in
g
tech
n
i
q
u
es
with
r
ea
l
-
tim
e
d
ata
s
tr
ea
m
s
p
r
esen
ts
a
c
o
m
p
ellin
g
f
r
o
n
tier
.
Fo
r
in
s
tan
ce
,
th
e
ap
p
licatio
n
o
f
atten
tio
n
m
ec
h
an
is
m
s
with
in
L
STM
f
r
am
ewo
r
k
s
ca
n
en
h
an
ce
th
e
m
o
d
el'
s
s
en
s
itiv
ity
to
cr
itical
m
o
m
en
ts
in
h
u
m
an
b
eh
av
io
r
,
allo
win
g
f
o
r
m
o
r
e
ac
cu
r
ate
d
etec
ti
o
n
o
f
an
o
m
alies
as
th
ey
o
cc
u
r
.
Mo
r
eo
v
er
,
ex
p
lo
r
in
g
th
e
u
s
e
o
f
g
e
n
er
ativ
e
a
d
v
er
s
ar
ial
n
et
wo
r
k
s
(
GANs)
co
u
ld
p
r
o
v
id
e
in
n
o
v
ativ
e
way
s
to
au
g
m
en
t
tr
ai
n
in
g
d
at
asets
,
ad
d
r
ess
in
g
is
s
u
es
o
f
b
ias
an
d
im
p
r
o
v
in
g
g
en
er
aliza
b
ilit
y
ac
r
o
s
s
d
iv
er
s
e
p
o
p
u
latio
n
s
[
2
1
]
.
As
th
e
in
teg
r
atio
n
o
f
ad
v
a
n
ce
d
m
ac
h
in
e
le
ar
n
in
g
tec
h
n
iq
u
es
c
o
n
tin
u
es
t
o
r
esh
ap
e
a
b
n
o
r
m
al
ac
tiv
ity
r
ec
o
g
n
itio
n
,
it
b
ec
o
m
es
in
cr
ea
s
in
g
ly
im
p
o
r
tan
t
t
o
ex
am
in
e
th
e
r
o
le
o
f
d
ata
q
u
ality
an
d
f
ea
tu
r
e
s
elec
tio
n
in
m
o
d
el
p
er
f
o
r
m
a
n
ce
.
T
h
e
ef
f
ec
tiv
e
n
ess
o
f
L
STM
m
o
d
els
ca
n
b
e
s
ig
n
if
ican
tl
y
in
f
lu
en
ce
d
b
y
th
e
ch
o
ice
o
f
in
p
u
t
f
ea
tu
r
es;
th
u
s
,
em
p
lo
y
in
g
m
eth
o
d
s
s
u
ch
as
d
im
en
s
io
n
ality
r
ed
u
ctio
n
o
r
g
eo
m
etr
ical
f
ea
tu
r
e
ex
tr
ac
tio
n
f
r
o
m
s
k
eleta
l d
ata
i
s
ess
en
tial f
o
r
o
p
tim
izin
g
ac
cu
r
ac
y
[
2
1
]
.
Mo
r
eo
v
er
,
th
e
in
teg
r
atio
n
o
f
ad
v
an
ce
d
tech
n
iq
u
es
s
u
ch
as
h
y
b
r
id
m
o
d
els
th
at
co
m
b
in
e
L
STM
n
etwo
r
k
s
with
C
NNs
h
as
s
h
o
wn
p
r
o
m
is
e
in
en
h
an
ci
n
g
ac
tiv
ity
r
ec
o
g
n
itio
n
s
y
s
tem
s
b
y
ca
p
tu
r
in
g
b
o
th
s
p
atial
an
d
tem
p
o
r
al
f
ea
tu
r
es
ef
f
ec
tiv
ely
.
Fo
r
in
s
tan
ce
,
r
ec
en
t
s
tu
d
ies
h
av
e
d
em
o
n
s
tr
ated
th
at
em
p
lo
y
in
g
a
v
is
u
al
g
eo
m
etr
y
g
r
o
u
p
(
VGG
)
-
L
ST
M
m
o
d
el
ca
n
s
ig
n
if
ican
tly
im
p
r
o
v
e
ac
cu
r
ac
y
ac
r
o
s
s
v
ar
i
o
u
s
d
atasets
,
ac
h
iev
in
g
im
p
r
ess
iv
e
r
esu
lts
lik
e
9
5
.
8
5
%
o
n
t
h
e
wir
eless
s
en
s
o
r
d
a
ta
m
in
in
g
(
W
I
SDM
)
d
ataset
[
2
]
.
Fu
r
th
er
m
o
r
e
,
ex
p
lo
r
in
g
m
u
ltimo
d
al
d
ata
s
o
u
r
ce
s
—
s
u
ch
as
in
teg
r
atin
g
au
d
io
o
r
e
n
v
ir
o
n
m
en
tal
s
en
s
o
r
s
—
ca
n
en
r
ich
th
e
co
n
tex
t
s
u
r
r
o
u
n
d
in
g
h
u
m
a
n
a
ctiv
ities
,
lead
in
g
to
a
m
o
r
e
co
m
p
r
eh
en
s
iv
e
u
n
d
er
s
tan
d
in
g
a
n
d
in
ter
p
r
etatio
n
o
f
co
m
p
lex
b
eh
av
io
r
s
.
3
.
2
.
RQ
2
:
t
r
a
ns
f
o
rm
er
s
in m
o
t
io
n:
enha
ncing
a
bn
o
rma
l a
ct
iv
it
y
det
ec
t
io
n
Fo
r
th
is
r
esear
ch
q
u
esti
o
n
,
3
2
p
ap
er
s
wer
e
co
n
s
id
er
e
d
b
y
f
ilter
in
g
th
em
b
ased
o
n
th
e
r
o
le
o
f
T
r
an
s
f
o
r
m
e
r
s
in
e
n
h
an
cin
g
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
ac
tiv
ity
d
etec
tio
n
.
Fin
ally
,
7
p
a
p
er
s
w
er
e
co
n
s
id
er
e
d
f
o
r
an
aly
s
is
.
R
ec
en
t
s
tu
d
ies
in
d
icate
th
at
in
teg
r
atin
g
T
r
an
s
f
o
r
m
er
m
o
d
els
with
s
k
eleto
n
jo
in
t
d
ata
ca
n
y
ield
ev
e
n
h
ig
h
er
ac
cu
r
ac
y
r
ates
th
an
tr
a
d
itio
n
al
m
eth
o
d
s
,
f
u
r
th
er
p
u
s
h
in
g
th
e
b
o
u
n
d
ar
ies
o
f
wh
at
i
s
ac
h
iev
ab
le
in
th
is
d
o
m
ain
[
2
2
]
.
I
n
ad
d
itio
n
to
t
h
e
in
teg
r
atio
n
o
f
ed
g
e
co
m
p
u
tin
g
an
d
h
u
m
an
ex
p
e
r
tis
e,
th
e
r
o
le
o
f
ad
v
a
n
ce
d
an
o
m
aly
d
etec
tio
n
tech
n
iq
u
es
in
ad
d
r
ess
in
g
cy
b
er
s
ec
u
r
ity
th
r
ea
ts
ca
n
n
o
t
b
e
o
v
e
r
s
tated
.
As
d
ig
ital
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
26
:
3
0
2
6
-
3
0
3
5
3030
in
f
r
astru
ctu
r
es
b
ec
o
m
e
in
cr
ea
s
in
g
ly
co
m
p
lex
,
tr
ad
itio
n
al
m
eth
o
d
s
o
f
ten
f
all
s
h
o
r
t
in
id
en
t
if
y
in
g
s
o
p
h
is
ticated
cy
b
er
-
attac
k
s
th
at
ex
p
l
o
it
s
y
s
tem
v
u
ln
e
r
ab
ilit
ies;
h
en
ce
,
DL
ap
p
r
o
ac
h
es
h
av
e
em
er
g
ed
as
a
v
ital
s
o
lu
tio
n
[
2
3
]
.
Fo
r
in
s
tan
ce
,
u
tili
zin
g
m
o
d
els
s
u
ch
as
C
NNs
co
m
b
in
ed
with
L
STM
n
etwo
r
k
s
h
as
s
h
o
wn
r
em
ar
k
ab
le
s
u
cc
ess
in
d
etec
tin
g
n
etwo
r
k
an
o
m
ali
es
b
y
an
aly
zin
g
v
ast
am
o
u
n
ts
o
f
s
eq
u
en
tial
d
ata,
th
u
s
en
h
an
cin
g
b
o
th
ac
c
u
r
ac
y
an
d
ef
f
icien
c
y
in
r
ea
l
-
tim
e
th
r
ea
t id
en
tific
atio
n
[
2
4
]
,
[
2
5
]
.
Fu
r
th
er
m
o
r
e
,
th
ese
m
eth
o
d
o
l
o
g
ies
n
o
t
o
n
l
y
b
o
ls
ter
d
e
f
en
s
es
ag
ain
s
t
im
m
ed
iate
attac
k
s
b
u
t
also
co
n
tr
ib
u
te
to
a
m
o
r
e
p
r
o
f
o
u
n
d
u
n
d
er
s
tan
d
i
n
g
o
f
attac
k
p
atter
n
s
o
v
e
r
tim
e,
allo
win
g
o
r
g
a
n
izatio
n
s
to
p
r
e
-
em
p
tiv
el
y
s
tr
en
g
th
en
th
eir
s
ec
u
r
ity
p
o
s
tu
r
es
an
d
m
itig
at
e
r
is
k
s
ef
f
ec
tiv
ely
.
As
o
r
g
an
iz
atio
n
s
in
cr
ea
s
in
g
ly
r
ely
o
n
s
o
p
h
is
ticated
an
o
m
al
y
d
etec
tio
n
tech
n
iq
u
es,
th
e
in
teg
r
atio
n
o
f
u
n
s
u
p
er
v
is
ed
lear
n
in
g
m
et
h
o
d
s
h
a
s
em
er
g
ed
as
a
v
ital
co
m
p
lem
e
n
t
to
tr
ad
itio
n
al
ap
p
r
o
ac
h
es.
T
h
ese
m
eth
o
d
s
ca
n
au
to
n
o
m
o
u
s
ly
id
en
tify
p
atter
n
s
with
o
u
t
ex
ten
s
iv
e
la
b
eled
d
atasets
,
m
ak
in
g
th
em
p
ar
ticu
la
r
ly
u
s
ef
u
l
i
n
d
y
n
am
ic
e
n
v
ir
o
n
m
en
ts
wh
er
e
n
ew
ty
p
es o
f
an
o
m
alies f
r
eq
u
e
n
tly
ar
is
e.
F
o
r
i
n
s
t
a
n
c
e
,
l
e
v
e
r
a
g
i
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
l
o
n
g
s
h
o
r
t
-
t
e
r
m
m
e
m
o
r
y
(
C
o
n
v
-
L
S
T
M
)
m
o
d
e
l
s
e
n
a
b
l
e
s
s
y
s
t
e
m
s
t
o
l
e
a
r
n
s
p
a
t
i
o
t
e
m
p
o
r
a
l
f
e
a
t
u
r
e
s
f
r
o
m
v
i
d
e
o
d
a
t
a
e
f
f
e
c
t
i
v
e
l
y
,
t
h
e
r
e
b
y
e
n
h
a
n
c
i
n
g
s
u
r
v
e
i
l
l
a
n
c
e
c
a
p
a
b
i
l
i
t
i
e
s
a
n
d
i
m
p
r
o
v
i
n
g
r
e
s
p
o
n
s
e
t
i
m
e
s
i
n
c
r
i
t
i
c
a
l
s
i
t
u
a
t
i
o
n
s
[
2
6
]
.
B
y
e
m
b
r
a
c
i
n
g
s
u
c
h
i
n
n
o
v
a
t
i
v
e
m
e
t
h
o
d
o
l
o
g
i
e
s
,
o
r
g
a
n
i
z
a
t
i
o
n
s
n
o
t
o
n
l
y
f
o
r
t
i
f
y
t
h
e
i
r
d
e
f
e
n
s
e
s
a
g
a
i
n
s
t
e
v
o
l
v
i
n
g
t
h
r
e
a
t
s
b
u
t
a
l
s
o
p
a
v
e
t
h
e
w
a
y
f
o
r
m
o
r
e
a
d
a
p
t
i
v
e
a
n
d
r
e
s
i
l
i
e
n
t
o
p
e
r
a
t
i
o
n
a
l
f
r
a
m
e
w
o
r
k
s
t
h
a
t
c
a
n
r
e
s
p
o
n
d
i
n
t
e
l
l
i
g
e
n
t
l
y
t
o
u
n
f
o
r
e
s
e
e
n
c
h
a
l
l
e
n
g
e
s
.
I
n
a
d
d
i
t
i
o
n
t
o
t
h
e
a
d
v
a
n
c
e
m
e
n
t
s
i
n
h
y
b
r
i
d
m
o
d
e
l
s
,
t
h
e
i
n
t
e
g
r
a
t
i
o
n
o
f
u
n
s
u
p
e
r
v
i
s
e
d
DL
t
e
c
h
n
i
q
u
e
s
h
a
s
e
m
e
r
g
e
d
a
s
a
t
r
a
n
s
f
o
r
m
a
t
i
v
e
f
o
r
c
e
i
n
a
n
o
m
a
l
y
d
e
t
e
c
t
i
o
n
[
2
7
]
,
[
2
8
]
.
T
h
e
s
e
m
e
t
h
o
d
s
a
r
e
p
a
r
t
i
c
u
l
a
r
l
y
a
d
e
p
t
a
t
u
n
c
o
v
e
r
i
n
g
s
u
b
t
l
e
d
e
v
i
a
t
i
o
n
s
f
r
o
m
n
o
r
m
a
l
b
e
h
a
v
i
o
r
p
a
t
t
e
r
n
s
w
i
t
h
o
u
t
r
e
l
y
i
n
g
o
n
e
x
t
e
n
s
i
v
e
l
a
b
e
l
e
d
d
a
t
a
s
e
t
s
,
t
h
e
r
e
b
y
a
d
d
r
e
s
s
i
n
g
o
n
e
o
f
t
h
e
s
i
g
n
i
f
i
c
a
n
t
c
h
a
l
l
e
n
g
e
s
f
a
c
e
d
b
y
t
r
a
d
i
t
i
o
n
a
l
a
p
p
r
o
a
c
h
e
s
.
3
.
3
.
RQ
3
:
L
ST
M
-
ba
s
e
d m
o
dels
:
a
s
up
po
rt
iv
e
f
ra
m
ewo
r
k
f
o
r
s
k
elet
o
n j
o
int
a
na
ly
s
is
Fo
r
th
is
r
esear
ch
q
u
esti
o
n
,
2
7
p
ap
e
r
s
wer
e
co
n
s
id
er
e
d
b
y
f
ilter
in
g
th
em
with
th
e
r
o
le
o
f
L
STM
m
o
d
els
in
an
al
y
zin
g
s
k
eleto
n
jo
in
t
d
ata
d
u
r
in
g
th
e
p
r
o
c
ess
o
f
r
ec
o
g
n
izin
g
a
b
n
o
r
m
al
b
eh
av
io
r
.
At
last
,
7
p
ap
er
s
wer
e
c
o
n
s
id
er
ed
f
o
r
an
aly
s
is
.
As
th
e
f
ield
p
r
o
g
r
ess
es,
in
teg
r
atin
g
th
ese
ad
v
an
ce
d
m
eth
o
d
o
l
o
g
ies
will
b
e
cr
u
cial
f
o
r
d
ev
elo
p
in
g
co
m
p
r
eh
en
s
iv
e
s
o
lu
tio
n
s
th
at
ad
d
r
ess
b
o
th
tech
n
ical
an
d
eth
ical
co
n
s
id
er
atio
n
s
in
h
er
en
t
in
d
e
p
lo
y
in
g
HAR
tech
n
o
lo
g
ies
ef
f
ec
tiv
ely
.
B
y
ca
p
tu
r
in
g
th
e
d
y
n
am
ics
o
f
m
o
tio
n
o
v
e
r
tim
e,
L
STM
-
b
ased
ap
p
r
o
ac
h
es
ca
n
ef
f
ec
tiv
ely
d
is
tin
g
u
is
h
b
etw
ee
n
n
o
r
m
al
a
n
d
ab
n
o
r
m
al
ac
tiv
ities
,
lead
in
g
to
im
p
r
o
v
e
d
d
etec
tio
n
r
ates
in
v
ar
io
u
s
ap
p
licatio
n
s
s
u
ch
as
s
u
r
v
eillan
ce
,
h
ea
lth
ca
r
e,
an
d
h
u
m
an
-
c
o
m
p
u
ter
in
ter
ac
tio
n
.
Fu
r
t
h
er
m
o
r
e,
th
e
in
co
r
p
o
r
atio
n
o
f
atten
tio
n
m
ec
h
an
is
m
s
with
in
th
ese
L
STM
f
r
am
ewo
r
k
s
ca
n
f
u
r
th
er
e
n
h
an
ce
p
e
r
f
o
r
m
an
ce
b
y
allo
win
g
th
e
m
o
d
el
to
f
o
c
u
s
o
n
cr
itical
jo
in
ts
an
d
m
o
v
e
m
en
ts
th
at
ar
e
m
o
s
t
in
d
icativ
e
o
f
a
b
n
o
r
m
al
b
e
h
av
i
o
r
.
A
d
d
itio
n
ally
,
th
e
in
teg
r
atio
n
o
f
m
u
lti
-
m
o
d
al
d
ata
s
o
u
r
ce
s
,
s
u
ch
as
au
d
i
o
a
n
d
en
v
ir
o
n
m
en
tal
co
n
tex
t,
ca
n
p
r
o
v
id
e
a
r
ich
er
u
n
d
e
r
s
tan
d
in
g
o
f
th
e
s
ce
n
a
r
io
s
b
ein
g
an
aly
ze
d
,
en
ab
lin
g
ev
e
n
m
o
r
e
r
o
b
u
s
t
p
r
ed
ictio
n
s
an
d
in
s
ig
h
ts
in
to
h
u
m
an
b
eh
av
i
o
r
.
B
y
lev
er
ag
in
g
th
ese
d
iv
er
s
e
d
ata
s
tr
ea
m
s
,
r
esear
ch
er
s
ca
n
d
e
v
elo
p
m
o
r
e
co
m
p
r
eh
e
n
s
iv
e
m
o
d
els
th
at
ac
co
u
n
t
f
o
r
th
e
c
o
m
p
lex
ities
o
f
h
u
m
an
in
te
r
ac
tio
n
s
an
d
en
v
ir
o
n
m
en
tal
in
f
lu
e
n
ce
s
,
u
ltima
tely
lead
in
g
to
im
p
r
o
v
ed
o
u
tco
m
es
in
v
ar
io
u
s
f
ield
s
,
in
clu
d
in
g
s
ec
u
r
ity
an
d
h
ea
lth
ca
r
e
m
an
a
g
em
en
t
[
2
9
]
.
T
h
e
p
r
i
m
a
r
y
c
h
a
l
l
e
n
g
e
s
i
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
s
k
e
l
e
t
o
n
j
o
i
n
t
s
d
a
t
a
i
n
c
l
u
d
e
t
h
e
d
i
f
f
i
c
u
l
t
y
i
n
c
a
p
t
u
r
i
n
g
d
i
s
c
r
i
m
i
n
a
t
i
v
e
s
p
a
t
i
o
t
e
m
p
o
r
a
l
f
e
a
t
u
r
e
s
,
c
o
m
p
u
t
a
t
i
o
n
a
l
i
n
e
f
f
i
c
i
e
n
c
y
,
a
n
d
t
h
e
h
a
n
d
l
i
n
g
o
f
o
c
c
l
u
s
i
o
n
s
a
n
d
n
o
i
s
e
.
T
r
a
d
i
t
i
o
n
a
l
m
e
t
h
o
d
s
o
f
t
e
n
s
t
r
u
g
g
l
e
w
i
t
h
r
e
c
o
g
n
i
z
i
n
g
a
c
t
i
o
n
s
t
h
a
t
h
a
v
e
s
u
b
t
l
e
d
i
f
f
e
r
e
n
c
e
s
i
n
m
o
t
i
o
n
p
a
t
t
e
r
n
s
d
u
e
t
o
t
h
e
i
r
r
e
l
i
a
n
c
e
o
n
n
a
i
v
e
a
t
t
e
n
t
i
o
n
m
e
c
h
a
n
i
s
m
s
,
w
h
i
c
h
a
r
e
n
o
t
a
d
e
p
t
a
t
d
i
s
t
i
n
g
u
i
s
h
i
n
g
s
u
c
h
n
u
a
n
c
e
s
[
3
0
]
.
A
d
d
i
t
i
o
n
a
l
l
y
,
t
h
e
c
o
m
p
l
e
x
i
t
y
o
f
t
r
a
n
s
f
o
r
m
e
r
m
o
d
e
l
s
c
a
n
l
e
a
d
t
o
h
i
g
h
c
o
m
p
u
t
a
t
i
o
n
a
l
c
o
s
t
s
a
n
d
r
e
s
o
u
r
c
e
d
e
m
a
n
d
s
,
w
h
i
c
h
a
r
e
s
i
g
n
i
f
i
c
a
n
t
b
a
r
r
i
e
r
s
t
o
t
h
e
i
r
p
r
a
c
t
i
c
a
l
a
p
p
l
i
c
a
t
i
o
n
[
3
1
]
.
A
n
o
t
h
e
r
c
h
a
l
l
e
n
g
e
i
s
t
h
e
l
i
m
i
t
e
d
r
e
c
e
p
t
i
v
e
f
i
e
l
d
o
f
g
r
a
p
h
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
t
w
o
r
k
s
(
G
C
N
s
)
,
w
h
i
c
h
r
e
s
t
r
i
c
t
s
t
h
e
i
r
a
b
i
l
i
t
y
t
o
c
a
p
t
u
r
e
l
o
n
g
-
r
a
n
g
e
d
e
p
e
n
d
e
n
c
i
e
s
b
e
t
w
e
e
n
j
o
i
n
t
s
a
c
r
o
s
s
f
r
a
m
e
s
[
4
]
.
M
o
r
eo
v
er
,
t
h
e
a
g
g
r
eg
a
t
i
o
n
o
f
jo
i
n
t
s
i
n
to
b
o
d
y
p
ar
t
s
c
an
o
b
s
c
u
r
e
f
in
e
-
g
r
a
in
ed
a
c
t
i
o
n
s
,
m
ak
i
n
g
i
t
d
i
f
f
ic
u
l
t
t
o
f
u
l
ly
e
x
p
l
o
i
t
p
a
r
t
-
l
e
v
e
l
d
a
ta
.
T
h
e
I
I
P
-
t
r
an
s
f
o
r
m
e
r
l
e
v
er
a
g
e
s
i
n
t
r
a
-
a
n
d
i
n
t
er
-
p
a
r
t
in
t
e
r
a
c
t
io
n
s
to
i
m
p
r
o
v
e
r
e
co
g
n
i
t
i
o
n
o
f
b
o
th
c
o
a
r
s
e
-
a
n
d
f
in
e
-
g
r
a
i
n
ed
a
c
t
io
n
s
[
5
]
.
A
d
d
i
t
i
o
n
a
l
l
y
,
t
h
e
l
a
n
g
u
ag
e
-
g
u
id
e
d
g
r
a
p
h
t
r
a
n
s
f
o
r
m
e
r
(
L
G
G
T
)
in
c
o
r
p
o
r
a
t
e
s
s
e
m
an
t
i
c
g
u
i
d
an
c
e
to
e
n
h
an
c
e
t
h
e
m
o
d
e
l
's
u
n
d
er
s
t
a
n
d
i
n
g
o
f
a
c
t
i
o
n
r
e
l
a
t
i
o
n
s
h
ip
s
[
6
]
.
T
h
e
s
e
in
n
o
v
a
t
i
o
n
s
d
em
o
n
s
t
r
a
t
e
th
e
p
o
t
en
t
i
a
l
o
f
t
r
an
s
f
o
r
m
e
r
s
t
o
o
v
e
r
c
o
m
e
th
e
in
h
er
e
n
t
c
h
a
l
l
en
g
e
s
in
s
k
e
l
e
to
n
-
b
a
s
ed
a
c
t
i
o
n
r
e
co
g
n
i
t
i
o
n
,
o
f
f
er
i
n
g
m
o
r
e
r
o
b
u
s
t
,
e
f
f
ic
i
e
n
t,
an
d
ac
c
u
r
a
te
m
o
d
e
l
s
f
o
r
r
e
a
l
-
wo
r
ld
ap
p
l
i
c
a
t
io
n
s
.
I
n
s
k
e
le
t
o
n
-
b
a
s
e
d
H
A
R
s
y
s
te
m
s
,
i
n
t
er
p
r
e
t
ab
i
l
i
t
y
i
s
a
l
s
o
c
r
u
c
ia
l
.
A
t
t
en
t
i
o
n
w
e
i
g
h
t
s
,
k
e
y
jo
i
n
t
m
o
v
e
m
en
t
s
,
m
o
t
io
n
tr
a
j
e
c
to
r
i
e
s
,
a
n
d
p
a
t
t
er
n
s
a
t
t
h
e
f
r
am
e
l
e
v
e
l
f
o
c
u
s
o
n
h
o
w
L
S
T
M
a
n
d
t
r
a
n
s
f
o
r
m
e
r
m
o
d
e
l
s
c
l
a
s
s
i
f
y
a
n
a
c
t
iv
i
t
y
.
T
h
e
s
e
e
x
p
l
a
i
n
a
b
l
e
f
e
a
t
u
r
e
s
ca
n
b
e
h
e
l
p
f
u
l
i
n
s
en
s
i
t
i
v
e
ap
p
l
i
c
a
t
io
n
s
a
s
t
h
ey
p
r
o
v
i
d
e
i
n
f
o
r
m
a
t
io
n
o
n
w
h
ic
h
b
o
d
y
j
o
in
t
s
o
r
t
i
m
e
s
t
e
p
s
m
o
s
t
i
n
f
lu
e
n
ce
d
th
e
f
i
n
a
l
c
l
a
s
s
i
f
ic
a
t
i
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Lo
n
g
s
h
o
r
t
-
term me
mo
r
y
b
a
s
ed
a
ctivity
d
etec
tio
n
u
s
in
g
s
ke
leto
n
…
(
B
a
kk
a
la
S
a
n
th
a
K
u
ma
r
)
3031
3
.
4
.
RQ
4
:
f
o
cus
ing
o
n sk
elet
o
n j
o
ints a
nd
t
ra
ns
f
o
rm
er
s
f
o
r
s
up
er
io
r
a
ct
iv
it
y
re
co
g
nitio
n
F
o
r
th
i
s
r
e
s
e
a
r
c
h
q
u
e
s
t
i
o
n
,
2
3
p
ap
e
r
s
w
e
r
e
co
n
s
id
e
r
ed
b
y
f
i
l
t
er
i
n
g
th
e
m
w
i
th
th
e
c
o
n
ce
p
t
o
f
i
n
t
e
g
r
a
t
i
n
g
s
k
e
l
e
t
o
n
jo
i
n
t
s
d
a
t
a
a
n
d
T
r
an
s
f
o
r
m
e
r
s
f
o
r
b
e
tt
e
r
a
c
t
iv
i
t
y
r
e
co
g
n
i
t
i
o
n
.
F
i
n
al
l
y
,
8
p
a
p
er
s
w
e
r
e
c
o
n
s
i
d
e
r
ed
f
o
r
an
a
ly
s
i
s
.
A
s
t
h
e
i
n
t
e
g
r
a
t
i
o
n
o
f
s
k
e
le
t
o
n
j
o
in
t
d
at
a
w
i
t
h
a
d
v
a
n
c
e
d
a
r
c
h
i
te
c
t
u
r
e
s
l
i
k
e
t
r
a
n
s
f
o
r
m
e
r
s
an
d
L
S
T
M
s
g
a
in
s
t
r
a
c
t
io
n
,
a
t
t
en
t
i
o
n
m
u
s
t
a
l
s
o
b
e
d
ir
e
c
t
ed
t
o
wa
r
d
e
n
h
an
c
in
g
u
s
e
r
in
t
e
r
a
c
t
i
o
n
w
i
t
h
t
h
e
s
e
s
y
s
t
e
m
s
.
Fo
r
ex
am
p
l
e
,
d
ev
e
l
o
p
i
n
g
in
t
u
i
t
iv
e
in
t
e
r
f
a
c
e
s
t
h
a
t
a
l
lo
w
u
s
er
s
t
o
p
r
o
v
id
e
r
e
a
l
-
t
im
e
f
e
e
d
b
a
ck
c
a
n
s
i
g
n
i
f
i
ca
n
t
ly
i
m
p
r
o
v
e
m
o
d
e
l
t
r
a
in
i
n
g
a
n
d
a
d
ap
t
a
b
i
l
i
ty
,
f
o
s
t
e
r
in
g
a
m
o
r
e
p
e
r
s
o
n
a
l
i
ze
d
e
x
p
er
i
e
n
c
e
.
T
h
e
m
u
l
t
i
-
s
c
a
le
te
m
p
o
r
a
l
tr
a
n
s
f
o
r
m
e
r
(
M
T
T
)
a
l
s
o
a
d
d
r
e
s
s
e
s
t
h
e
n
e
ed
f
o
r
ca
p
t
u
r
i
n
g
l
o
n
g
-
t
e
r
m
t
e
m
p
o
r
a
l
d
e
p
en
d
en
c
i
e
s
,
wh
i
ch
ar
e
c
r
u
c
i
a
l
f
o
r
s
eq
u
en
t
i
a
l
s
k
e
l
e
t
o
n
d
a
ta
,
b
y
in
t
eg
r
a
t
in
g
m
u
l
t
i
-
s
c
a
le
te
m
p
o
r
a
l
e
m
b
ed
d
in
g
s
an
d
t
r
a
n
s
f
o
r
m
e
r
e
n
c
o
d
e
r
s
[
2
4
]
.
T
h
e
c
l
o
u
d
en
v
i
r
o
n
m
e
n
t
o
f
f
e
r
s
a
s
c
a
l
a
b
l
e
a
n
d
e
f
f
i
c
i
e
n
t
p
l
a
tf
o
r
m
f
o
r
d
e
p
lo
y
in
g
th
e
s
e
ad
v
an
c
ed
m
o
d
el
s
,
a
l
l
o
w
in
g
f
o
r
r
e
a
l
-
t
im
e
p
r
o
c
e
s
s
i
n
g
a
n
d
an
a
l
y
s
i
s
o
f
l
a
r
g
e
d
a
ta
s
e
t
s
.
T
h
e
i
n
t
e
g
r
a
t
i
o
n
o
f
L
S
T
M
an
d
t
r
an
s
f
o
r
m
e
r
s
in
t
h
e
c
lo
u
d
c
an
l
ev
e
r
a
g
e
t
h
e
s
t
r
en
g
th
s
o
f
b
o
th
m
o
d
e
l
s
,
w
i
t
h
L
S
T
M
h
a
n
d
l
i
n
g
t
e
m
p
o
r
a
l
s
e
q
u
en
c
es
a
n
d
t
r
a
n
s
f
o
r
m
e
r
s
c
ap
t
u
r
i
n
g
c
o
m
p
l
ex
d
e
p
en
d
en
c
i
e
s
,
th
u
s
e
n
h
a
n
c
in
g
th
e
o
v
er
a
l
l
a
cc
u
r
a
c
y
a
n
d
ef
f
i
c
ie
n
c
y
o
f
H
A
R
s
y
s
t
em
s
.
M
o
r
eo
v
er
,
t
h
e
u
s
e
o
f
DL
t
ec
h
n
iq
u
e
s
,
s
u
ch
a
s
t
h
o
s
e
c
o
m
b
i
n
in
g
C
N
N
an
d
L
S
T
M
,
h
a
s
s
h
o
w
n
p
r
o
m
i
s
e
i
n
e
x
t
r
a
c
t
in
g
s
p
a
t
i
o
te
m
p
o
r
a
l
f
e
a
t
u
r
e
s
f
r
o
m
s
k
e
l
e
to
n
d
at
a
,
f
u
r
th
e
r
i
m
p
r
o
v
in
g
r
ec
o
g
n
i
t
i
o
n
a
c
cu
r
a
cy
[
3
2
]
.
T
h
e
d
e
v
e
l
o
p
m
en
t
o
f
i
n
t
e
l
l
i
g
en
t
H
A
R
s
y
s
t
e
m
s
u
s
i
n
g
s
k
e
l
e
to
n
i
n
f
o
r
m
a
t
io
n
,
as
d
i
s
c
u
s
s
e
d
[
3
3
]
,
u
n
d
e
r
s
c
o
r
e
s
t
h
e
p
o
t
en
t
i
a
l
o
f
t
h
e
s
e
t
e
c
h
n
o
lo
g
i
e
s
in
v
ar
i
o
u
s
a
p
p
l
i
ca
t
i
o
n
s
,
f
r
o
m
m
o
n
i
to
r
i
n
g
t
h
e
e
l
d
er
l
y
to
s
u
r
v
e
i
l
la
n
c
e.
T
h
e
c
lo
u
d
e
n
v
i
r
o
n
m
e
n
t
n
o
t
o
n
ly
f
ac
i
l
i
t
a
t
e
s
t
h
e
d
e
p
lo
y
m
e
n
t
o
f
t
h
e
s
e
m
o
d
e
ls
b
u
t
a
l
s
o
s
u
p
p
o
r
t
s
t
h
e
in
t
e
g
r
a
t
i
o
n
o
f
m
u
l
t
i
p
l
e
d
at
a
s
o
u
r
c
e
s
,
s
u
ch
a
s
i
m
ag
e
f
e
at
u
r
e
e
x
tr
a
c
t
io
n
a
n
d
o
b
j
ec
t
d
e
te
c
t
i
o
n
,
to
im
p
r
o
v
e
a
c
t
i
o
n
r
e
co
g
n
i
t
i
o
n
[
1
0
]
.
T
h
e
s
p
a
t
i
o
-
te
m
p
o
r
a
l
i
m
ag
e
f
o
r
m
at
i
o
n
t
ec
h
n
i
q
u
e
p
r
o
p
o
s
ed
b
y
T
a
s
n
i
m
e
t
a
l
.
[
1
1
]
a
l
s
o
h
i
g
h
l
i
g
h
t
s
t
h
e
p
o
te
n
t
i
a
l
o
f
u
s
i
n
g
p
r
e
-
tr
a
i
n
ed
m
o
d
e
l
s
an
d
f
u
s
io
n
t
e
ch
n
iq
u
e
s
t
o
en
h
an
ce
t
h
e
p
e
r
f
o
r
m
a
n
ce
o
f
H
A
R
s
y
s
t
e
m
s
i
n
th
e
c
lo
u
d
.
F
i
n
a
l
ly
,
a
d
d
r
e
s
s
i
n
g
c
h
a
ll
en
g
e
s
s
u
ch
a
s
m
i
s
s
i
n
g
j
o
in
t
r
ec
o
v
er
y
,
a
s
e
x
p
l
o
r
e
d
b
y
Z
h
a
l
g
a
s
b
a
y
ev
a
n
d
T
u
[
3
4
]
,
i
s
c
r
u
c
i
a
l
f
o
r
i
m
p
r
o
v
in
g
t
h
e
r
o
b
u
s
t
n
e
s
s
o
f
s
k
e
l
et
o
n
-
b
a
s
e
d
ac
t
i
v
i
ty
r
e
c
o
g
n
i
t
io
n
s
y
s
t
em
s
i
n
d
y
n
a
m
i
c
a
n
d
u
n
ce
r
t
a
in
e
n
v
i
r
o
n
m
e
n
ts
[
3
5
]
,
[
3
6
]
.
O
v
e
r
a
l
l,
t
h
e
co
m
b
i
n
a
t
io
n
o
f
L
S
T
M,
t
r
a
n
s
f
o
r
m
e
r
s
,
a
n
d
c
lo
u
d
co
m
p
u
t
in
g
p
r
e
s
e
n
t
s
a
p
o
w
e
r
f
u
l
ap
p
r
o
a
ch
to
ad
v
an
c
in
g
t
h
e
f
ie
l
d
o
f
HAR
u
s
i
n
g
s
k
e
l
e
t
o
n
j
o
in
t
s
.
L
a
t
e
n
cy
,
p
r
iv
a
c
y
,
b
a
n
d
w
i
d
th
,
a
n
d
c
o
m
p
u
t
a
t
i
o
n
a
l
co
s
t
a
r
e
c
r
i
t
i
c
a
l
ch
a
l
l
en
g
e
s
f
o
r
c
l
o
u
d
-
e
n
ab
l
e
d
H
A
R
s
y
s
t
e
m
s
f
o
r
r
e
a
l
-
wo
r
ld
d
e
p
lo
y
m
e
n
t
.
A
f
e
a
s
i
b
le
s
o
lu
t
i
o
n
i
s
t
o
d
o
t
h
e
s
k
e
l
e
t
o
n
ex
t
r
a
c
t
io
n
an
d
b
a
s
i
c
p
r
o
c
e
s
s
i
n
g
o
n
e
d
g
e
d
ev
i
c
e
s
,
a
n
d
t
h
e
l
ar
g
e
-
s
c
a
l
e
s
t
o
r
a
g
e
an
d
c
o
m
p
l
ex
L
S
T
M
-
t
r
a
n
s
f
o
r
m
e
r
in
f
er
e
n
c
e
o
n
t
h
e
c
l
o
u
d
s
e
r
v
e
r
.
T
h
e
in
t
eg
r
a
t
io
n
o
f
ed
g
e
an
d
c
l
o
u
d
c
an
h
e
l
p
t
o
c
u
t
l
a
te
n
cy
a
n
d
i
n
c
r
e
a
s
e
s
c
a
l
a
b
i
l
i
t
y
.
B
u
t
,
in
c
e
r
t
a
i
n
a
p
p
l
i
c
a
t
io
n
s
l
ik
e
h
e
a
lt
h
c
a
r
e,
s
u
r
v
e
i
l
la
n
c
e,
an
d
e
l
d
e
r
l
y
m
o
n
i
t
o
r
i
n
g
,
p
r
iv
a
cy
-
p
r
es
e
r
v
i
n
g
t
ec
h
n
i
q
u
e
s
a
r
e
r
eq
u
ir
e
d
f
o
r
u
s
i
n
g
H
A
R
s
y
s
t
e
m
.
T
h
e
r
e
s
u
l
t
s
o
f
t
h
i
s
r
e
v
ie
w
a
r
e
a
p
p
l
i
c
ab
l
e
t
o
a
f
e
w
p
r
a
c
t
i
c
a
l
a
p
p
l
i
c
a
t
io
n
s
o
f
H
A
R
s
u
ch
a
s
f
a
l
l
d
e
te
c
t
i
o
n
f
o
r
th
e
e
ld
e
r
ly
,
s
m
a
r
t
s
u
r
v
ei
l
l
a
n
c
e
,
r
eh
a
b
i
l
i
ta
t
i
o
n
,
s
p
o
r
t
s
a
n
a
l
y
s
i
s
,
H
R
I
,
an
d
h
e
a
l
t
h
c
a
r
e
m
o
n
i
to
r
in
g
.
T
r
a
n
s
f
e
r
l
e
a
r
n
i
n
g
,
d
o
m
a
in
ad
ap
t
a
ti
o
n
,
l
i
g
h
t
w
e
ig
h
t
e
d
g
e
p
r
o
c
e
s
s
i
n
g
,
a
n
d
s
c
a
la
b
l
e
i
n
f
er
e
n
ce
t
o
th
e
c
l
o
u
d
a
r
e
a
l
l
p
o
s
s
i
b
l
e
t
ec
h
n
i
q
u
e
s
t
o
ad
a
p
t
th
e
m
o
d
e
l
s
t
o
v
ar
i
o
u
s
e
n
v
i
r
o
n
m
e
n
t
s
.
T
h
e
m
o
d
e
l
s
c
a
n
b
e
t
r
an
s
f
e
r
r
ed
to
d
i
f
f
e
r
en
t
e
n
v
i
r
o
n
m
e
n
t
s
w
i
th
tr
an
s
f
e
r
l
e
a
r
n
in
g
,
ad
a
p
t
e
d
t
o
v
ar
i
o
u
s
e
n
v
ir
o
n
m
en
t
s
w
i
t
h
d
o
m
a
i
n
a
d
ap
t
a
t
i
o
n
,
p
r
o
c
e
s
s
e
d
a
t
t
h
e
ed
g
e
w
i
t
h
l
ig
h
t
we
i
g
h
t
m
o
d
e
l
s
,
an
d
s
c
al
e
d
t
o
th
e
c
l
o
u
d
f
o
r
i
n
f
e
r
e
n
ce
.
4.
CO
M
P
ARA
T
I
V
E
ANA
L
YS
I
S O
F
E
XI
ST
I
NG
WO
RK
S
T
h
e
f
ield
o
f
HAR
h
as
s
ig
n
if
ican
tly
ad
v
an
ce
d
with
th
e
e
m
er
g
en
ce
o
f
DL
an
d
m
ac
h
in
e
lear
n
in
g
m
o
d
els
[
3
7
]
,
[
3
8
]
,
p
ar
tic
u
lar
ly
in
en
v
i
r
o
n
m
e
n
ts
th
at
u
tili
ze
clo
u
d
c
o
m
p
u
tin
g
.
T
h
is
tech
n
o
lo
g
y
en
ab
les
th
e
p
r
ed
ictio
n
an
d
class
if
icatio
n
o
f
h
u
m
an
ac
tiv
ities
b
ased
o
n
t
h
e
co
llected
s
en
s
o
r
d
ata
[
3
9
]
.
Du
e
to
th
e
d
iv
er
s
e
co
m
p
u
tin
g
en
v
ir
o
n
m
en
ts
an
d
m
eth
o
d
s
,
th
e
d
e
v
elo
p
m
e
n
t
o
f
v
ar
io
u
s
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
an
d
m
o
d
els
h
as
b
ee
n
ca
r
r
ie
d
o
u
t.
So
m
e
o
f
t
h
ese
in
clu
d
e
th
e
u
s
e
o
f
t
r
an
s
f
o
r
m
er
n
etwo
r
k
s
an
d
l
o
n
g
-
t
er
m
m
em
o
r
y
.
T
h
is
s
u
b
s
ec
tio
n
aim
s
to
r
e
v
iew
th
e
v
ar
io
u
s
im
p
lem
en
tatio
n
s
o
f
HAR
u
s
in
g
t
r
an
s
f
o
r
m
er
n
etw
o
r
k
s
,
L
STM
m
o
d
els,
an
d
o
th
er
s
in
clo
u
d
en
v
ir
o
n
m
e
n
ts
.
T
h
er
e
was
n
o
f
o
r
m
al
m
eta
-
an
aly
s
is
co
n
d
u
cted
d
u
e
to
th
e
d
if
f
er
e
n
ce
s
in
th
e
d
ata
s
ets,
d
ata
f
o
r
m
at,
ev
alu
atio
n
m
etr
ics
,
an
d
ex
p
e
r
im
en
tal
co
n
d
itio
n
s
u
s
ed
ac
r
o
s
s
th
e
s
tu
d
ies.
T
ab
le
2
s
u
m
m
ar
izes
th
e
ex
is
tin
g
HA
R
s
tu
d
ies
b
ased
o
n
m
eth
o
d
,
d
ataset,
p
er
f
o
r
m
a
n
ce
,
a
n
d
k
ey
in
f
lu
en
cin
g
f
ac
to
r
s
.
T
h
e
co
m
p
a
r
is
o
n
allo
ws th
e
ev
alu
atio
n
o
f
t
h
e
r
elativ
e
p
er
f
o
r
m
an
ce
o
f
th
e
L
STM
,
C
NN
-
L
STM
,
g
ated
r
ec
u
r
r
en
t
u
n
it (
GR
U
)
,
t
r
an
s
f
o
r
m
e
r
,
GC
N,
an
d
h
y
b
r
id
m
o
d
els o
n
v
ar
i
o
u
s
HAR d
ataset
s
.
I
n
ad
d
itio
n
to
L
STM
m
o
d
els
an
d
t
r
an
s
f
o
r
m
e
r
m
o
d
els,
o
th
er
m
et
h
o
d
s
,
s
u
ch
as
GC
Ns
an
d
m
u
ltimo
d
al
f
u
s
io
n
in
s
k
eleto
n
-
b
ased
HAR
,
ar
e
also
s
ig
n
if
ican
t.
T
h
e
ad
v
an
ta
g
e
o
f
GC
N
-
b
ased
m
o
d
els
is
th
at
th
ey
ca
n
lear
n
s
p
atial
r
elatio
n
s
h
ip
s
b
etwe
en
b
o
d
y
jo
in
ts
,
an
d
L
STM
m
o
d
els
an
d
t
r
an
s
f
o
r
m
er
m
o
d
els
ar
e
p
r
im
ar
ily
u
s
ed
to
m
o
d
el
tem
p
o
r
al
an
d
lo
n
g
-
r
an
g
e
m
o
v
em
e
n
ts
.
B
y
4
co
m
b
in
in
g
s
k
eleto
n
d
ata
an
d
R
GB
v
id
eo
,
wea
r
ab
le
s
en
s
o
r
s
,
au
d
io
,
o
r
en
v
ir
o
n
m
en
tal
in
f
o
r
m
atio
n
,
th
e
p
er
f
o
r
m
an
ce
ca
n
f
u
r
th
er
b
e
im
p
r
o
v
ed
b
y
m
u
ltimo
d
al
f
u
s
io
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
26
:
3
0
2
6
-
3
0
3
5
3032
T
ab
le
2
.
C
o
m
p
a
r
ativ
e
an
aly
s
is
with
th
e
s
tate
-
of
-
th
e
-
a
r
t w
o
r
k
s
R
e
f
.
P
r
o
b
l
e
m
st
a
t
e
me
n
t
M
e
t
h
o
d
s
p
r
o
p
o
s
e
d
D
a
t
a
s
e
t
u
s
e
d
R
e
s
u
l
t
s
o
b
t
a
i
n
e
d
I
n
f
l
u
e
n
c
i
n
g
f
a
c
t
o
r
s
[
8
]
H
A
R
u
s
i
n
g
e
d
g
e
-
c
o
m
p
u
t
i
n
g
OI
-
El
e
A
t
t
G
-
G
R
U
U
C
I
mac
h
i
n
e
l
e
a
r
n
i
n
g
r
e
p
o
si
t
o
r
y
,
U
C
I
H
A
R
,
a
n
d
s
e
l
f
-
c
r
e
a
t
e
d
8
5
.
4
%
a
v
e
r
a
g
e
a
c
c
u
r
a
c
y
F
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
[
9
]
H
A
R
f
o
r
p
e
d
e
s
t
r
i
a
n
n
a
v
i
g
a
t
i
o
n
D
L
a
l
g
o
r
i
t
h
ms
w
i
t
h
sen
s
o
r
d
a
t
a
,
LST
M
,
a
n
d
C
N
N
U
C
I
U
p
t
o
9
1
.
9
2
%
a
c
c
u
r
a
c
y
a
n
d
9
5
.
5
5
%
o
n
smar
t
p
h
o
n
e
p
o
st
u
r
e
r
e
c
o
g
n
i
t
i
o
n
S
e
v
e
n
t
r
a
d
i
t
i
o
n
a
l
M
L
a
l
g
o
r
i
t
h
ms
[
1
2
]
H
A
R
u
s
i
n
g
mo
b
i
l
e
h
e
a
l
t
h
d
a
t
a
C
N
N
w
i
t
h
LST
M
a
n
d
S
V
M
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
M
H
EA
LTH
9
9
.
8
%
i
mp
r
e
ssi
v
e
a
c
c
u
r
a
c
y
Tr
a
n
sf
e
r
l
e
a
r
n
i
n
g
[
4
0
]
HAR
A
G
R
U
a
l
g
o
r
i
t
h
m
W
I
S
D
M
9
7
.
0
8
%
t
e
st
i
n
g
a
c
c
u
r
a
c
y
a
n
d
0
.
2
2
1
t
e
s
t
i
n
g
l
o
ss
H
y
p
e
r
-
p
a
r
a
met
e
r
t
u
n
i
n
g
a
n
d
k
-
f
o
l
d
c
r
o
ss
-
v
a
l
i
d
a
t
i
o
n
[
4
1
]
P
e
r
f
o
r
ma
n
c
e
e
v
a
l
u
a
t
i
o
n
w
.
r
.
t
H
A
R
F
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
u
si
n
g
3
D
-
D
C
N
N
u
si
n
g
LST
M
.
U
C
F
5
0
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
8
6
.
5
7
%
a
c
c
u
r
a
c
y
3
D
C
N
N
-
LST
M
a
p
p
r
o
a
c
h
[
4
2
]
H
A
R
u
s
i
n
g
DL
A
r
e
si
d
u
a
l
B
i
LST
M
n
e
t
w
o
r
k
U
C
F
-
C
r
i
me
K
a
g
g
l
e
1
0
0
%
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
c
c
u
r
a
c
y
C
o
m
b
i
n
i
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
a
n
d
r
e
c
u
r
r
e
n
t
l
a
y
e
r
s
[
4
3
]
S
e
n
s
o
r
-
b
a
s
e
d
HAR
A
mo
d
i
f
i
e
d
B
R
N
N
w
i
t
h
LST
M
M
o
b
i
f
a
l
l
4
%
i
m
p
r
o
v
e
m
e
n
t
o
v
e
r
p
r
e
v
i
o
u
s
me
t
h
o
d
s
F
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
[
4
4
]
H
A
R
u
s
i
n
g
h
y
b
r
i
d
D
L
H
y
b
r
i
d
D
L
c
o
m
b
i
n
e
d
w
i
t
h
C
N
N
a
n
d
LSTM
2
0
p
a
r
t
i
c
i
p
a
n
t
s
d
a
t
a
se
t
w
i
t
h
1
2
d
i
f
f
e
r
e
n
t
p
h
y
s
i
c
a
l
a
c
t
i
v
i
t
i
e
s
9
0
.
8
9
%
a
c
c
u
r
a
c
y
C
o
m
b
i
n
a
t
i
o
n
o
f
C
N
N
a
n
d
LS
TM
[
4
5
]
D
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
(DNN)
-
b
a
sed
HAR
LSTM
c
o
mb
i
n
e
d
w
i
t
h
R
N
N
W
I
S
D
M
9
4
%
a
c
c
u
r
a
c
y
w
i
t
h
3
0
%
l
o
ss
L2
r
e
g
u
l
a
r
i
z
a
t
i
o
n
a
n
d
d
r
o
p
o
u
t
.
[
4
6
]
H
A
R
u
s
i
n
g
sen
s
o
r
d
a
t
a
B
i
LST
M
,
C
N
N
,
a
n
d
C
o
n
v
LST
M
W
I
S
D
M
8
4
.
8
%
w
i
t
h
B
i
LST
M
a
n
d
8
4
.
3
%
w
i
t
h
C
o
n
v
LST
M
.
W
a
t
c
h
a
c
c
e
l
e
r
o
m
e
t
e
r
d
a
t
a
[
4
7
]
H
A
R
u
s
i
n
g
DNN
A
h
y
b
r
i
d
mo
d
e
l
c
o
m
b
i
n
i
n
g
C
N
N
a
n
d
LSTM
a
l
g
o
r
i
t
h
ms
U
C
I
-
HAR
9
6
.
8
%
a
c
c
u
r
a
c
y
.
P
r
i
n
c
i
p
a
l
c
o
mp
o
n
e
n
t
a
n
a
l
y
si
s
(
P
C
A
)
a
n
d
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
[
4
8
]
H
A
R
u
s
i
n
g
LSTM
a
n
d
R
N
N
LSTM
-
R
N
N
D
L
O
P
P
O
R
TU
N
I
TY
9
4
%
o
v
e
r
a
l
l
a
c
c
u
r
a
c
y
.
9
1
%
a
c
c
u
r
a
c
y
o
n
n
o
n
-
r
e
p
e
t
i
t
i
v
e
mo
t
i
o
n
s a
n
d
8
0
%
a
c
c
u
r
a
c
y
o
n
r
e
p
e
t
i
t
i
v
e
o
n
e
s
C
o
m
p
a
r
a
t
i
v
e
a
n
a
l
y
si
s
a
n
d
p
e
r
f
o
r
m
a
n
c
e
met
r
i
c
s
[
4
9
]
H
A
R
u
s
i
n
g
w
e
a
r
a
b
l
e
sen
s
o
r
d
a
t
a
C
N
N
w
i
t
h
G
R
U
W
I
S
D
M
smar
t
p
h
o
n
e
a
n
d
sm
a
r
t
w
a
t
c
h
a
c
t
i
v
i
t
y
a
n
d
b
i
o
me
t
r
i
c
s
Th
e
C
N
N
-
G
R
U
m
o
d
e
l
a
c
h
i
e
v
e
d
t
h
e
h
i
g
h
e
s
t
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
c
c
u
r
a
c
y
Le
v
e
r
a
g
i
n
g
sp
a
t
i
a
l
a
n
d
t
e
m
p
o
r
a
l
f
e
a
t
u
r
e
s
[
5
0
]
H
A
R
u
s
i
n
g
sel
f
-
a
t
t
e
n
t
i
o
n
mo
d
e
l
C
N
N
w
i
t
h
LST
M
H
-
a
c
t
i
v
i
t
y
,
M
H
EA
L
TH
,
a
n
d
U
C
I
-
HAR
9
9
.
9
3
%
a
c
c
u
r
a
c
y
.
A
sel
f
-
a
t
t
e
n
t
i
o
n
l
a
y
e
r
t
o
e
x
t
r
a
c
t
t
e
m
p
o
r
a
l
f
e
a
t
u
r
e
s
[
5
1
]
H
A
R
b
a
se
d
o
n
C
N
N
2D
-
C
N
N
W
I
S
D
M
a
c
h
i
e
v
i
n
g
a
n
a
c
c
u
r
a
c
y
o
f
9
7
.
2
0
%
o
n
t
h
e
W
I
S
D
M
d
a
t
a
se
t
2D
-
C
N
N
w
i
t
h
h
y
p
e
r
p
a
r
a
me
t
e
r
t
u
n
i
n
g
[
5
2
]
H
A
R
u
s
i
n
g
t
r
a
n
s
i
t
i
o
n
a
l
a
c
t
i
v
i
t
i
e
s
LSTM
w
i
t
h
B
i
LST
M
a
n
d
C
N
N
D
a
t
a
s
e
t
A
(
H
A
P
T)
a
n
d
d
a
t
a
se
t
B
9
6
.
1
%
a
c
c
u
r
a
c
y
f
o
r
t
r
a
n
s
i
t
i
o
n
a
l
a
c
t
i
v
i
t
i
e
s a
n
d
9
8
.
4
%
f
o
r
b
a
si
c
a
c
t
i
v
i
t
i
e
s
H
y
b
r
i
d
mo
d
e
l
c
o
m
b
i
n
e
d
w
i
t
h
d
e
c
i
si
o
n
f
u
si
o
n
[
5
3
]
H
A
R
u
s
i
n
g
n
o
i
s
y
d
a
t
a
C
N
N
w
i
t
h
B
i
LST
M
p
a
r
a
l
l
e
l
mo
d
e
l
u
si
n
g
a
t
t
e
n
t
i
o
n
m
e
c
h
a
n
i
sm
U
C
I
H
A
R
d
a
t
a
se
t
a
n
d
t
h
e
W
I
S
D
M
H
A
R
9
6
.
7
1
%
a
c
c
u
r
a
c
y
D
i
me
n
si
o
n
a
l
i
t
y
r
e
d
u
c
t
i
o
n
a
n
d
h
y
p
e
r
p
a
r
a
me
t
e
r
t
u
n
i
n
g
[
5
4
]
H
A
R
u
s
i
n
g
DL
C
N
N
w
i
t
h
LST
M
U
C
I
-
H
A
R
,
O
P
P
O
R
TU
N
I
TY
,
U
N
I
M
I
B
-
S
H
A
R
,
P
A
M
A
P
2
,
a
n
d
W
I
S
D
M
9
8
.
8
2
%
i
m
p
r
e
ss
i
v
e
a
c
c
u
r
a
c
y
o
n
s
p
e
c
i
f
i
c
d
a
t
a
se
t
s
H
y
b
r
i
d
a
r
c
h
i
t
e
c
t
u
r
e
c
o
m
b
i
n
e
d
w
i
t
h
LS
TM
a
n
d
M
L
a
l
g
o
r
i
t
h
ms
5.
CO
NCLU
SI
O
N
T
h
e
co
m
p
r
e
h
en
s
iv
e
r
ev
iew
o
f
L
STM
-
b
ased
ac
tiv
ity
d
etec
ti
o
n
u
tili
zin
g
s
k
eleto
n
jo
in
ts
u
n
d
er
s
co
r
es
th
e
tr
an
s
f
o
r
m
ativ
e
ca
p
ab
ilit
ie
s
o
f
DL
tech
n
iq
u
es
with
in
t
h
e
s
p
h
er
e
o
f
HAR
.
T
h
e
co
m
b
in
atio
n
o
f
L
STM
n
etwo
r
k
s
with
em
er
g
in
g
t
r
an
s
f
o
r
m
er
ar
c
h
itectu
r
es
s
ig
n
if
ie
s
a
n
o
tab
le
a
d
v
an
ce
in
ef
f
ec
t
iv
ely
ca
p
tu
r
i
n
g
t
h
e
tem
p
o
r
al
a
n
d
s
p
atial
in
tr
icac
ie
s
o
f
h
u
m
a
n
m
o
v
em
e
n
ts
,
th
u
s
im
p
r
o
v
i
n
g
th
e
p
r
ec
is
io
n
an
d
r
e
liab
ilit
y
o
f
ac
tiv
ity
r
ec
o
g
n
itio
n
s
y
s
tem
s
.
No
n
eth
eless
,
d
esp
ite
th
e
en
co
u
r
ag
in
g
p
r
o
g
r
ess
,
th
e
d
o
m
ain
s
till
f
ac
es
co
n
s
id
er
a
b
le
o
b
s
tacle
s
,
in
clu
d
in
g
im
p
lem
en
tin
g
tech
n
o
lo
g
ies
in
r
e
al
-
wo
r
ld
s
ce
n
ar
io
s
.
T
ac
k
li
n
g
t
h
ese
is
s
u
es
th
r
o
u
g
h
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Lo
n
g
s
h
o
r
t
-
term me
mo
r
y
b
a
s
ed
a
ctivity
d
etec
tio
n
u
s
in
g
s
ke
leto
n
…
(
B
a
kk
a
la
S
a
n
th
a
K
u
ma
r
)
3033
in
n
o
v
ativ
e
s
tr
ateg
ies
lik
e
en
s
em
b
le
lear
n
in
g
,
m
u
ltimo
d
al
d
ata
in
teg
r
atio
n
,
an
d
f
ed
e
r
ated
lear
n
in
g
will
b
e
ess
en
tial
f
o
r
cr
ea
tin
g
r
esil
ien
t
s
y
s
tem
s
th
at
em
p
h
asize
u
s
er
p
r
i
v
ac
y
wh
ile
en
h
an
cin
g
th
eir
o
p
er
atio
n
al
ef
f
icien
cy
.
As
r
esear
ch
ad
v
a
n
ce
s
,
th
e
k
n
o
wled
g
e
ac
q
u
ir
e
d
f
r
o
m
t
h
is
r
ev
iew
will
lay
th
e
g
r
o
u
n
d
wo
r
k
f
o
r
f
u
tu
r
e
in
v
esti
g
atio
n
s
aim
ed
at
r
ef
i
n
in
g
ac
tiv
ity
r
ec
o
g
n
itio
n
s
y
s
tem
s
,
p
av
i
n
g
t
h
e
way
f
o
r
a
d
ee
p
er
u
n
d
er
s
tan
d
i
n
g
a
n
d
in
ter
p
r
etatio
n
o
f
h
u
m
an
b
e
h
av
io
r
s
in
in
tr
icate
co
n
te
x
ts
.
T
h
e
im
p
o
r
t
an
ce
o
f
d
etec
tin
g
u
n
u
s
u
al
ac
tiv
ities
in
m
o
d
er
n
s
u
r
v
eillan
ce
h
as b
ee
n
h
ig
h
lig
h
t
ed
in
th
is
r
ev
iew.
I
t e
m
p
h
asizes th
e
im
p
o
r
tan
ce
o
f
h
av
in
g
s
k
eleto
n
jo
in
t
d
ata
as
a
s
tar
tin
g
p
o
in
t
f
o
r
id
en
tify
in
g
s
u
ch
ac
tiv
ities
.
T
h
e
in
teg
r
atio
n
o
f
th
e
T
r
an
s
f
o
r
m
e
r
an
d
L
STM
m
o
d
els
ca
n
p
r
o
v
id
e
a
m
o
r
e
ac
cu
r
ate
an
d
r
o
b
u
s
t
s
o
lu
tio
n
to
id
en
tify
in
g
c
o
m
p
lex
ac
tiv
ity
p
atter
n
s
.
T
h
e
d
ev
elo
p
m
en
t
o
f
a
clo
u
d
-
b
ased
L
STM
f
r
am
ewo
r
k
th
at
is
ca
p
ab
le
o
f
h
an
d
lin
g
jo
i
n
t
s
k
eleto
n
an
aly
s
is
is
ex
p
ec
ted
t
o
h
elp
im
p
r
o
v
e
th
e
d
etec
tio
n
o
f
u
n
u
s
u
al
ac
tiv
ities
.
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
Ba
k
k
a
la S
a
n
t
h
a
Ku
m
a
r
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
R
.
S
h
a
n
k
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
T
h
e
au
th
o
r
s
h
av
e
n
o
co
n
f
lict o
f
in
ter
ests
to
d
ec
lar
e
th
at
ar
e
r
elev
an
t to
th
e
c
o
n
ten
t o
f
th
is
a
r
ticle.
DATA AV
AI
L
AB
I
L
I
T
Y
Data
a
v
aila
b
i
lit
y
is
n
o
t
a
p
p
li
ca
b
le
t
o
t
h
is
p
ap
er
as
n
o
n
e
w
d
a
t
a
w
er
e
cr
ea
te
d
o
r
an
al
y
z
e
d
i
n
t
h
is
s
t
u
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
M
.
G
h
i
s
l
i
e
r
i
,
G
.
L.
C
e
r
o
n
e
,
M
.
K
n
a
f
l
i
t
z
,
a
n
d
V
.
A
g
o
s
t
i
n
i
,
“
L
o
n
g
sh
o
r
t
-
t
e
r
m
memo
r
y
(
LST
M
)
r
e
c
u
r
r
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
f
o
r
m
u
s
c
l
e
a
c
t
i
v
i
t
y
d
e
t
e
c
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
N
e
u
r
o
En
g
i
n
e
e
ri
n
g
a
n
d
R
e
h
a
b
i
l
i
t
a
t
i
o
n
,
v
o
l
.
1
8
,
n
o
.
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
8
6
/
s
1
2
9
8
4
-
0
2
1
-
0
0
9
4
5
-
w.
[
2
]
K
.
X
i
a
,
J
.
H
u
a
n
g
,
a
n
d
H
.
W
a
n
g
,
“
LS
TM
-
C
N
N
a
r
c
h
i
t
e
c
t
u
r
e
f
o
r
h
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
5
6
8
5
5
–
5
6
8
6
6
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
8
2
2
2
5
.
[
3
]
C
.
S
i
,
W
.
C
h
e
n
,
W
.
W
a
n
g
,
L.
W
a
n
g
,
a
n
d
T.
Ta
n
,
“
A
n
a
t
t
e
n
t
i
o
n
e
n
h
a
n
c
e
d
g
r
a
p
h
c
o
n
v
o
l
u
t
i
o
n
a
l
LST
M
n
e
t
w
o
r
k
f
o
r
sk
e
l
e
t
o
n
-
b
a
s
e
d
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
I
EE
E
C
o
m
p
u
t
e
r
S
o
c
i
e
t
y
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
si
o
n
a
n
d
Pa
t
t
e
rn
Re
c
o
g
n
i
t
i
o
n
,
2
0
1
9
,
p
p
.
1
2
2
7
–
1
2
3
6
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
.
2
0
1
9
.
0
0
1
3
2
.
[
4
]
J.
D
o
a
n
d
M
.
K
i
m
,
“
S
k
a
t
e
F
o
r
m
e
r
:
s
k
e
l
e
t
a
l
-
t
e
m
p
o
r
a
l
t
r
a
n
sf
o
r
mer
f
o
r
h
u
m
a
n
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
C
o
m
p
u
t
e
r
Vi
si
o
n
–
(
E
C
C
V
2
0
2
4
)
.
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
4
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
1
-
7
2
9
4
0
-
9
.
[
5
]
Q
.
W
a
n
g
,
S
.
S
h
i
,
J.
H
e
,
J
.
P
e
n
g
,
T.
Li
u
,
a
n
d
R
.
W
e
n
g
,
“
I
I
P
-
t
r
a
n
sf
o
r
mer:
i
n
t
r
a
-
i
n
t
e
r
-
p
a
r
t
t
r
a
n
sf
o
r
m
e
r
f
o
r
sk
e
l
e
t
o
n
-
b
a
se
d
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
2
0
2
3
I
E
EE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
B
i
g
D
a
t
a
,
Bi
g
D
a
t
a
2
0
2
3
,
2
0
2
3
,
p
p
.
9
3
6
–
9
4
5
,
d
o
i
:
1
0
.
1
1
0
9
/
B
i
g
D
a
t
a
5
9
0
4
4
.
2
0
2
3
.
1
0
3
8
6
9
7
0
.
[
6
]
L.
W
e
n
g
,
W
.
L
o
u
,
a
n
d
F
.
G
a
o
,
“
L
a
n
g
u
a
g
e
g
u
i
d
e
d
g
r
a
p
h
t
r
a
n
s
f
o
r
m
e
r
f
o
r
s
k
e
l
e
t
o
n
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
C
o
m
m
u
n
i
c
a
t
i
o
n
s
i
n
C
o
m
p
u
t
e
r
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
,
S
i
n
g
a
p
o
r
e
:
S
p
r
i
n
g
e
r
,
2
0
2
4
,
p
p
.
2
8
3
–
299
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
981
-
99
-
8141
-
0_22.
[
7
]
J.
K
o
n
g
,
Y
.
B
i
a
n
,
a
n
d
M
.
Ji
a
n
g
,
“
M
TT:
mu
l
t
i
-
sc
a
l
e
t
e
m
p
o
r
a
l
t
r
a
n
sf
o
r
mer
f
o
r
sk
e
l
e
t
o
n
-
b
a
s
e
d
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
I
EE
E
S
i
g
n
a
l
Pro
c
e
ssi
n
g
L
e
t
t
e
rs
,
v
o
l
.
2
9
,
p
p
.
5
2
8
–
5
3
2
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
LSP
.
2
0
2
2
.
3
1
4
2
6
7
5
.
[
8
]
X
.
H
u
a
n
g
,
Y
.
Y
u
a
n
,
C
.
C
h
a
n
g
,
Y
.
G
a
o
,
C
.
Z
h
e
n
g
,
a
n
d
L
.
Y
a
n
,
“
H
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
me
t
h
o
d
b
a
se
d
o
n
e
d
g
e
c
o
m
p
u
t
i
n
g
-
a
ssi
s
t
e
d
a
n
d
G
R
U
d
e
e
p
l
e
a
r
n
i
n
g
n
e
t
w
o
r
k
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
3
,
n
o
.
1
6
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
3
1
6
9
0
5
9
.
[
9
]
J.
Y
e
,
X
.
L
i
,
X
.
Zh
a
n
g
,
Q
.
Z
h
a
n
g
,
a
n
d
W
.
C
h
e
n
,
“
D
e
e
p
l
e
a
r
n
i
n
g
-
b
a
s
e
d
h
u
ma
n
a
c
t
i
v
i
t
y
r
e
a
l
-
t
i
m
e
r
e
c
o
g
n
i
t
i
o
n
f
o
r
p
e
d
e
s
t
r
i
a
n
n
a
v
i
g
a
t
i
o
n
,
”
S
e
n
s
o
rs
,
v
o
l
.
2
0
,
n
o
.
9
,
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
0
0
9
2
5
7
4
.
[
1
0
]
J.
D
.
D
o
m
i
n
g
o
,
J.
G
.
-
G
.
-
B
e
r
me
j
o
,
a
n
d
E.
Z
a
l
a
ma
,
“
I
mp
r
o
v
i
n
g
h
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
i
n
t
e
g
r
a
t
i
n
g
LST
M
w
i
t
h
d
i
f
f
e
r
e
n
t
d
a
t
a
so
u
r
c
e
s
:
f
e
a
t
u
r
e
s,
o
b
j
e
c
t
d
e
t
e
c
t
i
o
n
a
n
d
s
k
e
l
e
t
o
n
t
r
a
c
k
i
n
g
,
”
I
E
EE
A
c
c
e
ss
,
v
o
l
.
1
0
,
p
p
.
6
8
2
1
3
–
6
8
2
3
0
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
2
.
3
1
8
6
4
6
5
.
[
1
1
]
N
.
Ta
s
n
i
m
,
M
.
K
.
I
sl
a
m
,
a
n
d
J.
H
.
B
a
e
k
,
“
D
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
h
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
sp
a
t
i
o
-
t
e
m
p
o
r
a
l
i
m
a
g
e
f
o
r
m
a
t
i
o
n
o
f
s
k
e
l
e
t
o
n
j
o
i
n
t
s,”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
1
,
n
o
.
6
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
1
0
6
2
6
7
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
26
:
3
0
2
6
-
3
0
3
5
3034
[
1
2
]
R
.
A
.
P
r
a
b
a
a
n
d
L
.
S
u
g
a
n
t
h
i
,
“
H
A
R
N
e
t
:
a
u
t
o
ma
t
i
c
r
e
c
o
g
n
i
t
i
o
n
o
f
h
u
ma
n
a
c
t
i
v
i
t
y
f
r
o
m
m
o
b
i
l
e
h
e
a
l
t
h
d
a
t
a
u
si
n
g
C
N
N
a
n
d
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
o
f
LST
M
w
i
t
h
S
V
M
,
”
A
u
t
o
m
a
t
i
k
a
,
v
o
l
.
6
5
,
n
o
.
1
,
p
p
.
1
6
7
–
1
7
8
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
8
0
/
0
0
0
5
1
1
4
4
.
2
0
2
3
.
2
2
9
0
7
3
6
.
[
1
3
]
B
.
K
i
t
c
h
e
n
h
a
m
a
n
d
S
.
C
h
a
r
t
e
r
s,
“
G
u
i
d
e
l
i
n
e
s
f
o
r
p
e
r
f
o
r
mi
n
g
s
y
s
t
e
ma
t
i
c
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
s
i
n
s
o
f
t
w
a
r
e
e
n
g
i
n
e
e
r
i
n
g
,
”
EB
S
E
T
e
c
h
n
i
c
a
l
Re
p
o
r
t
,
EB
S
E
-
2
0
0
7
-
0
1
,
K
e
e
l
e
U
n
i
v
e
r
si
t
y
a
n
d
U
n
i
v
e
r
si
t
y
o
f
D
u
r
h
a
m
,
U
n
i
t
e
d
K
i
n
g
d
o
m
,
2
0
0
7
.
[
1
4
]
M
.
J.
P
a
g
e
e
t
a
l
.
,
“
Th
e
P
R
I
S
M
A
2
0
2
0
s
t
a
t
e
m
e
n
t
:
A
n
u
p
d
a
t
e
d
g
u
i
d
e
l
i
n
e
f
o
r
r
e
p
o
r
t
i
n
g
sy
st
e
mat
i
c
r
e
v
i
e
w
s,”
B
MJ
,
v
o
l
.
3
7
2
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
3
6
/
b
m
j
.
n
7
1
.
[
1
5
]
A
.
El
k
h
o
l
y
,
M
.
E
.
H
u
ss
e
i
n
,
W
.
G
o
maa
,
D
.
D
a
m
e
n
,
a
n
d
E.
S
a
b
a
,
“
Ef
f
i
c
i
e
n
t
a
n
d
r
o
b
u
st
sk
e
l
e
t
o
n
-
b
a
s
e
d
q
u
a
l
i
t
y
a
ss
e
ssm
e
n
t
a
n
d
a
b
n
o
r
m
a
l
i
t
y
d
e
t
e
c
t
i
o
n
i
n
h
u
ma
n
a
c
t
i
o
n
p
e
r
f
o
r
ma
n
c
e
,
”
I
EEE
J
o
u
rn
a
l
o
f
Bi
o
m
e
d
i
c
a
l
a
n
d
H
e
a
l
t
h
I
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
2
4
,
n
o
.
1
,
p
p
.
2
8
0
–
2
9
1
,
A
u
g
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
JB
H
I
.
2
0
1
9
.
2
9
0
4
3
2
1
.
[
1
6
]
C
.
N
.
P
h
y
o
,
T
.
T.
Zi
n
,
a
n
d
P
.
T
i
n
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
r
e
c
o
g
n
i
z
i
n
g
h
u
m
a
n
a
c
t
i
v
i
t
i
e
s
u
s
i
n
g
m
o
t
i
o
n
s
o
f
sk
e
l
e
t
a
l
j
o
i
n
t
s,”
I
E
EE
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
o
n
su
m
e
r E
l
e
c
t
r
o
n
i
c
s
,
v
o
l
.
6
5
,
n
o
.
2
,
p
p
.
2
4
3
–
2
5
2
,
M
a
y
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
TC
E
.
2
0
1
9
.
2
9
0
8
9
8
6
.
[
1
7
]
B
.
X
.
B
.
Y
u
,
Y
.
L
i
u
,
a
n
d
K
.
C
.
C
.
C
h
a
n
,
“
S
k
e
l
e
t
o
n
-
b
a
s
e
d
d
e
t
e
c
t
i
o
n
o
f
a
b
n
o
r
mal
i
t
i
e
s
i
n
h
u
ma
n
a
c
t
i
o
n
s
u
si
n
g
g
r
a
p
h
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
t
w
o
r
k
s,”
i
n
Pro
c
e
e
d
i
n
g
s
-
2
0
2
0
2
n
d
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
T
r
a
n
sd
i
sc
i
p
l
i
n
a
ry
AI
,
T
ra
n
sAI
2
0
2
0
,
2
0
2
0
,
p
p
.
1
3
1
–
137
,
d
o
i
:
1
0
.
1
1
0
9
/
Tr
a
n
sA
I
4
9
8
3
7
.
2
0
2
0
.
0
0
0
3
0
.
[
1
8
]
T.
A
h
m
a
d
,
S
.
T
.
H
.
R
i
z
v
i
,
a
n
d
N
.
K
a
n
w
a
l
,
“
Tr
a
n
sf
o
r
m
i
n
g
s
p
a
t
i
o
-
t
e
mp
o
r
a
l
sel
f
-
a
t
t
e
n
t
i
o
n
u
s
i
n
g
a
c
t
i
o
n
e
m
b
e
d
d
i
n
g
f
o
r
sk
e
l
e
t
o
n
-
b
a
s
e
d
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
V
i
su
a
l
C
o
m
m
u
n
i
c
a
t
i
o
n
a
n
d
I
m
a
g
e
R
e
p
r
e
se
n
t
a
t
i
o
n
,
v
o
l
.
9
5
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
v
c
i
r
.
2
0
2
3
.
1
0
3
8
9
2
.
[
1
9
]
D
.
O
.
Esa
n
,
P
.
A
.
O
w
o
l
a
w
i
,
a
n
d
C
.
Tu
,
“
D
e
t
e
c
t
i
o
n
o
f
a
n
o
m
a
l
o
u
s
b
e
h
a
v
i
o
u
r
a
l
p
a
t
t
e
r
n
s
i
n
u
n
i
v
e
r
s
i
t
y
e
n
v
i
r
o
n
men
t
u
si
n
g
C
N
N
-
LSTM
,
”
i
n
2
0
2
0
2
3
r
d
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
n
f
o
rm
a
t
i
o
n
Fu
s
i
o
n
,
FU
S
I
O
N
2
0
2
0
,
2
0
2
0
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
2
3
9
1
9
/
F
U
S
I
O
N
4
5
0
0
8
.
2
0
2
0
.
9
1
9
0
4
0
6
.
[
2
0
]
K
.
Zh
o
u
,
T.
W
u
,
C
.
W
a
n
g
,
J
.
W
a
n
g
,
a
n
d
C
.
L
i
,
“
S
k
e
l
e
t
o
n
b
a
se
d
a
b
n
o
r
mal
b
e
h
a
v
i
o
r
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
s
p
a
t
i
o
-
t
e
m
p
o
r
a
l
c
o
n
v
o
l
u
t
i
o
n
a
n
d
a
t
t
e
n
t
i
o
n
-
b
a
s
e
d
LS
TM
,
”
Pro
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
7
4
,
p
p
.
4
2
4
–
4
3
2
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.
2
0
2
0
.
0
6
.
1
1
0
.
[
2
1
]
R
.
N
a
l
e
,
M
.
S
a
w
a
r
b
a
n
d
h
e
,
N
.
C
h
e
g
o
g
o
j
u
,
a
n
d
V
.
S
a
t
p
u
t
e
,
“
S
u
sp
i
c
i
o
u
s
h
u
m
a
n
a
c
t
i
v
i
t
y
d
e
t
e
c
t
i
o
n
u
si
n
g
p
o
s
e
e
st
i
ma
t
i
o
n
a
n
d
LSTM
,
”
i
n
2
0
2
1
I
n
t
e
rn
a
t
i
o
n
a
l
S
y
m
p
o
si
u
m
o
f
Asi
a
n
C
o
n
t
r
o
l
Ass
o
c
i
a
t
i
o
n
o
n
I
n
t
e
l
l
i
g
e
n
t
R
o
b
o
t
i
c
s
a
n
d
I
n
d
u
st
r
i
a
l
A
u
t
o
m
a
t
i
o
n
,
I
RI
A
2
0
2
1
,
2
0
2
1
,
p
p
.
1
9
7
–
2
0
2
,
d
o
i
:
1
0
.
1
1
0
9
/
I
R
I
A
5
3
0
0
9
.
2
0
2
1
.
9
5
8
8
7
1
9
.
[
2
2
]
G
.
P
a
n
d
e
y
,
A
.
K
.
K
a
r
n
,
a
n
d
M
.
J
h
a
,
“
H
u
m
a
n
A
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
C
N
N
-
LST
M
-
G
R
U
m
o
d
e
l
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
Re
s
e
a
rc
h
J
o
u
rn
a
l
o
n
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
ri
n
g
H
u
b
,
v
o
l
.
2
,
n
o
.
0
4
,
p
p
.
8
8
9
–
8
9
4
,
2
0
2
4
,
d
o
i
:
1
0
.
4
7
3
9
2
/
i
r
j
a
e
h
.
2
0
2
4
.
0
1
2
5
.
[
2
3
]
P
.
K
a
l
l
e
p
a
l
l
y
a
n
d
M
.
R
a
j
e
sh
,
“
LST
M
-
b
a
se
d
d
e
e
p
l
e
a
r
n
i
n
g
a
r
c
h
i
t
e
c
t
u
r
e
f
o
r
r
e
c
o
g
n
i
t
i
o
n
o
f
h
u
ma
n
a
c
t
i
v
i
t
i
e
s
,
”
i
n
A
d
v
a
n
c
e
s
i
n
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
,
S
i
n
g
a
p
o
r
e
:
S
p
r
i
n
g
e
r
,
2
0
2
3
,
p
p
.
1
6
3
–
1
7
2
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
19
-
5
4
4
3
-
6
_
1
4
.
[
2
4
]
Y
.
L
i
,
T
.
G
u
o
,
X
.
L
i
u
,
a
n
d
R
.
X
i
a
,
“
S
k
e
l
e
t
o
n
-
b
a
s
e
d
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
w
i
t
h
L
i
e
g
r
o
u
p
a
n
d
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s
,
”
i
n
2
0
1
9
I
E
E
E
4
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
S
i
g
n
a
l
a
n
d
I
m
a
g
e
P
r
o
c
e
s
s
i
n
g
,
I
C
S
I
P
2
0
1
9
,
2
0
1
9
,
p
p
.
2
6
–
30
,
d
o
i
:
1
0
.
1
1
0
9
/
S
I
P
R
O
C
E
S
S
.
2
0
1
9
.
8
8
6
8
5
4
8
.
[
2
5
]
H
.
Y
u
n
e
t
a
l
.
,
“
D
e
v
e
l
o
p
m
e
n
t
o
f
a
c
t
i
o
n
-
r
e
c
o
g
n
i
t
i
o
n
t
e
c
h
n
o
l
o
g
y
u
si
n
g
LSTM
b
a
s
e
d
o
n
s
k
e
l
e
t
o
n
d
a
t
a
,
”
i
n
T
h
e
9
t
h
I
I
AE
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
n
d
u
st
r
i
a
l
A
p
p
l
i
c
a
t
i
o
n
En
g
i
n
e
e
ri
n
g
2
0
2
1
(
I
C
I
AE
2
0
2
1
)
,
2
0
2
1
,
p
p
.
3
5
–
39
,
d
o
i
:
1
0
.
1
2
7
9
2
/
i
c
i
a
e
2
0
2
1
.
0
0
9
.
[
2
6
]
G
.
W
u
,
Z.
G
u
o
,
L.
Li
,
a
n
d
C
.
W
a
n
g
,
“
V
i
d
e
o
a
b
n
o
r
mal
e
v
e
n
t
d
e
t
e
c
t
i
o
n
b
a
se
d
o
n
C
N
N
a
n
d
LST
M
,
”
i
n
2
0
2
0
I
EEE
5
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
i
g
n
a
l
a
n
d
I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
I
C
S
I
P
2
0
2
0
,
2
0
2
0
,
p
p
.
3
3
4
–
3
3
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
I
P
4
9
8
9
6
.
2
0
2
0
.
9
3
3
9
4
2
8
.
[
2
7
]
G
.
P
a
o
l
e
t
t
i
,
C
.
B
e
y
a
n
,
a
n
d
A
.
D
.
B
u
e
,
“
S
K
ELT
ER
:
u
n
s
u
p
e
r
v
i
se
d
sk
e
l
e
t
o
n
a
c
t
i
o
n
d
e
n
o
i
si
n
g
a
n
d
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
t
r
a
n
sf
o
r
m
e
r
s,”
Fro
n
t
i
e
rs
i
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
5
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
8
9
/
f
c
o
m
p
.
2
0
2
3
.
1
2
0
3
9
0
1
.
[
2
8
]
T.
A
c
h
a
r
y
a
,
A
.
A
n
n
a
ma
l
a
i
,
a
n
d
M
.
F
.
C
h
o
u
i
k
h
a
,
“
Ef
f
i
c
a
c
y
o
f
C
N
N
-
b
i
d
i
r
e
c
t
i
o
n
a
l
LS
TM
h
y
b
r
i
d
mo
d
e
l
f
o
r
n
e
t
w
o
r
k
-
b
a
se
d
a
n
o
ma
l
y
d
e
t
e
c
t
i
o
n
,
”
i
n
1
3
t
h
I
EEE
S
y
m
p
o
si
u
m
o
n
C
o
m
p
u
t
e
r
Ap
p
l
i
c
a
t
i
o
n
s
a
n
d
I
n
d
u
st
ri
a
l
El
e
c
t
ro
n
i
c
s
(
I
S
C
AI
E
2
0
2
3
)
,
2
0
2
3
,
p
p
.
3
4
8
–
3
5
3
,
d
o
i
:
1
0
.
1
1
0
9
/
I
S
C
A
I
E5
7
7
3
9
.
2
0
2
3
.
1
0
1
6
5
0
8
8
.
[
2
9
]
J.
L
i
u
,
J
.
B
a
i
,
H
.
Li
,
a
n
d
B
.
S
u
n
,
“
I
mp
r
o
v
e
d
LST
M
-
b
a
se
d
a
b
n
o
r
ma
l
s
t
r
e
a
m
d
a
t
a
d
e
t
e
c
t
i
o
n
a
n
d
c
o
r
r
e
c
t
i
o
n
sy
s
t
e
m
f
o
r
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s,
”
I
EE
E
T
r
a
n
s
a
c
t
i
o
n
s
o
n
I
n
d
u
st
ri
a
l
I
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
1
8
,
n
o
.
2
,
p
p
.
1
2
8
2
–
1
2
9
0
,
J
a
n
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
I
.
2
0
2
1
.
3
0
7
9
5
0
4
.
[
3
0
]
L.
W
e
n
e
t
a
l
.
,
“
F
r
e
q
u
e
n
c
y
g
u
i
d
a
n
c
e
mat
t
e
r
s:
s
k
e
l
e
t
a
l
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
b
y
f
r
e
q
u
e
n
c
y
-
a
w
a
r
e
mi
x
e
d
t
r
a
n
sf
o
r
m
e
r
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
3
2
n
d
AC
M
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
ren
c
e
o
n
M
u
l
t
i
m
e
d
,
M
e
l
b
o
u
r
n
e
,
A
u
st
r
a
l
i
a
,
2
0
2
4
.
[
3
1
]
S
.
O
r
a
k
i
,
H
.
Z
h
u
a
n
g
,
a
n
d
J
.
Li
a
n
g
,
“
LO
R
TSA
R
:
l
o
w
-
r
a
n
k
t
r
a
n
sf
o
r
m
e
r
f
o
r
sk
e
l
e
t
o
n
-
b
a
s
e
d
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
Ad
v
a
n
c
e
s
i
n
Vi
su
a
l
C
o
m
p
u
t
i
n
g
(
I
S
V
C
2
0
2
4
)
,
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
4
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
1
-
7
7
3
9
2
-
1
.
[
3
2
]
R
.
K
.
M
u
r
t
h
y
,
S
.
D
h
a
n
r
a
j
,
T.
N
.
M
a
n
j
u
n
a
t
h
,
A
.
N
.
P
r
a
s
a
d
,
P
.
K
.
P
a
r
e
e
k
,
a
n
d
H
.
N
.
K
u
m
a
r
,
“
H
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
C
N
N
a
n
d
l
o
n
g
-
t
e
r
m
sh
o
r
t
-
t
e
r
m
m
e
mo
r
y
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
H
e
a
l
t
h
S
c
i
e
n
c
e
s
,
v
o
l
.
6
,
n
o
.
S
6
,
p
p
.
1
0
7
9
7
–
1
0
8
0
9
,
2
0
2
2
,
d
o
i
:
1
0
.
5
3
7
3
0
/
i
j
h
s.
v
6
n
s
6
.
1
2
9
1
9
.
[
3
3
]
S
.
A
.
R
o
sel
i
n
e
,
S
.
K
a
r
t
h
i
k
,
a
n
d
I
.
N
.
V
.
D
.
S
r
u
t
i
,
“
I
n
t
e
l
l
i
g
e
n
t
h
u
m
a
n
a
n
o
ma
l
y
d
e
t
e
c
t
i
o
n
u
si
n
g
LST
M
a
u
t
o
e
n
c
o
d
e
r
s,”
i
n
3
r
d
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
A
d
v
a
n
c
e
s
i
n
C
o
m
p
u
t
i
n
g
,
C
o
m
m
u
n
i
c
a
t
i
o
n
a
n
d
A
p
p
l
i
e
d
I
n
f
o
rm
a
t
i
c
s,
AC
C
AI
2
0
2
4
,
2
0
2
4
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
A
I
6
1
0
6
1
.
2
0
2
4
.
1
0
6
0
2
4
5
4
.
[
3
4
]
Y
.
Z
h
a
l
g
a
s
b
a
y
e
v
a
n
d
N
.
A
.
T
u
,
“
T
w
o
-
B
r
a
n
c
h
st
a
c
k
e
d
t
r
a
n
sf
o
r
m
e
r
f
o
r
2
D
s
k
e
l
e
t
o
n
-
b
a
se
d
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
2
0
2
3
1
7
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
U
b
i
q
u
i
t
o
u
s
I
n
f
o
rm
a
t
i
o
n
M
a
n
a
g
e
m
e
n
t
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
,
I
M
C
O
M
2
0
2
3
,
2
0
2
3
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
M
C
O
M
5
6
9
0
9
.
2
0
2
3
.
1
0
0
3
5
6
6
0
.
[
3
5
]
J.
Z
h
a
n
g
,
Y
.
Ji
a
,
W
.
X
i
e
,
a
n
d
Z
.
Tu
,
“
Z
o
o
m
t
r
a
n
sf
o
r
mer
f
o
r
s
k
e
l
e
t
o
n
-
b
a
s
e
d
g
r
o
u
p
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
,
”
I
EEE
T
ra
n
sa
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
2
,
n
o
.
1
2
,
p
p
.
8
6
4
6
–
8
6
5
9
,
F
e
b
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
S
V
T.
2
0
2
2
.
3
1
9
3
5
7
4
.
A
.
[
3
6
]
A
.
Zh
u
e
t
a
l
.
,
“
E
x
p
l
o
r
i
n
g
a
r
i
c
h
sp
a
t
i
a
l
–
t
e
m
p
o
r
a
l
d
e
p
e
n
d
e
n
t
r
e
l
a
t
i
o
n
a
l
m
o
d
e
l
f
o
r
s
k
e
l
e
t
o
n
-
b
a
se
d
a
c
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
b
y
b
i
d
i
r
e
c
t
i
o
n
a
l
LSTM
-
C
N
N
,
”
N
e
u
r
o
c
o
m
p
u
t
i
n
g
,
v
o
l
.
4
1
4
,
p
p
.
9
0
–
1
0
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m
.
2
0
2
0
.
0
7
.
0
6
8
.
[
3
7
]
B
.
L
i
n
d
e
ma
n
n
,
B
.
M
a
sc
h
l
e
r
,
N
.
S
a
h
l
a
b
,
a
n
d
M
.
W
e
y
r
i
c
h
,
“
A
s
u
r
v
e
y
o
n
a
n
o
ma
l
y
d
e
t
e
c
t
i
o
n
f
o
r
t
e
c
h
n
i
c
a
l
s
y
st
e
ms
u
s
i
n
g
LST
M
n
e
t
w
o
r
k
s,”
C
o
m
p
u
t
e
rs
i
n
I
n
d
u
st
r
y
,
v
o
l
.
1
3
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
i
n
d
.
2
0
2
1
.
1
0
3
4
9
8
.
[
3
8
]
S
.
N
a
t
h
a
,
“
A
s
y
st
e
ma
t
i
c
r
e
v
i
e
w
o
f
a
n
o
m
a
l
y
d
e
t
e
c
t
i
o
n
u
si
n
g
ma
c
h
i
n
e
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
Q
u
a
i
d
-
e
-
Aw
a
m
U
n
i
v
e
rs
i
t
y
Re
se
a
rc
h
J
o
u
r
n
a
l
o
f
E
n
g
i
n
e
e
r
i
n
g
,
S
c
i
e
n
c
e
&
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
0
,
n
o
.
1
,
p
p
.
8
3
–
9
4
,
2
0
2
2
,
d
o
i
:
1
0
.
5
2
5
8
4
/
q
r
j
.
2
0
0
1
.
1
1
.
[
3
9
]
S
.
M
e
k
r
u
k
sa
v
a
n
i
c
h
a
n
d
A
.
J
i
t
p
a
t
t
a
n
a
k
u
l
,
“
LS
TM
n
e
t
w
o
r
k
s
u
si
n
g
s
mart
p
h
o
n
e
d
a
t
a
f
o
r
se
n
s
o
r
-
b
a
s
e
d
h
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
i
n
smar
t
h
o
mes,
”
S
e
n
s
o
rs
,
v
o
l
.
2
1
,
n
o
.
5
,
p
p
.
1
–
2
5
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
1
0
5
1
6
3
6
.
[
4
0
]
S
.
M
o
h
se
n
,
“
R
e
c
o
g
n
i
t
i
o
n
o
f
h
u
ma
n
a
c
t
i
v
i
t
y
u
s
i
n
g
G
R
U
d
e
e
p
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m,
”
Mu
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
8
2
,
n
o
.
3
0
,
p
p
.
4
7
7
3
3
–
4
7
7
4
9
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
0
4
2
-
0
2
3
-
1
5
5
7
1
-
y.
[
4
1
]
K
.
S
.
L
o
w
a
n
d
S
.
K
.
E
n
g
,
“
P
e
r
f
o
r
ma
n
c
e
e
v
a
l
u
a
t
i
o
n
o
f
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s f
o
r
h
u
m
a
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
sy
s
t
e
m,
”
J
o
u
rn
a
l
o
f
Ph
y
s
i
c
s:
C
o
n
f
e
r
e
n
c
e
S
e
ri
e
s
,
v
o
l
.
2
6
4
1
,
n
o
.
1
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
4
2
-
6
5
9
6
/
2
6
4
1
/
1
/
0
1
2
0
1
2
.
[
4
2
]
S
.
S
a
b
b
u
a
n
d
V
.
G
a
n
e
s
a
n
,
“
LST
M
-
b
a
sed
n
e
u
r
a
l
n
e
t
w
o
r
k
t
o
r
e
c
o
g
n
i
z
e
h
u
m
a
n
a
c
t
i
v
i
t
i
e
s
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
A
p
p
l
i
e
d
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
S
o
f
t
C
o
m
p
u
t
i
n
g
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
2
/
1
6
8
1
0
9
6
.
[
4
3
]
P
.
M
a
r
a
g
a
t
h
a
v
a
l
l
i
,
R
.
V
i
v
i
t
h
a
,
M
.
S
a
r
a
n
y
a
,
a
n
d
U
.
D
i
n
e
s
h
,
“
I
n
c
o
r
p
o
r
a
t
i
n
g
LSTM
me
t
h
o
d
o
n
mo
d
i
f
i
e
d
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
f
o
r
sen
s
o
r
b
a
se
d
h
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
s
y
st
e
m,
”
I
O
P
C
o
n
f
e
re
n
c
e
S
e
r
i
e
s:
M
a
t
e
r
i
a
l
s
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
ri
n
g
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
5
7
-
8
9
9
X
/
1
0
6
5
/
1
/
0
1
2
0
4
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Lo
n
g
s
h
o
r
t
-
term me
mo
r
y
b
a
s
ed
a
ctivity
d
etec
tio
n
u
s
in
g
s
ke
leto
n
…
(
B
a
kk
a
la
S
a
n
th
a
K
u
ma
r
)
3035
[
4
4
]
I
.
U
.
K
h
a
n
,
S
.
A
f
z
a
l
,
a
n
d
J
.
W
.
L
e
e
,
“
H
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
v
i
a
h
y
b
r
i
d
d
e
e
p
l
e
a
r
n
i
n
g
b
a
s
e
d
m
o
d
e
l
,
”
S
e
n
s
o
rs
,
v
o
l
.
2
2
,
n
o
.
1
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s2
2
0
1
0
3
2
3
.
[
4
5
]
S
.
W
.
P
i
e
n
a
a
r
a
n
d
R
.
M
a
l
e
k
i
a
n
,
“
H
u
man
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
LST
M
-
R
N
N
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
a
r
c
h
i
t
e
c
t
u
r
e
,
”
i
n
2
0
1
9
I
E
EE
2
n
d
Wi
r
e
l
e
ss A
f
ri
c
a
C
o
n
f
e
re
n
c
e
,
WA
C
2
0
1
9
,
2
0
1
9
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
A
F
R
I
C
A
.
2
0
1
9
.
8
8
4
3
4
0
3
.
[
4
6
]
B
.
O
l
u
w
a
l
a
d
e
,
S
.
N
e
e
l
a
,
J
.
W
a
w
i
r
a
,
T
.
A
d
e
j
u
m
o
,
a
n
d
S
.
P
u
r
k
a
y
a
st
h
a
,
“
H
u
man
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
s
o
n
smar
t
p
h
o
n
e
s
a
n
d
sm
a
r
t
w
a
t
c
h
e
s
se
n
so
r
d
a
t
a
,
”
i
n
H
EALT
H
I
N
F
2
0
2
1
-
1
4
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
H
e
a
l
t
h
I
n
f
o
rm
a
t
i
c
s;
P
a
rt
o
f
t
h
e
1
4
t
h
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
d
i
c
a
l
En
g
i
n
e
e
r
i
n
g
S
y
s
t
e
m
s
a
n
d
T
e
c
h
n
o
l
o
g
i
e
s,
BI
O
S
T
EC
2
0
2
1
,
2
0
2
1
,
p
p
.
6
4
5
–
650
,
d
o
i
:
1
0
.
5
2
2
0
/
0
0
1
0
3
2
5
9
0
6
4
5
0
6
5
0
.
[
4
7
]
A
.
K
h
A
b
o
o
a
n
d
L.
M
.
I
b
r
a
h
i
m,
“
H
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
a
h
y
b
r
i
d
C
N
N
-
LSTM
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
W
e
b
o
l
o
g
y
,
v
o
l
.
1
9
,
n
o
.
1
,
2
0
2
2
.
[
4
8
]
D
.
G
a
u
r
a
n
d
S
.
K
.
D
u
b
e
y
,
“
D
e
v
e
l
o
p
men
t
o
f
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
m
o
d
e
l
u
si
n
g
LS
TM
-
R
N
N
d
e
e
p
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m,”
J
o
u
rn
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
a
n
d
O
r
g
a
n
i
za
t
i
o
n
a
l
S
c
i
e
n
c
e
s
,
v
o
l
.
4
6
,
n
o
.
2
,
p
p
.
2
7
7
–
2
9
1
,
2
0
2
2
,
d
o
i
:
1
0
.
3
1
3
4
1
/
j
i
o
s.
4
6
.
2
.
1
.
[
4
9
]
S
.
G
u
p
t
a
,
“
D
e
e
p
l
e
a
r
n
i
n
g
b
a
s
e
d
h
u
man
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
(
H
A
R
)
u
s
i
n
g
w
e
a
r
a
b
l
e
se
n
s
o
r
d
a
t
a
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
M
a
n
a
g
e
m
e
n
t
D
a
t
a
I
n
si
g
h
t
s
,
v
o
l
.
1
,
n
o
.
2
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
j
i
m
e
i
.
2
0
2
1
.
1
0
0
0
4
6
.
[
5
0
]
M
.
A
.
K
h
a
t
u
n
e
t
a
l
.
,
“
D
e
e
p
C
N
N
-
LS
TM
w
i
t
h
s
e
l
f
-
a
t
t
e
n
t
i
o
n
mo
d
e
l
f
o
r
h
u
man
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
w
e
a
r
a
b
l
e
se
n
so
r
,
”
I
EE
E
J
o
u
rn
a
l
o
f
T
r
a
n
sl
a
t
i
o
n
a
l
E
n
g
i
n
e
e
r
i
n
g
i
n
H
e
a
l
t
h
a
n
d
M
e
d
i
c
i
n
e
,
v
o
l
.
1
0
,
p
p
.
1
–
1
6
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
J
TEH
M
.
2
0
2
2
.
3
1
7
7
7
1
0
.
[
5
1
]
R
.
R
a
j
a
n
d
A
.
K
o
s
,
“
A
n
i
m
p
r
o
v
e
d
h
u
m
a
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
t
e
c
h
n
i
q
u
e
b
a
s
e
d
o
n
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
S
c
i
e
n
t
i
f
i
c
Re
p
o
r
t
s
,
v
o
l
.
1
3
,
n
o
.
1
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
0
2
3
-
4
9
7
3
9
-
1.
[
5
2
]
S
.
I
r
f
a
n
,
N
.
A
n
j
u
m,
N
.
M
a
s
o
o
d
,
A
.
S
.
K
h
a
t
t
a
k
,
a
n
d
N
.
R
a
m
z
a
n
,
“
A
n
o
v
e
l
h
y
b
r
i
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
f
o
r
h
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
b
a
se
d
o
n
t
r
a
n
s
i
t
i
o
n
a
l
a
c
t
i
v
i
t
i
e
s,
”
S
e
n
so
r
s
,
v
o
l
.
2
1
,
n
o
.
2
4
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
1
2
4
8
2
2
7
.
[
5
3
]
X
.
Y
i
n
,
Z.
Li
u
,
D
.
L
i
u
,
a
n
d
X
.
R
e
n
,
“
A
n
o
v
e
l
C
N
N
-
b
a
se
d
B
i
-
LST
M
p
a
r
a
l
l
e
l
m
o
d
e
l
w
i
t
h
a
t
t
e
n
t
i
o
n
m
e
c
h
a
n
i
sm
f
o
r
h
u
m
a
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
w
i
t
h
n
o
i
s
y
d
a
t
a
,
”
S
c
i
e
n
t
i
f
i
c
Re
p
o
rt
s
,
v
o
l
.
1
2
,
n
o
.
1
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
0
2
2
-
1
1
8
8
0
-
8.
[
5
4
]
W
.
Z
.
Te
e
,
R
.
D
a
v
e
,
J.
S
e
l
i
y
a
,
a
n
d
M
.
V
a
n
a
mal
a
,
“
A
c
l
o
s
e
l
o
o
k
i
n
t
o
h
u
m
a
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
m
o
d
e
l
s
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
i
n
2
0
2
2
3
r
d
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
i
n
g
,
N
e
t
w
o
rks
a
n
d
I
n
t
e
rn
e
t
o
f
T
h
i
n
g
s,
C
N
I
O
T
2
0
2
2
,
2
0
2
2
,
p
p
.
2
0
1
–
2
0
6
,
d
o
i
:
1
0
.
1
1
0
9
/
C
N
I
O
T5
5
8
6
2
.
2
0
2
2
.
0
0
0
4
3
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Ba
k
k
a
la
S
a
n
th
a
K
u
m
a
r
re
c
e
iv
e
d
M
.
Tec
h
.
in
Co
m
p
u
t
e
r
S
c
ien
c
e
a
n
d
En
g
i
n
e
e
rin
g
fro
m
JN
TU
Un
i
v
e
rsity
,
Hy
d
e
ra
b
a
d
.
He
is
n
o
w
w
o
rk
i
n
g
a
s
a
n
a
ss
istan
t
p
r
o
fe
ss
o
r
in
t
h
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
i
n
e
e
rin
g
a
t
K
L
De
e
m
e
d
to
b
e
Un
i
v
e
rsity
,
Va
d
d
e
sw
a
ra
m
,
Vijay
a
wa
d
a
,
An
d
h
ra
P
ra
d
e
sh
.
He
is
a
h
i
g
h
l
y
a
c
c
o
m
p
li
sh
e
d
re
se
a
rc
h
e
r
with
a
stro
n
g
trac
k
re
c
o
r
d
o
f
p
u
b
li
c
a
ti
o
n
s
a
n
d
p
a
ten
ts.
He
is
a
lso
a
g
lo
b
a
ll
y
c
e
rti
fied
train
e
r
f
o
r
Az
u
re
a
n
d
AWS
.
He
h
a
s
1
4
y
e
a
rs
o
f
a
c
a
d
e
m
ic
e
x
p
e
rien
c
e
.
His
r
e
se
a
rc
h
in
tere
sts
li
e
i
n
t
h
e
a
re
a
s
o
f
ima
g
e
p
r
o
c
e
ss
in
g
,
c
lo
u
d
c
o
m
p
u
ti
n
g
,
a
n
d
m
a
c
h
in
e
lea
rn
i
n
g
.
He
is
c
u
rre
n
tl
y
p
u
rs
u
in
g
a
P
h
.
D
.
in
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
in
e
e
rin
g
a
t
K
o
n
e
ru
Lak
sh
m
a
ia
h
Ed
u
c
a
ti
o
n
F
o
u
n
d
a
ti
o
n
,
Va
d
d
e
sw
a
ra
m
,
In
d
ia
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
sa
n
th
a
k
u
m
a
r.
b
a
k
k
a
la@
g
m
a
il
.
c
o
m
.
Dr
.
R
.
S
h
a
n
k
a
r
re
c
e
iv
e
d
B.
E
.
i
n
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
fr
o
m
t
h
e
Un
iv
e
rsity
o
f
M
a
d
ra
s,
Ch
e
n
n
a
i,
M
.
Tec
h
.
in
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
fro
m
A
n
n
a
Un
iv
e
rsity
,
C
h
e
n
n
a
i
,
a
n
d
P
h
.
D
.
i
n
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
fro
m
M
a
n
o
n
m
a
n
iam
S
u
n
d
a
ra
n
a
r
Un
iv
e
rsit
y
,
Ti
r
u
n
e
lv
e
li
.
He
is
n
o
w
wo
rk
i
n
g
a
s
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
in
e
e
rin
g
a
t
K
o
n
e
r
u
Lak
s
h
m
a
iah
Ed
u
c
a
ti
o
n
F
o
u
n
d
a
ti
o
n
,
Va
d
d
e
sw
a
ra
m
,
In
d
ia
.
He
h
a
s
2
4
y
e
a
rs
o
f
a
c
a
d
e
m
ic
e
x
p
e
rien
c
e
.
He
h
a
s
p
u
b
li
sh
e
d
h
is
re
se
a
rc
h
wo
rk
i
n
se
v
e
ra
l
r
e
p
u
ted
j
o
u
r
n
a
ls
a
n
d
c
o
n
fe
re
n
c
e
s
in
t
h
e
field
s
o
f
m
o
b
il
e
c
o
m
p
u
ti
n
g
,
c
lo
u
d
c
o
m
p
u
ti
n
g
,
a
n
d
a
rti
ficia
l
i
n
telli
g
e
n
c
e
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
sh
a
n
k
a
rra
jen
d
ra
n
7
5
@k
l
u
n
i
v
e
rsity
.
in
.
Evaluation Warning : The document was created with Spire.PDF for Python.