I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
15
,
No
.
4
,
A
u
g
u
s
t
20
25
,
p
p
.
3
7
6
9
~
3
7
7
8
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
4
.
pp
3
7
6
9
-
3
7
7
8
3769
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Indo
nesia
n speec
h emo
tion re
co
g
nition: f
ea
ture
extr
a
ction a
nd
neura
l net
wo
rk a
ppro
a
ches
I
zz
a
Nur
Af
if
a
h
1
,
T
ri
B
u
di S
a
nto
s
o
2
,
T
it
o
n Dut
o
no
3
1
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
c
s
a
n
d
C
o
m
p
u
t
e
r
En
g
i
n
e
e
r
i
n
g
,
P
o
l
i
t
e
k
n
i
k
El
e
k
t
r
o
n
i
k
a
N
e
g
e
r
i
S
u
r
a
b
a
y
a
,
S
u
r
a
b
a
y
a
,
I
n
d
o
n
e
s
i
a
2
D
e
p
a
r
t
me
n
t
o
f
C
r
e
a
t
i
v
e
M
u
l
t
i
me
d
i
a
Te
c
h
n
o
l
o
g
y
,
P
o
l
i
t
e
k
n
i
k
El
e
k
t
r
o
n
i
k
a
N
e
g
e
r
i
S
u
r
a
b
a
y
a
,
S
u
r
a
b
a
y
a
,
I
n
d
o
n
e
s
i
a
3
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
P
o
l
i
t
e
k
n
i
k
E
l
e
k
t
r
o
n
i
k
a
N
e
g
e
r
i
S
u
r
a
b
a
y
a
,
S
u
r
a
b
a
y
a
,
I
n
d
o
n
e
si
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
3
1
,
2
0
2
4
R
ev
is
ed
Ma
r
2
6
,
2
0
2
5
Acc
ep
ted
Ma
y
2
4
,
2
0
2
5
Th
is
stu
d
y
e
x
p
lo
re
d
th
e
c
h
a
ll
e
n
g
e
s
o
f
e
m
o
ti
o
n
re
c
o
g
n
it
i
o
n
in
I
n
d
o
n
e
sia
n
sp
e
e
c
h
u
sin
g
d
e
e
p
lea
rn
i
n
g
tec
h
n
iq
u
e
s,
a
d
d
re
ss
in
g
th
e
c
o
m
p
lex
n
u
a
n
c
e
s
o
f
e
m
o
ti
o
n
a
l
e
x
p
re
ss
io
n
i
n
sp
o
k
e
n
lan
g
u
a
g
e
th
a
t
p
o
se
d
sig
n
ifi
c
a
n
t
d
iffi
c
u
lt
ies
fo
r
a
u
to
m
a
ti
c
re
c
o
g
n
it
io
n
sy
ste
m
s.
Th
e
re
se
a
rc
h
fo
c
u
se
d
o
n
th
e
a
p
p
li
c
a
ti
o
n
o
f
fe
a
tu
re
e
x
trac
ti
o
n
m
e
th
o
d
s
a
n
d
th
e
imp
lem
e
n
tatio
n
o
f
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
two
rk
s
(CNN
)
a
n
d
a
h
y
b
ri
d
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
s
-
l
o
n
g
sh
o
rt
-
term
m
e
m
o
ry
(CNN
-
L
S
T
M
)
m
o
d
e
l
to
id
e
n
ti
fy
e
m
o
ti
o
n
a
l
sta
tes
fro
m
sp
e
e
c
h
d
a
ta.
B
y
a
n
a
ly
z
in
g
k
e
y
fe
a
tu
re
s
o
f
s
p
e
e
c
h
sig
n
a
ls,
in
c
lu
d
in
g
m
e
l
fre
q
u
e
n
c
y
c
e
p
stra
l
c
o
e
fficie
n
t
(M
F
CC)
,
z
e
ro
c
ro
ss
in
g
ra
te
(ZCR),
ro
o
t
m
e
a
n
sq
u
a
re
e
n
e
rg
y
(RM
S
E)
,
p
it
c
h
,
a
n
d
sp
e
c
tral
c
e
n
tr
o
id
,
t
h
e
stu
d
y
e
v
a
lu
a
ted
t
h
e
m
o
d
e
ls’
a
b
il
it
y
to
c
a
p
tu
re
b
o
th
sp
a
ti
a
l
a
n
d
tem
p
o
ra
l
p
a
tt
e
r
n
s
i
n
th
e
d
a
ta.
Tes
ti
n
g
wa
s
c
o
n
d
u
c
ted
u
sin
g
a
n
I
n
d
o
n
e
sia
n
d
a
tas
e
t
c
o
m
p
risi
n
g
2
0
0
sa
m
p
les
.
Th
e
CNN
m
o
d
e
l,
u
ti
li
z
in
g
fo
u
r
f
e
a
tu
re
s
(M
F
CC,
ZCR,
RM
S
E
,
a
n
d
p
it
c
h
),
a
n
d
th
e
CNN
-
L
S
TM
m
o
d
e
l,
wh
i
c
h
u
se
d
t
h
re
e
fe
a
tu
re
s
(M
F
CC,
ZCR,
a
n
d
RM
S
E),
b
o
th
a
c
h
ie
v
e
d
a
n
e
m
o
ti
o
n
c
las
sifica
ti
o
n
a
c
c
u
ra
c
y
o
f
a
p
p
ro
x
ima
tely
8
8
%
.
T
h
e
re
su
l
t
sh
o
we
d
th
a
t
th
e
CNN
-
LS
TM
m
o
d
e
l
a
c
h
iev
e
d
c
o
m
p
a
ra
b
le
p
e
rfo
rm
a
n
c
e
with
a
sim
p
ler
fe
a
t
u
re
se
t
c
o
m
p
a
re
d
to
th
e
CNN
m
o
d
e
l.
T
h
is
h
ig
h
li
g
h
ted
th
e
sig
n
ifi
c
a
n
c
e
o
f
c
h
o
o
si
n
g
th
e
a
p
p
ro
p
riate
tec
h
n
iq
u
e
s
in
fe
a
tu
re
e
x
trac
ti
o
n
a
n
d
c
las
sifica
ti
o
n
t
o
e
n
h
a
n
c
e
th
e
a
c
c
u
ra
c
y
o
f
i
d
e
n
ti
f
y
in
g
e
m
o
ti
o
n
s fr
o
m
sp
e
e
c
h
d
a
ta w
h
il
e
a
lso
m
a
n
a
g
in
g
c
o
m
p
u
tatio
n
a
l
c
o
m
p
lex
it
y
.
K
ey
w
o
r
d
s
:
C
o
h
en
’
s
Kap
p
a
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
L
o
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
Mel
-
f
r
eq
u
e
n
cy
c
ep
s
tr
al
co
ef
f
icien
ts
Sp
ee
ch
em
o
tio
n
r
ec
o
g
n
itio
n
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
T
r
i Bu
d
i San
to
s
o
Dep
ar
tm
en
t o
f
C
r
ea
tiv
e
Mu
ltime
d
ia
T
ec
h
n
o
lo
g
y
,
Po
litek
n
ik
E
lek
tr
o
n
ik
a
Neg
er
i Su
r
a
b
ay
a
J
alan
R
ay
a
I
T
S,
Kep
u
tih
,
Su
k
o
lilo
,
Su
r
ab
ay
a
,
E
ast
Ja
v
a
6
0
1
1
1
,
I
n
d
o
n
esia
E
m
ail: tr
ib
u
d
i@
p
en
s
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
Sp
ee
ch
co
m
m
u
n
icatio
n
s
er
v
e
s
as
th
e
s
im
p
lest
an
d
ef
f
ec
tiv
e
ap
p
r
o
ac
h
th
at
p
e
o
p
le
h
a
v
e
i
n
o
r
d
e
r
to
co
m
m
u
n
icate
in
f
o
r
m
atio
n
.
T
h
e
im
p
o
r
tan
ce
o
f
s
p
ee
ch
b
e
co
m
es
ev
id
en
t
wh
en
alter
n
at
iv
e
co
m
m
u
n
icatio
n
m
eth
o
d
s
,
s
u
ch
as
tex
t
m
ess
ag
es
o
r
em
ails
,
ar
e
co
m
m
o
n
ly
u
s
ed
b
u
t
ca
n
ea
s
ily
b
e
m
is
in
t
er
p
r
eted
.
W
h
en
we
attem
p
t
to
ex
p
r
ess
em
o
tio
n
s
in
wr
itin
g
,
em
o
jis
o
f
te
n
b
ec
o
m
e
n
ec
ess
ar
y
aid
s
in
tex
t
m
e
s
s
ag
in
g
[
1
]
.
T
h
u
s
,
s
p
ee
ch
is
th
e
m
o
s
t
ef
f
ec
tiv
e
m
eth
o
d
to
c
o
m
m
u
n
icate
in
h
u
m
an
life
,
as
it
ca
r
r
ies
a
wea
lth
o
f
in
f
o
r
m
atio
n
th
r
o
u
g
h
b
o
th
lin
g
u
is
tic
an
d
p
a
r
alin
g
u
is
tic
elem
en
ts
[
2
]
.
T
h
e
ad
v
an
ce
m
en
t
o
f
in
f
o
r
m
at
io
n
an
d
co
m
m
u
n
icatio
n
tech
n
o
lo
g
y
(
I
C
T
)
tech
n
o
lo
g
y
h
as
o
p
en
ed
u
p
n
ew
p
o
s
s
ib
ilit
ies
f
o
r
h
o
w
h
u
m
an
s
in
ter
ac
t
with
co
m
p
u
ter
s
.
Giv
en
th
at
u
n
d
e
r
s
tan
d
in
g
em
o
tio
n
al
s
tates
en
h
an
ce
s
in
ter
p
e
r
s
o
n
al
co
m
p
r
eh
en
s
io
n
,
th
e
r
e
is
a
n
ee
d
to
in
teg
r
ate
th
is
co
n
c
ep
t
in
t
o
co
m
p
u
ter
s
y
s
tem
s
.
T
h
is
id
ea
in
s
p
ir
ed
th
e
estab
lis
h
m
e
n
t
o
f
s
p
ee
ch
em
o
tio
n
r
ec
o
g
n
itio
n
(
SER),
a
f
ield
f
o
cu
s
ed
o
n
id
e
n
tify
in
g
an
d
in
ter
p
r
etin
g
em
o
tio
n
al
s
tates
co
n
v
ey
e
d
th
r
o
u
g
h
s
p
ee
c
h
.
Ma
n
y
s
tu
d
ies
h
av
e
b
ee
n
co
n
d
u
ct
ed
to
ex
p
lo
r
e
SER,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
7
6
9
-
3778
3770
b
u
t
th
e
to
p
ic
s
t
ill
p
r
esen
ts
s
i
g
n
if
ican
t
ch
allen
g
es.
SER
tech
n
o
lo
g
y
h
as
p
o
ten
tial
u
s
es
ac
r
o
s
s
s
ev
er
al
f
ield
s
,
in
clu
d
in
g
h
ea
lth
ca
r
e,
ca
ll
ce
n
ter
s
,
an
d
ed
u
ca
tio
n
[
3
]
–
[
5
]
.
I
n
h
ea
lth
ca
r
e,
it
ca
n
h
elp
i
n
th
e
d
iag
n
o
s
is
o
f
p
s
y
ch
o
lo
g
ica
l
p
r
o
b
lem
s
lik
e
d
ep
r
ess
io
n
,
au
tis
m
,
an
d
o
th
er
m
en
tal
d
is
o
r
d
er
s
.
I
n
ca
ll
ce
n
te
r
s
,
it
h
elp
s
m
ea
s
u
r
e
cu
s
to
m
er
s
atis
f
ac
tio
n
.
I
n
ed
u
c
atio
n
,
p
ar
ticu
la
r
ly
in
d
is
tan
ce
lear
n
in
g
,
it
ca
n
en
h
an
ce
th
e
le
ar
n
in
g
e
x
p
er
ien
ce
.
Desp
ite
its
s
ig
n
if
ican
t
p
o
ten
ti
al,
ch
allen
g
es
r
em
ain
,
s
u
ch
a
s
th
e
lack
o
f
d
iv
er
s
e
d
atasets
,
ch
o
o
s
in
g
th
e
r
ig
h
t
f
ea
tu
r
es,
an
d
t
h
e
ch
o
ice
o
f
ef
f
ec
tiv
e
in
tellig
en
t r
ec
o
g
n
itio
n
t
ec
h
n
iq
u
es
[
6
]
–
[
8
]
.
T
h
e
m
ajo
r
ity
o
f
SER
r
esear
c
h
h
as
f
o
cu
s
ed
o
n
lan
g
u
ag
es
with
ab
u
n
d
an
t
r
eso
u
r
ce
s
a
n
d
wid
esp
r
ea
d
u
s
e,
s
u
ch
as
E
n
g
lis
h
o
r
Ger
m
an
[
9
]
–
[
1
1
]
.
Alth
o
u
g
h
th
es
e
s
tu
d
ies
h
av
e
d
ee
p
en
ed
o
u
r
u
n
d
er
s
tan
d
in
g
o
f
d
etec
tin
g
em
o
tio
n
s
in
s
p
ee
ch
,
th
er
e
r
em
ain
s
a
co
n
s
id
er
ab
le
g
a
p
in
ex
p
lo
r
i
n
g
r
eso
u
r
ce
-
lim
ited
lan
g
u
ag
es
lik
e
I
n
d
o
n
esian
.
I
n
r
ec
en
t
y
ea
r
s
,
r
esear
ch
o
n
em
o
tio
n
d
etec
tio
n
in
I
n
d
o
n
esian
s
p
ee
ch
h
as
b
eg
u
n
to
em
er
g
e,
co
v
er
in
g
ar
ea
s
s
u
ch
as
em
o
ti
o
n
d
etec
tio
n
i
n
f
ilm
s
[
1
2
]
,
r
ec
o
g
n
itio
n
u
s
in
g
ac
o
u
s
tic
an
d
lex
ical
f
ea
tu
r
es
[
1
3
]
,
an
d
au
to
m
atic
em
o
tio
n
r
ec
o
g
n
itio
n
[
1
4
]
.
Desp
ite
I
n
d
o
n
e
s
ian
b
ein
g
s
p
o
k
e
n
b
y
o
v
er
2
0
0
m
illi
o
n
p
eo
p
le,
r
esear
ch
atten
tio
n
in
SER
r
e
m
ain
s
lim
ited
.
T
h
e
s
ca
r
city
o
f
co
r
p
o
r
a
an
d
s
tan
d
a
r
d
ized
d
a
tab
ases
h
am
p
er
s
th
e
p
r
o
g
r
ess
o
f
SER
r
esear
ch
i
n
I
n
d
o
n
esian
.
C
r
o
s
s
-
lin
g
u
al
e
m
o
t
io
n
r
ec
o
g
n
iti
o
n
e
x
p
er
im
e
n
ts
h
av
e
b
ee
n
c
o
n
d
u
cted
d
u
e
to
th
ese
lim
itatio
n
s
[
1
5
]
.
I
n
s
im
p
le
te
r
m
s
,
SER
co
n
s
is
ts
o
f
two
p
r
im
ar
y
c
o
m
p
o
n
en
ts
:
f
ea
tu
r
e
ex
tr
ac
tio
n
a
n
d
class
if
ic
atio
n
[
1
6
]
,
[
1
7
]
.
Featu
r
e
ex
tr
ac
tio
n
in
v
o
l
v
es
id
en
tify
in
g
ch
ar
ac
ter
is
tics
r
elate
d
to
em
o
tio
n
with
in
s
p
ee
ch
s
ig
n
als
[
1
8
]
.
T
h
e
g
o
al
is
to
ex
t
r
ac
t
em
o
tio
n
al
in
f
o
r
m
atio
n
f
r
o
m
s
p
o
k
en
la
n
g
u
ag
e
b
y
co
n
v
er
tin
g
th
e
r
aw
s
p
ee
ch
s
ig
n
als
in
to
r
elev
an
t
f
ea
tu
r
e
s
ets.
SER
f
r
am
ewo
r
k
s
d
iv
id
e
ch
ar
ac
ter
is
tic
s
in
to
f
o
u
r
ca
teg
o
r
ies:
p
r
o
s
o
d
i
c
f
ea
tu
r
es,
s
p
ec
tr
al
f
ea
tu
r
es,
v
o
ice
q
u
ality
f
ea
tu
r
e
s
,
an
d
T
ea
g
er
en
e
r
g
y
o
p
er
ato
r
(
T
E
O)
-
b
ased
f
ea
tu
r
es
[
2
]
.
T
h
e
ch
allen
g
e
lies
in
ch
o
o
s
in
g
th
e
m
o
s
t
es
s
en
tial
f
ea
tu
r
es
th
at
ar
e
ab
le
to
d
if
f
er
en
tiate
b
etwe
en
d
if
f
e
r
en
t
e
m
o
tio
n
s
[
1
9
]
.
Me
l
-
f
r
eq
u
e
n
cy
ce
p
s
tr
al
co
e
f
f
icien
t
s
(
MFC
C
)
is
ef
f
ec
tiv
e
in
ca
p
tu
r
in
g
im
p
o
r
tan
t
s
p
ec
tr
al
c
h
ar
a
cter
is
tics
b
ased
o
n
h
u
m
an
p
e
r
ce
p
tio
n
o
f
f
r
eq
u
en
c
y
,
m
ak
in
g
it
r
elev
an
t
f
o
r
d
etec
tin
g
s
p
ec
t
r
u
m
ch
an
g
es
ass
o
ciate
d
with
em
o
tio
n
s
.
ze
r
o
cr
o
s
s
in
g
r
ate
m
ea
s
u
r
es
h
o
w
f
r
eq
u
en
tly
th
e
v
alu
e
o
f
t
h
e
au
d
io
s
ig
n
al
ch
an
g
es
f
r
o
m
ab
o
v
e
to
b
elo
w
ze
r
o
,
p
r
o
v
id
i
n
g
in
f
o
r
m
atio
n
ab
o
u
t
t
h
e
tem
p
o
r
al
asp
ec
ts
th
at
m
ay
ch
an
g
e
with
em
o
tio
n
.
R
o
o
t
m
ea
n
s
q
u
ar
e
en
er
g
y
(
R
MSE
)
m
ea
s
u
r
es
th
e
a
v
er
ag
e
en
er
g
y
o
f
th
e
s
p
ee
c
h
s
ig
n
al,
wh
ich
ca
n
r
ef
lect
v
a
r
y
in
g
s
o
u
n
d
in
ten
s
ity
lev
els
ass
o
ciate
d
with
em
o
tio
n
s
.
Pit
ch
m
ea
s
u
r
es
th
e
f
u
n
d
am
e
n
tal
f
r
eq
u
e
n
cy
o
f
th
e
s
p
ee
ch
,
wh
e
r
e
ch
an
g
es
in
p
itch
ar
e
o
f
te
n
lin
k
ed
to
e
m
o
tio
n
al
v
ar
iatio
n
.
Sp
ec
tr
al
C
en
tr
o
id
m
ea
s
u
r
es
th
e
av
e
r
ag
e
f
r
e
q
u
en
c
y
lo
ca
tio
n
with
in
t
h
e
s
p
ec
tr
u
m
,
r
ef
lectin
g
th
e
b
r
i
g
h
t
n
ess
o
f
th
e
s
o
u
n
d
,
wh
ich
m
ay
ch
an
g
e
with
d
if
f
er
e
n
t
en
er
g
y
d
is
tr
ib
u
tio
n
s
d
u
e
to
em
o
tio
n
s
[
2
0
]
–
[
2
4
]
.
C
las
s
if
icatio
n
is
th
e
s
ec
o
n
d
cr
u
cial
s
tep
in
SER.
I
t
in
v
o
lv
es a
p
p
ly
in
g
m
ac
h
in
e
lear
n
in
g
m
o
d
els
to
th
e
ex
tr
ac
ted
f
ea
tu
r
es
to
id
e
n
tify
th
e
em
o
tio
n
s
ex
p
r
ess
ed
in
s
p
ee
ch
.
T
h
e
r
e
ar
e
two
m
ain
a
p
p
r
o
ac
h
es
to
SER
class
if
icatio
n
:
co
n
v
en
tio
n
al
c
lass
if
ier
s
an
d
d
ee
p
lear
n
in
g
class
if
ier
s
.
R
ec
en
t
d
ev
elo
p
m
en
ts
in
d
icate
th
at
p
r
o
b
lem
s
in
SER
ar
e
b
ein
g
ad
d
r
ess
ed
with
m
o
r
e
em
p
h
asis
o
n
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es,
esp
ec
ially
d
ee
p
lear
n
in
g
ap
p
r
o
ac
h
es.
Dee
p
l
ea
r
n
in
g
m
et
h
o
d
s
h
a
v
e
d
em
o
n
s
tr
ated
s
ig
n
if
ican
t
im
p
r
o
v
e
m
en
ts
in
em
o
tio
n
r
ec
o
g
n
itio
n
,
o
f
f
er
in
g
ad
v
a
n
tag
es
s
u
ch
as
s
ca
lab
ilit
y
,
p
ar
a
m
eter
tu
n
in
g
,
an
d
cu
s
to
m
iza
b
le
f
u
n
ctio
n
s
[
2
]
.
Sev
er
al
r
esear
ch
er
s
h
av
e
ex
p
l
o
r
ed
v
a
r
i
o
u
s
n
e
u
r
al
n
etwo
r
k
m
eth
o
d
o
l
o
g
ies,
in
clu
d
i
n
g
ar
tif
icial
n
eu
r
al
n
etwo
r
k
s
(
ANN)
,
co
n
v
o
lu
tio
n
al
n
e
u
r
al
n
etwo
r
k
s
(
C
NN)
,
d
ee
p
n
eu
r
al
n
etwo
r
k
s
(
DNN)
,
r
ec
u
r
r
en
t
n
eu
r
al
n
etwo
r
k
s
(
R
NN)
,
an
d
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
[
2
5
]
–
[
2
8
]
.
C
NN
an
d
L
STM
a
r
e
in
c
r
ea
s
in
g
ly
r
ec
o
g
n
ized
f
o
r
SER task
s
b
ec
au
s
e
th
ey
ef
f
ec
tiv
ely
ca
p
tu
r
e
tem
p
o
r
al
d
ep
e
n
d
en
cies a
n
d
s
p
atial
p
atter
n
s
in
s
eq
u
en
tial d
ata.
B
ased
o
n
th
e
ch
allen
g
es
f
ac
e
d
in
SER
r
esear
ch
f
o
r
th
e
I
n
d
o
n
esian
lan
g
u
ag
e,
th
is
s
tu
d
y
aim
s
to
ad
d
r
ess
th
e
g
a
p
b
y
p
r
o
v
id
in
g
a
co
m
p
r
eh
e
n
s
iv
e
co
m
p
ar
is
o
n
o
f
s
p
ee
ch
e
m
o
tio
n
r
ec
o
g
n
itio
n
s
y
s
tem
s
f
o
r
d
eter
m
in
in
g
em
o
tio
n
al
s
tates.
I
t
ev
alu
ated
th
e
c
o
n
s
is
ten
cy
a
n
d
r
eliab
ilit
y
o
f
em
o
tio
n
lab
el
in
g
u
s
in
g
C
o
h
en
’
s
k
ap
p
a,
ap
p
lied
s
ev
er
al
f
ea
t
u
r
e
ex
tr
ac
tio
n
ap
p
r
o
ac
h
es
in
cl
u
d
in
g
m
el
f
r
eq
u
en
cy
ce
p
s
tr
al
co
ef
f
icien
t
(
MFC
C
)
,
ze
r
o
cr
o
s
s
in
g
r
ate
(
Z
C
R
)
,
r
o
o
t
m
ea
n
s
q
u
ar
e
en
er
g
y
(
R
MSE
)
,
p
itch
,
an
d
s
p
ec
tr
al
ce
n
tr
o
id
,
a
n
d
co
m
b
in
e
d
th
ese
f
ea
tu
r
es
with
class
if
icatio
n
tech
n
iq
u
es
th
at
u
s
ed
C
NN
an
d
L
STM
.
T
h
e
s
tr
u
ctu
r
e
o
f
th
is
p
ap
er
is
as
f
o
llo
ws:
T
h
e
s
tu
d
y
ch
r
o
n
o
lo
g
y
,
as
well
as
th
e
r
esear
ch
d
esig
n
,
m
e
th
o
d
o
lo
g
y
,
d
ataset
co
llectio
n
,
f
ea
tu
r
e
ex
tr
ac
tio
n
s
tr
ateg
ies,
an
d
class
if
icatio
n
alg
o
r
ith
m
s
,
a
r
e
c
o
v
er
ed
in
s
ec
t
io
n
2
.
T
h
e
r
esear
c
h
r
esu
lts
ap
p
ea
r
in
s
ec
tio
n
3
alo
n
g
with
a
c
o
m
p
r
e
h
en
s
iv
e
d
i
s
cu
s
s
io
n
,
wh
ile
s
ec
tio
n
4
p
r
o
v
id
es th
e
co
n
clu
s
io
n
.
2.
M
E
T
H
O
D
I
n
th
is
r
esear
ch
,
th
e
p
r
o
ce
s
s
o
f
r
ec
o
g
n
izin
g
em
o
tio
n
s
in
s
p
ee
ch
was o
r
g
an
ized
i
n
to
th
r
ee
m
ain
s
tag
es:
d
ata
co
llectio
n
,
f
ea
tu
r
e
ex
t
r
a
ctio
n
,
an
d
class
if
icatio
n
.
Du
r
in
g
d
ata
co
llectio
n
s
tag
e,
a
d
ataset
o
f
s
p
ee
ch
s
am
p
les
r
ep
r
esen
tin
g
v
ar
io
u
s
em
o
tio
n
al
s
tates
was
g
ath
er
ed
,
an
d
in
ter
-
r
ater
r
eliab
ilit
y
was
em
p
lo
y
ed
to
en
s
u
r
e
co
n
s
is
ten
t
em
o
tio
n
la
b
elin
g
ac
r
o
s
s
d
if
f
er
e
n
t
ev
alu
ato
r
s
.
On
ce
th
e
d
ataset
was
p
r
ep
ar
ed
,
f
ea
tu
r
e
ex
tr
ac
tio
n
was
ca
r
r
ied
o
u
t
to
id
en
tify
an
d
p
r
o
ce
s
s
k
ey
ch
ar
ac
ter
is
tics
o
f
a
s
p
ee
ch
s
ig
n
al.
T
h
ese
ex
tr
ac
ted
f
ea
tu
r
es
wer
e
th
en
u
s
ed
in
th
e
class
if
icat
io
n
s
tag
e
to
ac
cu
r
ately
ca
teg
o
r
ize
th
e
em
o
tio
n
al
s
tates
co
n
v
ey
ed
i
n
th
e
s
p
ee
ch
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
I
n
d
o
n
esia
n
s
p
ee
ch
em
o
tio
n
r
e
co
g
n
itio
n
:
fea
tu
r
e
ex
tr
a
ctio
n
a
n
d
n
eu
r
a
l
…
(
I
z
z
a
N
u
r
A
fifa
h
)
3771
2
.
1
.
Da
t
a
c
o
llect
io
n
T
h
e
au
d
io
d
ataset
f
o
r
th
is
r
esear
ch
co
n
s
is
ts
o
f
s
p
ee
ch
r
ec
o
r
d
in
g
s
in
I
n
d
o
n
esian
.
T
h
e
d
ig
ital
au
d
io
d
ata
was
s
to
r
ed
i
n
W
AV
f
i
le
f
o
r
m
at.
T
h
e
d
ataset
in
clu
d
es
r
ec
o
r
d
in
g
s
f
r
o
m
1
0
m
al
e
an
d
1
0
f
em
ale
p
ar
ticip
an
ts
,
ag
e
d
2
0
to
2
2
y
ea
r
s
.
E
ac
h
au
d
i
o
r
ec
o
r
d
in
g
last
s
b
etwe
en
o
n
e
to
th
r
ee
s
ec
o
n
d
s
,
with
ea
ch
p
ar
ticip
an
t c
o
n
tr
i
b
u
tin
g
f
o
u
r
r
ec
o
r
d
in
g
s
p
e
r
em
o
tio
n
.
No
t a
ll r
ec
o
r
d
in
g
s
,
h
o
wev
er
,
wer
e
s
u
itab
le
o
r
u
s
ab
le
d
u
e
to
f
ac
to
r
s
s
u
ch
as
p
o
o
r
au
d
i
o
q
u
ality
o
r
in
co
n
s
is
ten
cy
in
th
e
em
o
tio
n
al
ex
p
r
ess
io
n
.
A
to
t
al
o
f
5
0
au
d
io
f
iles
wer
e
u
s
ed
to
r
ep
r
esen
t
f
o
u
r
em
o
tio
n
al
e
x
p
r
ess
io
n
s
(
an
g
r
y
,
h
a
p
p
y
,
n
e
u
tr
al,
an
d
s
ad
[
1
2
]
)
,
r
esu
ltin
g
in
ap
p
r
o
x
im
ately
2
0
0
au
d
io
f
iles
in
t
o
tal.
T
h
e
d
ataset
co
n
tain
e
d
r
ec
o
r
d
in
g
s
with
s
am
p
lin
g
r
a
tes
th
at
v
ar
ie
d
f
r
o
m
4
4
.
1
to
4
8
k
Hz.
T
o
en
s
u
r
e
c
o
n
s
is
ten
cy
f
o
r
au
d
io
an
aly
s
is
,
all
f
iles
wer
e
r
esam
p
led
to
4
8
k
Hz,
p
r
eser
v
in
g
h
ig
h
au
d
io
q
u
ality
.
T
o
ev
alu
ate
t
h
e
co
n
s
is
ten
cy
an
d
r
eliab
ilit
y
o
f
em
o
tio
n
lab
elin
g
,
C
o
h
en
'
s
Kap
p
a
an
aly
s
is
was
co
n
d
u
cte
d
[
2
9
]
.
T
h
is
s
tatis
tical
m
eth
o
d
p
r
o
v
id
es
d
ee
p
er
in
s
ig
h
t
in
to
th
e
ag
r
ee
m
e
n
t
lev
els
b
etwe
en
an
n
o
tato
r
s
,
u
s
in
g
a
s
ca
le
f
r
o
m
-
1
to
1
.
A
v
alu
e
o
f
-
1
r
ep
r
esen
ts
co
m
p
l
ete
d
is
ag
r
ee
m
en
t,
0
in
d
icate
s
r
an
d
o
m
ag
r
ee
m
en
t,
an
d
1
r
ef
lects p
er
f
ec
t a
g
r
ee
m
e
n
t
[
3
0
]
.
T
ab
le
1
co
n
tain
s
C
o
h
e
n
'
s
Kap
p
a
v
alu
es a
n
d
ass
o
ciate
d
in
ter
p
r
etatio
n
s
.
T
ab
le
1
.
I
n
ter
p
r
etatio
n
o
f
C
o
h
en
’
s
Kap
p
a
C
o
h
e
n
’
s
K
a
p
p
a
S
t
a
t
i
s
t
i
c
S
t
r
e
n
g
t
h
o
f
a
g
r
e
e
me
n
t
<
0
.
0
0
P
o
o
r
0
.
0
0
–
0
.
2
0
S
l
i
g
h
t
0
.
2
1
–
0
.
4
0
F
a
i
r
0
.
4
1
–
0
.
6
0
M
o
d
e
r
a
t
e
0
.
6
1
–
0
.
8
0
S
u
b
s
t
a
n
t
i
a
l
0
.
8
1
–
1
.
0
0
A
l
mo
s
t
P
e
r
f
e
c
t
2
.
2
.
F
e
a
t
ure
ex
t
r
a
ct
io
n
On
e
o
f
th
e
m
o
s
t
im
p
o
r
tan
t
s
tep
s
in
p
r
o
ce
s
s
in
g
s
p
ee
ch
d
a
ta
f
o
r
em
o
tio
n
class
if
icatio
n
is
f
ea
tu
r
e
ex
tr
ac
tio
n
.
T
h
is
p
r
o
ce
s
s
in
v
o
lv
es
tr
an
s
f
o
r
m
in
g
r
aw
au
d
io
s
ig
n
als
in
to
r
elev
an
t
f
ea
tu
r
es
f
o
r
a
n
aly
s
is
.
T
h
e
tech
n
iq
u
es e
m
p
lo
y
ed
in
t
h
is
s
tu
d
y
wer
e
c
h
o
s
en
to
c
o
lle
ct
b
o
t
h
tem
p
o
r
al
a
n
d
s
p
ec
tr
al
ch
a
r
ac
ter
is
tics
o
f
s
p
ee
ch
.
T
h
e
tech
n
iq
u
es u
s
ed
in
th
is
s
tu
d
y
f
o
r
ex
tr
ac
tin
g
f
ea
t
u
r
es f
r
o
m
s
p
ee
ch
in
clu
d
e
MFC
C
,
Z
C
R
,
R
M
SE
,
p
itch
,
an
d
s
p
ec
tr
al
ce
n
tr
o
id
.
T
h
e
f
ea
tu
r
es we
r
e
ex
tr
ac
ted
a
n
d
ca
lcu
lated
in
d
iv
id
u
ally
f
r
o
m
ea
ch
a
u
d
io
.
2
.
2
.
1
.
M
el
f
re
qu
ency
ce
ps
t
ra
l c
o
ef
f
icient
T
h
e
f
ir
s
t
f
ea
tu
r
e
ex
tr
ac
tio
n
m
eth
o
d
em
p
lo
y
ed
was
MFC
C
.
MFC
C
i
s
in
s
p
ir
ed
b
y
th
e
way
th
e
h
u
m
an
ea
r
p
r
o
ce
s
s
es
s
o
u
n
d
[
3
1
]
,
[
3
2
]
.
T
h
ese
co
ef
f
icien
ts
f
o
cu
s
o
n
th
e
m
o
s
t
im
p
o
r
tan
t
asp
ec
ts
o
f
s
o
u
n
d
,
s
u
ch
as
th
e
s
h
ap
e
o
f
v
o
ca
l
f
o
r
m
an
ts
an
d
o
th
er
c
h
ar
ac
ter
is
tics
,
wh
ich
ar
e
ess
en
tial
f
o
r
task
s
lik
e
em
o
t
io
n
r
ec
o
g
n
itio
n
an
d
s
p
ee
ch
an
aly
s
is
[
3
3
]
.
B
y
em
p
h
asizin
g
f
r
eq
u
en
cies
th
at
ar
e
m
o
s
t
im
p
o
r
tan
t
f
o
r
h
o
w
h
u
m
an
s
h
ea
r
,
MFC
C
s
p
r
o
v
id
e
a
clea
r
r
e
p
r
esen
tatio
n
o
f
s
p
ee
ch
s
ig
n
als.
Fig
u
r
e
1
illu
s
tr
ates th
e
MFC
C
f
ea
tu
r
e
ex
tr
ac
t
io
n
p
r
o
ce
s
s
.
T
h
e
ex
tr
ac
tio
n
o
f
MFC
C
f
ea
t
u
r
es
f
r
o
m
s
p
ee
ch
d
ata
b
eg
an
with
p
r
e
-
em
p
h
asis
,
wh
ich
b
o
o
s
ted
th
e
h
ig
h
er
f
r
eq
u
en
cies
to
en
h
a
n
ce
clar
ity
.
Ne
x
t,
th
e
au
d
i
o
s
ig
n
al
Nex
t,
th
e
au
d
io
s
ig
n
al
was
s
eg
m
en
ted
in
to
s
m
all
f
r
am
es
o
f
2
5
m
s
,
with
a
5
0
%
o
v
er
lap
.
E
ac
h
f
r
am
e
was
th
en
p
r
o
ce
s
s
ed
u
s
in
g
a
Ham
m
in
g
win
d
o
w
to
m
in
im
ize
ed
g
e
e
f
f
ec
ts
b
ef
o
r
e
u
n
d
e
r
g
o
i
n
g
f
ast
Fo
u
r
ier
tr
a
n
s
f
o
r
m
(
FF
T
)
,
wh
ich
tr
an
s
f
o
r
m
ed
th
e
a
u
d
io
d
ata
f
r
o
m
th
e
tim
e
d
o
m
ain
in
to
th
e
f
r
eq
u
en
cy
d
o
m
ain
u
s
in
g
an
NFFT
s
ize
o
f
5
1
2
.
Af
ter
th
at,
a
Me
l
-
f
ilter
b
an
k
was
ap
p
lied
,
co
n
s
is
tin
g
o
f
4
0
f
ilter
s
s
p
ac
ed
ac
co
r
d
in
g
to
th
e
Me
l
s
ca
le
to
m
im
ic
th
e
h
u
m
an
ea
r
’
s
f
r
eq
u
en
cy
r
esp
o
n
s
e.
T
o
m
an
ag
e
th
e
wid
e
r
a
n
g
e
o
f
v
al
u
es,
lo
g
c
o
m
p
r
ess
io
n
was
ap
p
lied
,
co
m
p
r
ess
in
g
th
e
v
al
u
es
b
etwe
en
0
an
d
1
.
I
n
th
e
f
in
al
s
tep
,
th
e
d
is
cr
ete
co
s
in
e
tr
an
s
f
o
r
m
(
DC
T
)
was u
s
ed
o
n
th
e
lo
g
-
co
m
p
r
ess
ed
s
ig
n
al
t
o
d
er
iv
e
d
MFC
C
s
.
Fig
u
r
e
1
.
MFC
C
f
lo
wch
ar
t
2
.
2
.
2
.
Z
er
o
cr
o
s
s
ing
ra
t
e
T
h
e
s
ec
o
n
d
f
ea
tu
r
e
em
p
l
o
y
ed
was
th
e
Z
C
R
,
wh
ich
was
ca
l
cu
lated
s
ep
ar
ately
.
Z
C
R
was
d
er
iv
ed
b
y
ass
es
s
in
g
th
e
f
r
eq
u
en
cy
o
f
ze
r
o
-
cr
o
s
s
in
g
s
in
th
e
s
ig
n
al
ac
r
o
s
s
a
f
r
am
e.
T
h
e
p
r
o
ce
s
s
in
v
o
lv
ed
co
u
n
tin
g
ea
ch
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
7
6
9
-
3778
3772
in
s
tan
ce
wh
er
e
th
e
a
u
d
io
s
ig
n
al
s
h
if
ts
f
r
o
m
ab
o
v
e
ze
r
o
to
b
elo
w
ze
r
o
o
r
th
e
r
ev
er
s
e
with
i
n
a
f
r
a
m
e.
Z
C
R
is
d
ef
in
ed
as sh
o
wn
in
(
1
)
:
=
1
−
1
∑
1
(
[
]
∙
[
−
1
]
<
0
)
−
1
=
1
(
1
)
wh
er
e
r
ep
r
esen
ts
th
e
to
tal
s
a
m
p
le
co
u
n
t
with
in
th
e
f
r
am
e,
an
d
1
(
[
]
∙
[
−
1
]
<
0
)
r
ep
r
esen
ts
a
f
u
n
ctio
n
th
at
o
u
tp
u
ts
1
w
h
en
th
e
r
e
is
a
s
ig
n
ch
an
g
e
b
etwe
en
[
]
an
d
[
−
1
]
,
an
d
0
o
th
er
wis
e
.
Z
C
R
r
ef
lecte
d
th
e
r
ate
o
f
ch
a
n
g
e
in
th
e
s
ig
n
al,
w
h
ich
ca
n
b
e
in
d
icativ
e
o
f
em
o
tio
n
al
s
ta
tes.
Hig
h
er
Z
C
R
v
alu
es
ar
e
ass
o
ciate
d
with
m
o
r
e
ten
s
e
o
r
ag
itated
e
m
o
t
io
n
al
s
tates,
s
u
ch
as
a
n
g
r
y
o
r
h
ap
p
y
,
wh
er
e
r
a
p
id
ch
an
g
es
in
p
itch
o
r
to
n
e
o
cc
u
r
.
C
o
n
v
e
r
s
ely
,
lo
wer
Z
C
R
v
al
u
es
in
d
icate
ca
lm
e
r
em
o
tio
n
s
,
s
u
ch
as
n
eu
tr
al
o
r
s
ad
,
wh
er
e
th
e
s
p
ee
ch
is
m
o
r
e
s
tead
y
an
d
less
v
ar
iab
le
[
3
4
]
.
2
.
2
.
3
.
Ro
o
t
m
ea
n sq
ua
re
ener
g
y
T
h
e
th
ir
d
f
ea
t
u
r
e
em
p
l
o
y
ed
w
as
th
e
R
MSE
,
o
r
th
e
r
o
o
t
m
e
an
s
q
u
ar
e
v
alu
e
o
f
a
s
ig
n
al,
wh
ich
was
d
er
iv
ed
b
y
co
m
p
u
ti
n
g
th
e
s
q
u
ar
e
r
o
o
t
o
f
th
e
m
ea
n
v
alu
e
o
f
t
h
e
s
q
u
ar
ed
s
am
p
les.
Fo
r
ea
ch
s
am
p
le
[
]
in
th
e
au
d
io
s
ig
n
al
,
th
e
s
q
u
a
r
e
was
ca
lcu
lated
as:
[
]
2
.
T
h
e
a
v
er
ag
e
o
f
all
r
esu
ltin
g
s
q
u
ar
ed
v
alu
es
was
th
en
ca
lcu
lated
,
an
d
t
h
e
s
q
u
ar
e
r
o
o
t
o
f
th
is
av
er
a
g
e
was u
s
ed
to
d
e
ter
m
in
e
th
e
R
MSE
,
as sh
o
wn
in
(
2
)
:
=
√
1
∑
[
]
2
=
1
(
2
)
wh
er
e
[
]
is
th
e
s
ig
n
al
v
alu
e
at
in
d
ex
,
an
d
is
th
e
to
tal
n
u
m
b
e
r
o
f
s
i
g
n
al
s
am
p
les.
R
MSE
r
ef
lecte
d
th
e
in
te
n
s
ity
o
r
v
o
lu
m
e
o
f
th
e
s
p
ee
ch
s
ig
n
al.
E
m
o
tio
n
s
s
u
ch
as
an
g
r
y
o
r
h
ap
p
y
in
v
o
lv
ed
h
ig
h
er
en
e
r
g
y
lev
el
s
d
u
e
to
l
o
u
d
er
an
d
m
o
r
e
f
o
r
ce
f
u
l
s
p
ee
ch
,
r
esu
ltin
g
in
h
ig
h
er
R
MSE
v
alu
es.
C
o
n
v
er
s
ely
,
em
o
tio
n
s
lik
e
s
ad
wer
e
ex
p
r
ess
ed
w
ith
s
o
f
ter
,
lo
wer
-
en
er
g
y
s
p
ee
c
h
,
lead
in
g
to
lo
wer
R
MSE
v
alu
es.
2
.
2
.
4
.
P
it
ch
T
h
e
f
o
u
r
th
f
ea
tu
r
e
a
n
aly
ze
d
was
p
itch
.
Pit
ch
esti
m
atio
n
f
r
o
m
an
a
u
d
io
s
ig
n
al
in
v
o
lv
es
s
ev
er
al
k
ey
s
tep
s
to
ac
cu
r
ately
d
eter
m
in
e
th
e
f
u
n
d
am
e
n
tal
f
r
eq
u
en
cy
o
r
p
itch
.
T
h
is
p
r
o
ce
s
s
in
clu
d
es
s
p
ec
tr
al
an
aly
s
is
u
s
in
g
tech
n
iq
u
es
s
u
ch
as
th
e
f
ast
Fo
u
r
ier
t
r
an
s
f
o
r
m
(
FF
T
)
-
b
ased
m
eth
o
d
s
lik
e
au
to
co
r
r
elatio
n
o
r
ce
p
s
tr
al
an
aly
s
is
.
Pit
ch
was c
alcu
lated
as sh
o
wn
in
(
3
)
:
ℎ
=
(
3
)
E
m
o
tio
n
al
s
tates
ex
p
r
ess
ed
th
r
o
u
g
h
v
a
r
iatio
n
s
in
th
e
p
itch
o
f
th
e
v
o
ice.
E
m
o
tio
n
s
s
u
c
h
as
an
g
r
y
o
r
h
a
p
p
y
ten
d
ed
to
p
r
o
d
u
ce
h
ig
h
er
p
itc
h
v
ar
iatio
n
s
,
wh
er
e
th
e
v
o
ice
r
ea
ch
ed
elev
ated
f
r
e
q
u
en
cies,
ad
d
in
g
an
en
e
r
g
etic
o
r
in
ten
s
e
q
u
ality
to
th
e
s
p
ee
c
h
.
I
n
co
n
tr
ast,
s
ad
o
r
n
eu
t
r
al
in
v
o
lv
e
d
a
lo
wer
,
m
o
r
e
s
tab
le
p
itch
,
co
n
v
ey
in
g
a
ca
lm
er
o
r
m
o
r
e
s
u
b
d
u
ed
to
n
e
an
d
s
ig
n
alin
g
r
ed
u
ce
d
em
o
tio
n
al
ar
o
u
s
al.
2
.
2
.
5
.
Sp
ec
t
ra
l c
ent
ro
id
T
h
e
f
if
th
f
ea
tu
r
e
ca
lcu
late
d
w
as
th
e
s
p
ec
tr
al
ce
n
tr
o
id
,
wh
ic
h
r
ep
r
esen
t
s
th
e
ce
n
ter
o
f
g
r
a
v
ity
o
f
th
e
au
d
io
s
ig
n
al’
s
f
r
eq
u
e
n
cy
s
p
e
ctr
u
m
,
p
r
o
v
id
i
n
g
an
av
er
a
g
e
f
r
eq
u
e
n
cy
weig
h
ted
b
y
th
e
am
p
litu
d
e
o
f
ea
ch
s
p
ec
tr
al
co
m
p
o
n
e
n
t.
I
t
is
co
m
m
o
n
ly
u
s
ed
to
d
escr
ib
e
h
o
w
en
er
g
y
is
d
is
tr
ib
u
ted
ac
r
o
s
s
th
e
f
r
eq
u
e
n
cy
r
a
n
g
e,
o
f
f
er
in
g
in
s
ig
h
t in
to
th
e
b
r
i
g
h
t
n
ess
o
r
s
h
ar
p
n
ess
o
f
a
s
o
u
n
d
.
Sp
ec
tr
al
ce
n
tr
o
id
was c
alcu
lat
ed
u
s
in
g
(
4
)
:
=
∑
(
)
∙
|
(
)
|
−
1
=
0
∑
|
(
)
|
−
1
=
0
(
4
)
wh
er
e
(
)
is
th
e
f
r
eq
u
en
c
y
at
in
d
e
x
,
an
d
|
(
)
|
is
th
e
m
ag
n
itu
d
e
o
f
th
e
s
p
ec
tr
u
m
at
in
d
e
x
.
Sp
ec
tr
al
ce
n
tr
o
id
d
is
tin
g
u
is
h
e
d
em
o
tio
n
al
s
tates
in
s
p
ee
ch
b
y
r
ef
lectin
g
th
e
b
r
ig
h
tn
ess
o
r
s
h
ar
p
n
ess
o
f
th
e
v
o
ice.
Hig
h
e
r
s
p
ec
tr
al
ce
n
tr
o
id
v
alu
es,
lin
k
ed
to
a
n
g
r
y
o
r
h
ap
p
y
,
in
d
icate
d
e
n
e
r
g
y
co
n
ce
n
tr
ated
in
h
ig
h
er
f
r
eq
u
en
cies.
Me
a
n
wh
ile,
lo
wer
v
alu
es,
ass
o
ciate
d
w
ith
n
eu
tr
al
o
r
s
ad
,
s
u
g
g
ested
a
s
o
f
ter
a
n
d
m
o
r
e
s
u
b
d
u
ed
t
o
n
e
.
On
ce
ea
ch
f
ea
tu
r
e
was
ex
tr
ac
t
ed
f
r
o
m
ea
ch
a
u
d
io
,
a
f
ea
tu
r
e
v
ec
to
r
was
f
o
r
m
ed
to
r
ep
r
esen
t
th
e
k
e
y
ac
o
u
s
tic
ch
ar
ac
ter
is
tics
o
f
th
e
s
o
u
n
d
.
T
h
is
v
ec
to
r
ca
p
t
u
r
ed
th
e
m
o
s
t
s
ig
n
if
ican
t
ch
ar
ac
t
er
is
tics
o
f
th
e
au
d
io
,
wh
ich
wer
e
ess
en
tial
in
d
is
tin
g
u
is
h
in
g
v
ar
io
u
s
em
o
tio
n
al
s
t
ates.
T
h
ese
v
alu
es
wer
e
th
e
n
u
s
ed
as
in
p
u
t
in
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
I
n
d
o
n
esia
n
s
p
ee
ch
em
o
tio
n
r
e
co
g
n
itio
n
:
fea
tu
r
e
ex
tr
a
ctio
n
a
n
d
n
eu
r
a
l
…
(
I
z
z
a
N
u
r
A
fifa
h
)
3773
class
if
icatio
n
p
r
o
ce
s
s
,
wh
er
e
th
ey
h
elp
ed
t
h
e
m
o
d
el
r
ec
o
g
n
ize
an
d
d
is
tin
g
u
is
h
b
etwe
en
v
ar
io
u
s
ty
p
es
o
f
em
o
tio
n
al
ex
p
r
ess
io
n
s
.
2
.
3
.
Cla
s
s
if
ica
t
io
n
T
h
e
ex
tr
ac
ted
f
ea
tu
r
es
wer
e
u
s
ed
as
in
p
u
ts
f
o
r
em
o
tio
n
r
e
co
g
n
itio
n
t
h
r
o
u
g
h
v
ar
io
u
s
cla
s
s
if
icatio
n
tech
n
iq
u
es.
I
n
t
h
is
s
tu
d
y
,
b
o
th
C
NN
an
d
C
NN
–
L
STM
m
o
d
els
wer
e
ap
p
lied
to
c
o
m
p
ar
e
in
e
m
o
tio
n
r
ec
o
g
n
itio
n
.
T
h
ese
m
o
d
els
we
r
e
ass
ess
ed
to
m
ea
s
u
r
e
th
eir
ac
cu
r
ac
y
in
em
o
tio
n
class
if
icatio
n
u
s
in
g
f
ea
tu
r
es
ex
tr
ac
ted
f
r
o
m
th
e
s
p
ee
ch
d
at
a.
T
h
e
ex
p
er
im
en
ts
wer
e
co
n
d
u
cted
with
d
ata
d
iv
id
ed
in
to
7
5
%
f
o
r
tr
ain
in
g
,
2
0
% f
o
r
test
in
g
,
an
d
5
% f
o
r
v
alid
atio
n
.
T
h
e
m
o
d
els we
r
e
im
p
lem
en
ted
an
d
test
ed
in
Go
o
g
l
e
C
o
lab
.
2
.
3
.
1
.
Co
nv
o
lutio
na
l neura
l
net
wo
rk
s
T
h
e
C
NN
m
o
d
el
u
s
ed
was
a
1
D
C
NN
d
esig
n
ed
to
class
if
y
in
p
u
t
d
ata
with
1
D
d
im
en
s
io
n
s
,
s
u
ch
as
tim
e
s
er
ies
o
r
s
en
s
o
r
d
ata,
wh
er
e
th
e
o
r
d
er
o
r
r
elativ
e
p
o
s
itio
n
o
f
th
e
d
ata
is
im
p
o
r
tan
t.
T
h
is
m
o
d
el
p
r
o
ce
s
s
ed
th
e
in
p
u
t
d
ata
th
r
o
u
g
h
m
u
ltip
le
co
n
v
o
l
u
tio
n
al
lay
er
s
,
ex
t
r
ac
tin
g
s
p
atial
f
ea
tu
r
es
th
at
h
elp
ed
in
em
o
tio
n
class
if
icatio
n
b
ased
o
n
th
e
s
p
e
ec
h
d
ata.
Fig
u
r
e
2
d
e
p
icts
th
e
ar
ch
itectu
r
e
o
f
a
o
n
e
-
d
im
en
s
io
n
al
C
NN
m
o
d
el.
T
h
e
1
D
C
NN
ar
ch
itectu
r
e
b
eg
an
with
an
i
n
p
u
t
lay
e
r
o
f
s
ize
(
2
3
,
3
2
)
,
f
o
llo
wed
b
y
s
ev
er
al
co
n
v
o
l
u
tio
n
al
lay
er
s
with
f
ilter
s
o
f
3
2
,
6
4
,
an
d
1
2
8
,
ea
ch
ac
co
m
p
an
ied
b
y
b
atch
n
o
r
m
a
lizatio
n
,
ac
tiv
atio
n
f
u
n
ctio
n
s
,
an
d
p
o
o
lin
g
lay
er
s
to
r
ed
u
ce
d
ata
d
im
e
n
s
io
n
ality
.
A
d
r
o
p
o
u
t
lay
er
was
th
en
ap
p
lied
to
p
r
e
v
en
t
o
v
er
f
itti
n
g
.
T
h
e
o
u
tp
u
t
f
r
o
m
t
h
e
last
co
n
v
o
lu
tio
n
al
lay
e
r
wa
s
f
latten
ed
an
d
p
ass
ed
to
d
en
s
e
lay
er
s
in
o
r
d
er
to
g
et
th
e
f
in
al
o
u
tp
u
t,
w
h
ich
wa
s
u
s
ed
to
ca
teg
o
r
ize
th
e
f
o
u
r
e
m
o
tio
n
s
.
Fig
u
r
e
2
.
C
NN1
D
ar
ch
itectu
r
e
2
.
3
.
2
.
Co
nv
o
lutio
na
l neura
l
ne
t
wo
rk
s
-
lo
ng
s
ho
rt
-
t
er
m
mem
o
ry
T
h
e
C
NN
m
o
d
el
f
o
llo
wed
b
y
a
L
STM
n
etwo
r
k
,
o
f
ten
r
ef
er
r
ed
to
as
a
C
NN
-
L
STM
m
o
d
el,
is
ty
p
ically
u
s
ed
f
o
r
p
r
o
ce
s
s
in
g
h
ig
h
-
d
im
e
n
s
io
n
al
d
ata
s
u
ch
a
s
au
d
io
o
r
v
id
eo
.
I
n
th
is
m
o
d
el,
C
NN
r
etr
iev
ed
s
p
atial
ch
ar
ac
ter
is
tics
f
r
o
m
t
h
e
in
p
u
t,
an
d
L
STM
ca
p
tu
r
e
d
th
e
tem
p
o
r
al
d
e
p
en
d
e
n
cies
am
o
n
g
th
e
d
e
r
iv
ed
f
ea
tu
r
es.
Fig
u
r
e
3
d
ep
icts
th
e
a
r
ch
itectu
r
e
o
f
th
e
C
NN
-
L
STM
m
o
d
el.
Fig
u
r
e
3
.
C
NN
-
L
STM
ar
ch
ite
ctu
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
7
6
9
-
3778
3774
T
h
e
C
NN
-
L
STM
ar
ch
itectu
r
e
was
s
im
ilar
to
th
e
1
D
C
NN
ar
ch
itectu
r
e.
I
n
th
is
m
o
d
el,
t
h
e
o
u
tp
u
t
f
r
o
m
th
e
co
n
v
o
lu
tio
n
al
lay
er
s
was
s
en
t
in
to
an
L
STM
lay
e
r
,
wh
ich
ca
p
tu
r
ed
lo
n
g
-
ter
m
d
ep
en
d
en
cies
in
th
e
d
ata.
T
h
e
o
u
t
p
u
t
f
r
o
m
th
e
L
S
T
M
lay
er
was
f
latten
e
d
an
d
tr
an
s
f
er
r
ed
t
o
a
d
en
s
e
lay
e
r
th
a
t
h
ad
f
o
u
r
n
e
u
r
o
n
s
,
ea
ch
o
f
wh
ich
r
ep
r
esen
ted
o
n
e
o
f
th
e
f
o
u
r
em
o
tio
n
al
ca
teg
o
r
ies.
T
h
is
f
in
al
d
e
n
s
e
lay
er
s
er
v
ed
as
th
e
o
u
tp
u
t,
p
r
o
v
id
i
n
g
th
e
p
r
ed
icted
em
o
ti
o
n
b
ased
o
n
th
e
e
x
tr
ac
ted
f
ea
t
u
r
es.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
is
s
ec
tio
n
p
r
o
v
id
es
th
e
ess
en
tial
r
esu
lts
o
f
th
e
r
esear
ch
an
d
a
d
is
cu
s
s
io
n
o
n
th
eir
s
ig
n
if
ica
n
ce
.
I
t
ex
p
lo
r
es
asp
ec
ts
s
u
ch
as
in
ter
-
r
ater
r
eliab
ilit
y
an
d
c
o
m
p
ar
es
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
m
eth
o
d
s
an
d
class
if
icatio
n
tech
n
iq
u
es
u
s
ed
.
T
h
ese
r
esu
lts
p
r
o
v
id
e
i
n
s
ig
h
t
in
to
th
e
ef
f
ec
t
iv
en
ess
o
f
em
o
tio
n
r
ec
o
g
n
itio
n
f
r
o
m
s
p
ee
c
h
an
d
h
ig
h
lig
h
t im
p
o
r
tan
t tr
e
n
d
s
o
b
s
er
v
ed
th
r
o
u
g
h
o
u
t
th
e
an
aly
s
is
.
3
.
1
.
I
nte
r
-
ra
t
er
re
lia
bil
it
y
re
s
ults
T
h
e
s
tu
d
y
f
o
u
n
d
th
at
th
e
an
n
o
tato
r
s
h
ad
th
e
h
ig
h
est
ag
r
ee
m
en
t
wh
en
lab
elin
g
s
eg
m
en
t
s
with
th
e
em
o
tio
n
"a
n
g
r
y
,
"
ac
h
iev
i
n
g
a
s
co
r
e
o
f
0
.
8
3
.
T
h
is
s
u
g
g
ests
th
at
"a
n
g
r
y
"
was
a
d
is
tin
ct
an
d
ea
s
ily
r
ec
o
g
n
izab
le
em
o
tio
n
,
lead
in
g
to
m
o
r
e
co
n
s
is
ten
t
lab
elin
g
am
o
n
g
th
e
a
n
n
o
tato
r
s
.
T
h
e
em
o
tio
n
s
"h
a
p
p
y
"
an
d
"sad
"
h
ad
ag
r
ee
m
en
t
lev
els
o
f
0
.
7
8
an
d
0
.
7
4
,
r
esp
ec
tiv
el
y
.
I
n
co
n
tr
ast,
"n
eu
tr
al"
h
ad
th
e
lo
west
ag
r
e
em
en
t
lev
el,
with
a
s
co
r
e
o
f
0
.
7
2
.
T
h
is
in
d
icate
d
t
h
at
th
e
ab
s
en
ce
o
f
em
o
tio
n
o
r
a
n
eu
tr
al
s
tate
is
m
o
r
e
s
u
b
jec
tiv
e
an
d
h
a
r
d
er
t
o
lab
el
co
n
s
is
ten
tly
.
T
h
e
am
b
ig
u
ity
an
d
s
u
b
tlety
in
n
e
u
tr
a
l
ex
p
r
ess
io
n
s
lik
ely
co
n
tr
ib
u
ted
to
th
is
lo
wer
ag
r
ee
m
en
t le
v
el.
T
h
e
o
v
er
all
C
o
h
en
’
s
Kap
p
a
r
esu
lts
ar
e
illu
s
tr
ate
d
in
Fig
u
r
e
4
.
T
h
e
o
v
e
r
all
ag
r
ee
m
e
n
t
ac
r
o
s
s
all
em
o
tio
n
s
in
th
e
co
r
p
u
s
was
0
.
6
9
,
w
h
ich
f
ell
in
t
o
th
e
"su
b
s
tan
tial"
ag
r
ee
m
en
t
ca
teg
o
r
y
.
T
h
is
o
v
e
r
all
Kap
p
a
v
alu
e
h
i
g
h
lig
h
ted
th
e
v
ar
iab
ilit
y
an
d
s
u
b
jectiv
ity
in
h
o
w
th
e
two
an
n
o
tato
r
s
p
er
ce
iv
ed
an
d
lab
e
led
em
o
tio
n
s
.
T
h
e
r
esu
lts
o
f
t
h
e
d
ataset
ca
lcu
latio
n
s
u
s
in
g
I
B
M
SP
S
S
s
o
f
twar
e
wer
e
s
u
m
m
ar
ized
in
T
ab
le
2
.
T
h
e
Kap
p
a
v
alu
e
o
f
0
.
6
9
8
i
n
d
icate
d
a
s
u
b
s
tan
tial
lev
el
o
f
ag
r
ee
m
en
t
b
etwe
en
th
e
a
n
n
o
tato
r
s
,
s
u
g
g
esti
n
g
th
at
th
ey
o
f
ten
lab
eled
em
o
tio
n
s
co
n
s
is
ten
tly
.
T
h
e
T
-
v
alu
e
o
f
1
7
.
0
3
8
an
d
a
s
ig
n
if
ican
ce
lev
el
o
f
less
th
an
0
.
0
0
1
co
n
f
ir
m
ed
th
a
t
th
e
f
in
d
in
g
s
wer
e
s
tatis
ticall
y
s
ig
n
if
ican
t,
m
ea
n
in
g
t
h
at
th
e
o
b
s
er
v
ed
r
esu
lts
wer
e
u
n
lik
ely
to
h
a
v
e
o
cc
u
r
r
ed
b
y
ch
an
c
e.
T
h
is
in
d
icate
d
th
e
r
eliab
ilit
y
o
f
th
e
m
e
asu
r
em
en
t
p
r
o
ce
s
s
.
Ho
wev
er
,
th
ese
r
esu
lts
also
h
ig
h
lig
h
ted
th
at
w
h
ile
an
n
o
t
ato
r
s
g
en
er
ally
ag
r
ee
,
ce
r
tai
n
em
o
tio
n
s
led
t
o
in
co
n
s
is
ten
cies in
lab
elin
g
.
Fig
u
r
e
4
.
C
o
h
e
n
’
s
Kap
p
a
r
esu
lts
T
ab
le
2
.
C
o
h
e
n
’
s
Kap
p
a
r
esu
lt
s
f
r
o
m
I
B
M
SP
SS
V
a
l
u
e
A
sy
mp
t
o
t
i
c
s
t
a
n
d
a
r
d
e
r
r
o
r
a
A
p
p
r
o
x
i
ma
t
e
T
b
A
p
p
r
o
x
i
ma
t
e
si
g
n
i
f
i
c
a
n
c
e
0
.
6
9
8
0
.
0
3
9
1
7
.
0
3
8
<
0
.
0
0
1
3
.
2
.
F
e
a
t
ure
ex
t
r
a
ct
io
n a
nd
cla
s
s
if
ica
t
io
n c
o
m
pa
riso
n
E
x
p
er
im
en
ts
wer
e
c
o
n
d
u
cted
u
s
in
g
C
NN
an
d
C
NN
-
L
STM
m
eth
o
d
s
with
in
p
u
t
d
ata
d
er
iv
ed
f
r
o
m
f
ea
tu
r
e
ex
tr
ac
tio
n
o
f
s
p
ee
ch
s
ig
n
als.
T
h
e
f
ea
tu
r
es
ex
tr
ac
ted
f
r
o
m
th
e
s
p
ee
ch
s
ig
n
als
in
clu
d
ed
MFC
C
,
Z
C
R
,
R
MSE
,
p
itch
,
an
d
s
p
ec
tr
al
ce
n
tr
o
id
.
T
h
ese
f
ea
tu
r
es
wer
e
an
aly
ze
d
to
ass
ess
its
im
p
ac
t
in
im
p
r
o
v
in
g
e
m
o
tio
n
class
if
icatio
n
ac
cu
r
ac
y
.
0
,
6
5
0
,
7
0
,
7
5
0
,
8
0
,
8
5
Ne
u
tral
S
a
d
Ha
p
p
y
An
g
ry
Ag
r
e
e
m
e
n
t
Le
v
e
l
Em
o
tio
n
s
Ag
r
e
e
m
e
n
t
lev
e
l
fo
r
e
a
c
h
e
m
o
tio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
I
n
d
o
n
esia
n
s
p
ee
ch
em
o
tio
n
r
e
co
g
n
itio
n
:
fea
tu
r
e
ex
tr
a
ctio
n
a
n
d
n
eu
r
a
l
…
(
I
z
z
a
N
u
r
A
fifa
h
)
3775
3
.
2
.
1
.
Co
nv
o
lutio
na
l neura
l
net
wo
rk
s
I
n
th
e
C
NN
m
eth
o
d
,
v
ar
io
u
s
f
ea
tu
r
e
co
m
b
in
atio
n
s
wer
e
test
ed
to
ac
h
iev
e
th
e
b
est
r
esu
lts
i
n
em
o
tio
n
class
if
icatio
n
.
T
h
ese
r
esu
lts
h
i
g
h
lig
h
ted
th
e
s
ig
n
if
ica
n
ce
o
f
ea
ch
f
ea
tu
r
e
in
h
el
p
in
g
th
e
m
o
d
el
to
d
if
f
e
r
en
tiate
em
o
tio
n
al
s
tates.
T
ab
le
3
p
r
esen
ts
th
e
ac
cu
r
ac
y
o
f
test
in
g
u
s
in
g
C
NN
with
d
if
f
er
en
t f
ea
t
u
r
e
co
m
b
in
atio
n
s
.
T
ab
le
3
.
Acc
u
r
ac
y
co
m
p
ar
is
o
n
u
s
in
g
C
NN
F
e
a
t
u
r
e
s
A
c
c
u
r
a
c
y
M
F
C
C
8
1
%
M
F
C
C
+
Z
C
R
8
1
%
M
F
C
C
+
Z
C
R
+
R
M
S
E
8
3
%
M
F
C
C
+
Z
C
R
+
R
M
S
E
+
P
i
t
c
h
8
8
%
M
F
C
C
+
Z
C
R
+
R
M
S
E
+
P
i
t
c
h
+
S
p
e
c
t
r
a
l
C
e
n
t
r
o
i
d
8
5
%
T
h
e
u
s
e
o
f
MFC
C
alo
n
e
r
esu
lted
in
an
ac
cu
r
ac
y
o
f
8
1
%.
M
FC
C
is
ef
f
ec
tiv
e
in
ca
p
tu
r
in
g
im
p
o
r
tan
t
s
p
ec
tr
al
in
f
o
r
m
atio
n
,
h
o
wev
er
u
s
in
g
MFC
C
alo
n
e
m
ay
n
o
t
f
u
lly
ca
p
tu
r
e
th
e
tem
p
o
r
al
d
i
m
en
s
io
n
s
in
s
p
ee
c
h
th
at
co
r
r
esp
o
n
d
to
em
o
tio
n
al
s
tates.
Ad
d
in
g
th
e
Z
C
R
to
MFC
C
d
id
n
o
t
s
ig
n
i
f
ican
tly
im
p
r
o
v
e
ac
c
u
r
ac
y
.
Z
C
R
d
eter
m
in
es
h
o
w
f
r
e
q
u
en
tly
t
h
e
s
ig
n
al
cr
o
s
s
es
th
e
ze
r
o
-
am
p
litu
d
e
lin
e
with
in
a
s
p
ec
if
ic
tim
e
f
r
am
e,
b
u
t
its
co
n
tr
ib
u
tio
n
to
em
o
tio
n
class
if
icatio
n
s
ee
m
ed
less
s
ig
n
if
ica
n
t
co
m
p
a
r
ed
t
o
o
t
h
er
f
ea
tu
r
es.
W
h
en
R
MSE
was
ad
d
ed
to
th
e
c
o
m
b
in
atio
n
o
f
MFC
C
an
d
Z
C
R
,
ac
cu
r
ac
y
in
cr
ea
s
ed
to
8
3
%.
R
MSE
,
wh
ich
m
ea
s
u
r
es th
e
en
er
g
y
o
f
th
e
s
p
ee
ch
s
ig
n
al,
en
r
ich
es
th
e
r
ep
r
esen
tatio
n
o
f
th
e
tem
p
o
r
al
an
d
s
tr
en
g
th
asp
ec
ts
o
f
th
e
s
ig
n
al
th
at
ar
e
r
elev
an
t
f
o
r
em
o
tio
n
d
etec
tio
n
.
T
h
e
in
c
r
ea
s
e
in
ac
cu
r
ac
y
s
u
g
g
ested
th
at
s
ig
n
al
e
n
er
g
y
in
f
o
r
m
atio
n
p
lay
ed
an
im
p
o
r
tan
t
r
o
le
in
d
if
f
er
e
n
tiatin
g
em
o
tio
n
al
e
x
p
r
ess
io
n
s
.
Ad
d
i
n
g
p
itch
t
o
th
e
co
m
b
in
atio
n
o
f
MFC
C
,
Z
C
R
,
an
d
R
MSE
led
to
a
s
ig
n
if
ican
t
in
cr
ea
s
e
in
ac
cu
r
ac
y
to
8
8
%.
Pit
ch
p
r
o
v
id
es
in
f
o
r
m
atio
n
ab
o
u
t
v
o
ice
in
t
o
n
atio
n
,
wh
ich
is
cr
u
cial
in
em
o
tio
n
r
e
co
g
n
itio
n
b
ec
a
u
s
e
v
ar
iatio
n
s
i
n
in
to
n
atio
n
ca
n
r
ef
lect
d
ee
p
em
o
tio
n
al
c
h
an
g
es.
T
h
e
s
u
b
s
tan
tial
co
n
tr
ib
u
tio
n
o
f
p
itch
u
n
d
er
s
co
r
ed
th
e
im
p
o
r
tan
ce
o
f
in
to
n
atio
n
in
d
is
tin
g
u
is
h
in
g
e
m
o
tio
n
al
ex
p
r
ess
io
n
s
.
Ho
wev
er
,
a
d
d
in
g
Sp
ec
tr
al
C
en
tr
o
id
to
t
h
e
c
o
m
b
in
atio
n
o
f
MFC
C
,
Z
C
R
,
R
MSE
,
an
d
p
itch
s
lig
h
tly
d
ec
r
ea
s
ed
ac
cu
r
ac
y
t
o
8
5
%.
Sp
ec
tr
al
C
en
tr
o
id
,
w
h
ich
d
escr
ib
es
th
e
ce
n
te
r
o
f
m
ass
o
f
th
e
s
p
ec
tr
al
s
ig
n
al,
d
id
n
o
t
s
ee
m
t
o
p
r
o
v
i
d
e
s
ig
n
if
ican
t
ad
d
itio
n
al
v
alu
e
in
th
e
co
n
te
x
t
o
f
em
o
tio
n
class
if
icatio
n
,
o
r
it
m
ig
h
t e
v
en
co
m
p
licate
th
e
m
o
d
el
with
o
u
t a
d
d
in
g
in
f
o
r
m
ativ
e
v
alu
e.
3
.
2
.
2
.
Co
nv
o
lutio
na
l neura
l
net
wo
rk
s
-
lo
ng
s
ho
rt
-
t
er
m
mem
o
ry
T
h
e
C
NN
-
L
STM
m
eth
o
d
al
s
o
d
em
o
n
s
tr
ated
s
tr
o
n
g
p
er
f
o
r
m
an
ce
in
em
o
tio
n
class
if
icatio
n
.
B
y
co
m
b
in
in
g
th
e
f
ea
tu
r
e
ex
tr
ac
t
io
n
ca
p
a
b
ilit
ies
with
th
e
s
eq
u
en
tial
m
o
d
elin
g
p
o
wer
o
f
L
S
T
M,
th
is
ap
p
r
o
ac
h
ca
p
tu
r
ed
b
o
th
th
e
r
elev
an
t
f
e
atu
r
es
f
r
o
m
th
e
s
p
ee
ch
s
ig
n
a
l
an
d
th
e
s
eq
u
e
n
tial
d
ep
en
d
e
n
cies
in
th
e
d
ata.
T
ab
le
4
s
h
o
ws th
e
ac
cu
r
ac
y
o
f
test
in
g
u
s
in
g
C
NN
-
L
STM
wit
h
v
ar
io
u
s
f
ea
tu
r
e
c
o
m
b
in
atio
n
s
.
T
ab
le
4
.
Acc
u
r
ac
y
co
m
p
ar
is
o
n
u
s
in
g
C
NN
-
L
STM
F
e
a
t
u
r
e
s
A
c
c
u
r
a
c
y
M
F
C
C
8
1
%
M
F
C
C
+
Z
C
R
7
7
%
M
F
C
C
+
Z
C
R
+
R
M
S
E
8
8
%
M
F
C
C
+
Z
C
R
+
R
M
S
E
+
P
i
t
c
h
8
3
%
M
F
C
C
+
Z
C
R
+
R
M
S
E
+
P
i
t
c
h
+
S
p
e
c
t
r
a
l
C
e
n
t
r
o
i
d
8
3
%
Usi
n
g
MFC
C
alo
n
e
r
esu
lted
in
an
ac
cu
r
ac
y
o
f
8
1
%.
MFC
C
is
k
n
o
wn
to
b
e
ef
f
ec
tiv
e
in
ex
tr
ac
tin
g
s
p
ec
tr
al
in
f
o
r
m
atio
n
f
r
o
m
au
d
io
s
ig
n
als,
b
u
t
th
is
ac
cu
r
ac
y
s
u
g
g
ested
t
h
at
th
e
i
n
f
o
r
m
a
tio
n
o
b
tain
e
d
f
r
o
m
MFC
C
alo
n
e
s
til
l
h
ad
lim
itati
o
n
s
in
f
u
lly
d
etec
tin
g
em
o
tio
n
al
v
ar
iati
o
n
s
.
Ad
d
in
g
th
e
Z
C
R
to
MFC
C
ac
tu
all
y
r
ed
u
ce
d
ac
cu
r
ac
y
to
7
7
%.
T
h
is
r
ed
u
ctio
n
m
ig
h
t
h
av
e
b
ee
n
d
u
e
t
o
Z
C
R
in
tr
o
d
u
cin
g
n
o
i
s
e
o
r
less
r
elev
an
t
in
f
o
r
m
atio
n
,
th
e
r
eb
y
d
is
r
u
p
tin
g
th
e
m
o
d
el’
s
a
b
ilit
y
to
class
i
f
y
em
o
tio
n
s
ac
cu
r
ately
.
Ho
w
ev
er
,
wh
e
n
R
MSE
was
ad
d
ed
to
t
h
e
co
m
b
in
atio
n
o
f
MFC
C
an
d
Z
C
R
,
ac
cu
r
ac
y
s
ig
n
if
ican
tly
in
cr
ea
s
ed
to
8
8
%.
R
MSE
p
r
o
v
id
es
ad
d
itio
n
al
in
f
o
r
m
atio
n
ab
o
u
t
th
e
in
ten
s
ity
o
f
t
h
e
s
p
ee
ch
s
ig
n
al,
wh
ich
is
cr
u
cial
f
o
r
d
is
tin
g
u
is
h
in
g
em
o
tio
n
s
.
T
h
is
in
cr
ea
s
e
in
ac
cu
r
ac
y
in
d
icate
d
th
at
R
MSE
ad
d
ed
cr
u
ci
al
in
f
o
r
m
ativ
e
v
alu
e
f
o
r
em
o
tio
n
d
etec
tio
n
.
Ad
d
in
g
Pit
ch
to
th
e
co
m
b
in
at
io
n
o
f
MFC
C
,
Z
C
R
,
an
d
R
M
SE
in
cr
ea
s
ed
ac
cu
r
ac
y
to
8
3
%.
Alth
o
u
g
h
Pit
ch
s
h
o
u
ld
p
r
o
v
id
e
ad
d
itio
n
al
in
f
o
r
m
atio
n
ab
o
u
t
th
e
f
u
n
d
am
en
tal
f
r
eq
u
en
cy
o
f
th
e
v
o
ice
r
el
ev
an
t
f
o
r
em
o
tio
n
class
if
icat
io
n
,
th
is
in
cr
ea
s
e
in
ac
cu
r
ac
y
was
n
o
t
as
s
ig
n
if
ic
an
t
as
in
th
e
p
r
ev
io
u
s
co
m
b
i
n
atio
n
.
T
h
is
m
ig
h
t
h
av
e
b
ee
n
b
ec
au
s
e
th
e
in
f
o
r
m
atio
n
p
r
o
v
id
ed
b
y
Pit
ch
d
i
d
n
o
t
ad
d
en
o
u
g
h
v
alu
e
to
th
e
m
o
d
el
o
r
th
er
e
was
r
ed
u
n
d
an
cy
with
ex
is
tin
g
f
ea
t
u
r
es.
Ad
d
in
g
s
p
ec
tr
al
ce
n
tr
o
i
d
to
th
e
co
m
b
in
atio
n
o
f
MFC
C
,
Z
C
R
,
R
MSE
,
an
d
Pit
ch
d
id
n
o
t f
u
r
th
er
in
cr
ea
s
e
ac
cu
r
ac
y
,
wh
ich
r
em
ai
n
ed
at
8
3
%.
Sp
ec
tr
al
ce
n
tr
o
id
,
wh
ich
d
escr
ib
ed
th
e
ce
n
ter
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
7
6
9
-
3778
3776
o
f
m
ass
o
f
th
e
s
o
u
n
d
s
p
ec
tr
u
m
,
d
id
n
o
t
s
ee
m
to
p
r
o
v
id
e
s
u
f
f
icien
tly
d
if
f
er
en
tiatin
g
in
f
o
r
m
atio
n
co
m
p
ar
ed
to
th
e
o
th
er
f
ea
tu
r
es o
r
m
ig
h
t h
a
v
e
h
ad
a
n
ex
ce
s
s
iv
e
o
v
e
r
lap
o
f
in
f
o
r
m
atio
n
.
3
.
2
.
3
.
Co
m
pa
riso
n
T
h
e
test
in
g
r
esu
lts
f
o
r
v
ar
io
u
s
f
ea
tu
r
e
ex
tr
ac
tio
n
s
an
d
class
if
icatio
n
m
eth
o
d
s
s
h
o
w
th
at
R
MSE
h
ad
a
s
ig
n
if
ican
t
im
p
ac
t
in
im
p
r
o
v
in
g
ac
cu
r
ac
y
f
o
r
b
o
th
t
h
e
C
NN
an
d
C
NN
-
L
STM
m
o
d
els.
I
n
t
h
e
C
NN
m
o
d
el,
th
e
co
m
b
in
atio
n
o
f
f
o
u
r
f
ea
t
u
r
es:
MFC
C
+
Z
C
R
+
R
MSE
+
Pit
ch
led
to
th
e
h
ig
h
est
ac
cu
r
ac
y
o
f
8
8
%.
Similar
ly
,
th
e
C
NN
-
L
STM
m
o
d
el
r
ea
ch
ed
th
e
s
am
e
ac
cu
r
ac
y
with
ju
s
t
th
r
ee
f
ea
tu
r
es:
MFC
C
,
Z
C
R
,
an
d
R
MSE
.
T
h
ese
r
esu
lts
em
p
h
asized
th
e
s
ig
n
if
ican
ce
o
f
s
elec
tin
g
th
e
r
ig
h
t
f
e
atu
r
es
p
r
o
v
id
in
g
th
e
m
o
s
t
v
al
u
ab
le
co
n
tr
ib
u
tio
n
s
,
wh
ile
also
h
ig
h
lig
h
ted
an
im
p
o
r
tan
t
tr
ad
e
-
o
f
f
b
etwe
en
co
m
p
u
tatio
n
al
ef
f
icien
cy
,
as
f
ewe
r
f
ea
tu
r
es
r
ed
u
ce
th
e
co
m
p
u
tatio
n
al
c
o
s
t o
f
f
ea
t
u
r
e
ex
tr
ac
tio
n
.
4.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
ex
p
lo
r
e
d
m
eth
o
d
s
to
en
h
a
n
ce
em
o
tio
n
r
ec
o
g
n
itio
n
f
r
o
m
I
n
d
o
n
esian
s
p
ee
c
h
u
s
in
g
f
ea
tu
r
e
ex
tr
ac
tio
n
tech
n
iq
u
es
an
d
m
a
ch
in
e
lear
n
in
g
class
if
icatio
n
m
o
d
els.
T
h
e
ex
p
er
im
e
n
ts
wer
e
co
n
d
u
cted
o
n
an
I
n
d
o
n
esian
lan
g
u
a
g
e
d
ataset
co
n
s
is
tin
g
o
f
2
0
0
s
am
p
les.
T
o
ass
ess
in
ter
-
r
ater
r
eliab
ilit
y
,
C
o
h
en
'
s
k
ap
p
a
an
aly
s
is
was
co
n
d
u
cted
,
wh
i
ch
r
ev
ea
led
a
s
u
b
s
tan
tial
ag
r
ee
m
en
t
lev
el
(
=
0
.
6
9
8
)
b
etwe
en
an
n
o
tato
r
s
,
h
ig
h
lig
h
tin
g
th
e
co
n
s
is
ten
cy
o
f
em
o
tio
n
lab
elin
g
.
T
h
e
class
if
icatio
n
ex
p
er
im
en
ts
co
m
p
ar
ed
C
NN
an
d
C
NN
-
L
STM
m
o
d
els.
B
o
th
th
e
C
NN
m
o
d
el,
wh
ich
u
s
ed
f
o
u
r
f
ea
tu
r
es
(
MFC
C
,
Z
C
R
,
R
MSE
,
an
d
Pit
ch
)
,
an
d
th
e
C
NN
-
L
STM
m
o
d
el,
wh
ich
u
s
ed
th
r
ee
f
ea
tu
r
es
(
MFC
C
,
Z
C
R
,
an
d
R
MSE
)
,
ac
h
iev
ed
an
em
o
tio
n
class
if
icatio
n
ac
cu
r
ac
y
o
f
ap
p
r
o
x
im
ately
8
8
%.
T
h
e
d
if
f
e
r
en
ce
in
th
e
n
u
m
b
e
r
o
f
f
ea
tu
r
es
s
u
g
g
ests
th
at
wh
ile
th
e
C
NN
m
o
d
el
in
v
o
lv
ed
m
o
r
e
co
m
p
u
tatio
n
al
task
s
d
u
e
t
o
th
e
ad
d
itio
n
al
f
ea
tu
r
e,
t
h
e
C
NN
-
L
STM
m
o
d
el
m
an
ag
ed
t
o
ac
h
iev
e
s
im
ilar
p
e
r
f
o
r
m
a
n
ce
with
f
ewe
r
f
ea
tu
r
es
,
p
o
ten
tially
o
f
f
er
in
g
a
m
o
r
e
ef
f
icien
t a
p
p
r
o
ac
h
.
Ov
er
all,
th
e
f
in
d
in
g
s
d
em
o
n
s
tr
ate
th
at
in
co
r
p
o
r
atin
g
d
iv
er
s
e
f
ea
tu
r
e
ex
tr
ac
tio
n
tech
n
iq
u
es
ca
n
en
h
an
ce
em
o
tio
n
r
ec
o
g
n
itio
n
p
er
f
o
r
m
an
ce
,
p
a
r
ticu
l
ar
ly
in
I
n
d
o
n
esian
SER.
Ho
wev
er
,
ca
r
ef
u
l
co
n
s
id
er
atio
n
is
n
ee
d
ed
t
o
b
alan
ce
co
m
p
u
tatio
n
al
ef
f
icien
c
y
an
d
f
ea
tu
r
e
c
o
m
p
lex
ity
,
as
ad
d
in
g
m
o
r
e
f
e
atu
r
es
ca
n
im
p
r
o
v
e
ac
cu
r
ac
y
b
u
t
m
ay
also
in
cr
e
ase
co
m
p
u
tatio
n
al
co
s
t.
Fu
tu
r
e
r
esear
ch
co
u
ld
ex
p
l
o
r
e
th
e
u
s
e
o
f
ad
v
an
ce
d
o
p
tim
izatio
n
tech
n
iq
u
es
o
r
f
e
atu
r
e
s
elec
tio
n
m
eth
o
d
s
to
f
u
r
th
er
r
ef
in
e
m
o
d
el
p
er
f
o
r
m
a
n
c
e
wh
ile
m
in
im
izin
g
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
.
ACK
NO
WL
E
DG
M
E
N
T
S
T
h
e
au
th
o
r
s
wo
u
ld
lik
e
to
t
h
an
k
PENS
m
a
n
ag
em
e
n
t
f
o
r
all
s
u
p
p
o
r
t
in
th
e
f
o
r
m
o
f
lab
o
r
ato
r
y
f
ac
ilit
ies a
n
d
all
th
e
eq
u
ip
m
e
n
t p
r
o
v
id
e
d
,
s
o
th
at
we
ca
n
ca
r
r
y
o
u
t th
is
r
esear
ch
well.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
h
as
b
ee
n
s
u
p
p
o
r
ted
b
y
th
e
Min
is
tr
y
o
f
E
d
u
ca
t
io
n
,
C
u
ltu
r
e,
R
esear
ch
,
an
d
T
ec
h
n
o
lo
g
y
o
f
th
e
R
ep
u
b
lic
o
f
I
n
d
o
n
esia
with
th
e
s
ch
em
e
o
f
P
en
elitia
n
Te
s
is
Ma
g
is
te
r
f
o
r
th
e
f
is
ca
l
y
ea
r
2
0
2
4
,
with
th
e
co
n
tr
ac
t n
u
m
b
er
o
f
5
2
4
/PL1
4
/
PT.
0
1
.
0
5
/I
I
I
/2
0
2
4
.
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
t
es,
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
I
zz
a
Nu
r
Af
if
ah
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
T
r
i Bu
d
i San
to
s
o
✓
✓
✓
✓
✓
✓
✓
T
ito
n
Du
to
n
o
✓
✓
✓
✓
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
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
I
n
d
o
n
esia
n
s
p
ee
ch
em
o
tio
n
r
e
co
g
n
itio
n
:
fea
tu
r
e
ex
tr
a
ctio
n
a
n
d
n
eu
r
a
l
…
(
I
z
z
a
N
u
r
A
fifa
h
)
3777
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
I
NF
O
RM
E
D
CO
NS
E
N
T
W
e
h
av
e
o
b
tain
ed
in
f
o
r
m
ed
c
o
n
s
en
t f
r
o
m
all
in
d
iv
id
u
als in
c
lu
d
ed
in
t
h
is
s
tu
d
y
.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
th
e
f
in
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
av
ailab
l
e
f
r
o
m
th
e
co
r
r
esp
o
n
d
in
g
a
u
t
h
o
r
s
,
I
NA
an
d
T
B
S,
u
p
o
n
r
ea
s
o
n
a
b
le
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
M
.
B
.
A
k
ç
a
y
a
n
d
K
.
O
ğ
u
z
,
“
S
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
:
E
m
o
t
i
o
n
a
l
m
o
d
e
l
s
,
d
a
t
a
b
a
s
e
s
,
f
e
a
t
u
r
e
s
,
p
r
e
p
r
o
c
e
s
s
i
n
g
m
e
t
h
o
d
s
,
s
u
p
p
o
r
t
i
n
g
m
o
d
a
l
i
t
i
e
s
,
a
n
d
c
l
a
s
s
i
f
i
e
r
s
,
”
S
p
e
e
c
h
C
o
m
m
u
n
i
c
a
t
i
o
n
,
v
o
l
.
1
1
6
,
p
p
.
5
6
–
7
6
,
J
a
n
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
s
p
e
c
o
m
.
2
0
1
9
.
1
2
.
0
0
1
.
[
2
]
T.
M
.
W
a
n
i
,
T
.
S
.
G
u
n
a
w
a
n
,
S
.
A
.
A
.
Q
a
d
r
i
,
M
.
K
a
r
t
i
w
i
,
a
n
d
E.
A
mb
i
k
a
i
r
a
j
a
h
,
“
A
c
o
m
p
r
e
h
e
n
si
v
e
r
e
v
i
e
w
o
f
s
p
e
e
c
h
e
mo
t
i
o
n
re
c
o
g
n
i
t
i
o
n
sy
s
t
e
m
s,”
I
EEE
Ac
c
e
ss
,
v
o
l
.
9
,
p
p
.
4
7
7
9
5
–
4
7
8
1
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
a
c
c
e
ss.
2
0
2
1
.
3
0
6
8
0
4
5
.
[
3
]
M
.
B
o
j
a
n
i
ć
,
V
.
D
e
l
i
ć
,
a
n
d
A
.
K
a
r
p
o
v
,
“
C
a
l
l
r
e
d
i
s
t
r
i
b
u
t
i
o
n
f
o
r
a
c
a
l
l
c
e
n
t
e
r
b
a
se
d
o
n
sp
e
e
c
h
e
mo
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
0
,
n
o
.
1
3
,
p
.
4
6
5
3
,
J
u
l
.
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
0
1
3
4
6
5
3
.
[
4
]
V
.
M
.
K
o
t
i
,
K
.
M
u
r
t
h
y
,
M
.
S
u
g
a
n
y
a
,
M
.
S
.
S
a
r
ma
,
G
.
V
.
S
.
S
.
S
e
s
h
u
K
u
mar,
a
n
d
B
.
N
,
“
S
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
e
x
t
r
e
me
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
EAI
En
d
o
rsed
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
t
e
rn
e
t
o
f
T
h
i
n
g
s
,
v
o
l
.
1
0
,
N
o
v
.
2
0
2
3
,
d
o
i
:
1
0
.
4
1
0
8
/
e
e
t
i
o
t
.
4
4
8
5
.
[
5
]
S
.
La
t
i
f
,
J.
Q
a
d
i
r
,
A
.
Q
a
y
y
u
m,
M
.
U
s
a
ma,
a
n
d
S
.
Y
o
u
n
i
s,
“
S
p
e
e
c
h
t
e
c
h
n
o
l
o
g
y
f
o
r
h
e
a
l
t
h
c
a
r
e
:
o
p
p
o
r
t
u
n
i
t
i
e
s
,
c
h
a
l
l
e
n
g
e
s,
a
n
d
s
t
a
t
e
o
f
t
h
e
a
r
t
,
”
I
EEE
Re
v
i
e
w
s
i
n
Bi
o
m
e
d
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
v
o
l
.
1
4
,
p
p
.
3
4
2
–
3
5
6
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
r
b
m
e
.
2
0
2
0
.
3
0
0
6
8
6
0
.
[
6
]
H
.
H
.
M
u
s
t
a
f
a
,
N
.
R
.
D
a
r
w
i
s
h
,
a
n
d
H
.
A
.
H
e
f
n
y
,
“
A
u
t
o
ma
t
i
c
s
p
e
e
c
h
e
mo
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
:
a
s
y
st
e
ma
t
i
c
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
S
p
e
e
c
h
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
7
,
n
o
.
1
,
p
p
.
2
6
7
–
2
8
5
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
7
7
2
-
0
2
4
-
1
0
0
9
6
-
7.
[
7
]
S
.
La
n
g
a
r
i
,
H
.
M
a
r
v
i
,
a
n
d
M
.
Za
h
e
d
i
,
“
Ef
f
i
c
i
e
n
t
s
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
mo
d
i
f
i
e
d
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
,
”
I
n
f
o
rm
a
t
i
c
s
i
n
Me
d
i
c
i
n
e
U
n
l
o
c
k
e
d
,
v
o
l
.
2
0
,
p
.
1
0
0
4
2
4
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
m
u
.
2
0
2
0
.
1
0
0
4
2
4
.
[
8
]
A
.
H
a
s
h
e
m
,
M
.
A
r
i
f
,
a
n
d
M
.
A
l
g
h
a
m
d
i
,
“
S
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
a
p
p
r
o
a
c
h
e
s:
A
s
y
s
t
e
m
a
t
i
c
r
e
v
i
e
w
,
”
S
p
e
e
c
h
C
o
m
m
u
n
i
c
a
t
i
o
n
,
v
o
l
.
1
5
4
,
p
.
1
0
2
9
7
4
,
O
c
t
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
s
p
e
c
o
m
.
2
0
2
3
.
1
0
2
9
7
4
.
[
9
]
M
.
L
i
u
,
“
E
n
g
l
i
sh
sp
e
e
c
h
e
mo
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
me
t
h
o
d
b
a
se
d
o
n
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
S
p
e
e
c
h
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
5
,
n
o
.
2
,
p
p
.
3
9
1
–
3
9
8
,
F
e
b
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
7
7
2
-
0
2
1
-
0
9
9
5
5
-
4.
[
1
0
]
M
.
H
.
P
h
a
m
,
F
.
M
.
N
o
o
r
i
,
a
n
d
J.
T
o
r
r
e
sen
,
“
E
mo
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
sp
e
e
c
h
d
a
t
a
w
i
t
h
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
i
n
2
0
2
1
I
EEE
2
n
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
S
i
g
n
a
l
,
C
o
n
t
ro
l
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
(
S
C
C
)
,
D
e
c
.
2
0
2
1
,
p
p
.
1
8
2
–
187
,
d
o
i
:
1
0
.
1
1
0
9
/
sc
c
5
3
7
6
9
.
2
0
2
1
.
9
7
6
8
3
7
2
.
[
1
1
]
D
.
I
ssa,
M
.
F
a
t
i
h
D
e
m
i
r
c
i
,
a
n
d
A
.
Y
a
z
i
c
i
,
“
S
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
w
i
t
h
d
e
e
p
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
,
”
B
i
o
m
e
d
i
c
a
l
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
a
n
d
C
o
n
t
ro
l
,
v
o
l
.
5
9
,
p
.
1
0
1
8
9
4
,
M
a
y
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
b
s
p
c
.
2
0
2
0
.
1
0
1
8
9
4
.
[
1
2
]
F
.
F
a
h
mi
,
M
.
A
.
J
i
w
a
n
g
g
i
,
a
n
d
M
.
A
d
r
i
a
n
i
,
“
S
p
e
e
c
h
-
e
m
o
t
i
o
n
d
e
t
e
c
t
i
o
n
i
n
a
n
I
n
d
o
n
e
s
i
a
n
m
o
v
i
e
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
1
st
J
o
i
n
t
Wo
r
k
sh
o
p
o
n
S
p
o
k
e
n
L
a
n
g
u
a
g
e
T
e
c
h
n
o
l
o
g
i
e
s
f
o
r
U
n
d
e
r
-
res
o
u
r
c
e
d
l
a
n
g
u
a
g
e
s
(
S
L
T
U
)
a
n
d
C
o
l
l
a
b
o
r
a
t
i
o
n
a
n
d
C
o
m
p
u
t
i
n
g
f
o
r
U
n
d
e
r
-
Re
s
o
u
rce
d
L
a
n
g
u
a
g
e
s (C
C
U
R
L
)
,
2
0
2
0
,
p
p
.
1
8
5
–
19
3.
[
1
3
]
P
.
K
u
r
n
i
a
w
a
t
i
,
D
.
P
.
Le
s
t
a
r
i
,
a
n
d
M
.
L.
K
h
o
d
r
a
,
“
S
p
e
e
c
h
e
mo
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
f
r
o
m
I
n
d
o
n
e
si
a
n
sp
o
k
e
n
l
a
n
g
u
a
g
e
u
s
i
n
g
a
c
o
u
s
t
i
c
a
n
d
l
e
x
i
c
a
l
f
e
a
t
u
r
e
s,
”
i
n
2
0
1
7
2
0
t
h
C
o
n
f
e
r
e
n
c
e
o
f
t
h
e
O
ri
e
n
t
a
l
C
h
a
p
t
e
r
o
f
t
h
e
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
o
r
d
i
n
a
t
i
n
g
C
o
m
m
i
t
t
e
e
o
n
S
p
e
e
c
h
D
a
t
a
b
a
se
s a
n
d
S
p
e
e
c
h
I
/
O
S
y
s
t
e
m
s
a
n
d
Ass
e
ssm
e
n
t
(
O
-
C
O
C
O
S
D
A)
,
N
o
v
.
2
0
1
7
,
p
p
.
1
–
7
,
d
o
i
:
1
0
.
1
1
0
9
/
i
c
sd
a
.
2
0
1
7
.
8
3
8
4
4
6
7
.
[
1
4
]
N
.
B
.
W
u
n
a
r
so
a
n
d
Y
.
E
.
S
o
e
l
i
s
t
i
o
,
“
T
o
w
a
r
d
s
I
n
d
o
n
e
s
i
a
n
sp
e
e
c
h
-
e
mo
t
i
o
n
a
u
t
o
ma
t
i
c
r
e
c
o
g
n
i
t
i
o
n
(
I
-
S
p
EA
R
)
,
”
i
n
2
0
1
7
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
N
e
w
M
e
d
i
a
S
t
u
d
i
e
s (CO
N
ME
D
I
A)
,
N
o
v
.
2
0
1
7
,
p
p
.
9
8
–
101
,
d
o
i
:
1
0
.
1
1
0
9
/
c
o
n
me
d
i
a
.
2
0
1
7
.
8
2
6
6
0
3
8
.
[
1
5
]
O
.
U
.
K
u
m
a
l
a
a
n
d
A
.
Za
h
r
a
,
“
I
n
d
o
n
e
si
a
n
s
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
c
r
o
ss
-
c
o
r
p
u
s
me
t
h
o
d
w
i
t
h
t
h
e
c
o
m
b
i
n
a
t
i
o
n
o
f
M
F
C
C
a
n
d
Te
a
g
e
r
e
n
e
r
g
y
f
e
a
t
u
r
e
s,”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
A
d
v
a
n
c
e
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
2
,
n
o
.
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
4
5
6
9
/
i
j
a
c
s
a
.
2
0
2
1
.
0
1
2
0
4
2
2
.
[
1
6
]
T.
L
i
u
a
n
d
X
.
Y
u
a
n
,
“
P
a
r
a
l
i
n
g
u
i
st
i
c
a
n
d
s
p
e
c
t
r
a
l
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
f
o
r
sp
e
e
c
h
e
m
o
t
i
o
n
c
l
a
ss
i
f
i
c
a
t
i
o
n
u
s
i
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
EU
RA
S
I
P
J
o
u
rn
a
l
o
n
A
u
d
i
o
,
S
p
e
e
c
h
,
a
n
d
M
u
s
i
c
Pr
o
c
e
ssi
n
g
,
v
o
l
.
2
0
2
3
,
n
o
.
1
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
1
1
8
6
/
s1
3
6
3
6
-
023
-
0
0
2
9
0
-
x.
[
1
7
]
A
.
A
g
g
a
r
w
a
l
e
t
a
l
.
,
“
Tw
o
-
w
a
y
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
f
o
r
sp
e
e
c
h
e
mo
t
i
o
n
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
,
”
S
e
n
so
rs
,
v
o
l
.
2
2
,
n
o
.
6
,
p
.
2
3
7
8
,
M
a
r
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s2
2
0
6
2
3
7
8
.
[
1
8
]
S
.
S
e
k
k
a
t
e
,
M
.
K
h
a
l
i
l
,
a
n
d
A
.
A
d
i
b
,
“
A
st
a
t
i
st
i
c
a
l
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
f
o
r
d
e
e
p
s
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
i
n
a
b
i
l
i
n
g
u
a
l
s
c
e
n
a
r
i
o
,
”
Mu
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
8
2
,
n
o
.
8
,
p
p
.
1
1
4
4
3
–
1
1
4
6
0
,
O
c
t
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
0
4
2
-
0
2
2
-
1
4
0
5
1
-
z.
[
1
9
]
R
.
Ja
h
a
n
g
i
r
,
Y
.
W
.
Te
h
,
F
.
H
a
n
i
f
,
a
n
d
G
.
M
u
j
t
a
b
a
,
“
D
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s
f
o
r
sp
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
:
s
t
a
t
e
o
f
t
h
e
a
r
t
a
n
d
r
e
sea
r
c
h
c
h
a
l
l
e
n
g
e
s
,
”
M
u
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
0
,
n
o
.
1
6
,
p
p
.
2
3
7
4
5
–
2
3
8
1
2
,
Ja
n
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
0
4
2
-
020
-
0
9
8
7
4
-
7.
[
2
0
]
S
.
Jai
n
a
n
d
B
.
K
i
s
h
o
r
e
,
“
C
o
mp
a
r
a
t
i
v
e
st
u
d
y
o
f
v
o
i
c
e
p
r
i
n
t
B
a
s
e
d
a
c
o
u
st
i
c
f
e
a
t
u
r
e
s
:
M
F
C
C
a
n
d
LP
C
C
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
Ad
v
a
n
c
e
d
e
n
g
i
n
e
e
r
i
n
g
,
M
a
n
a
g
e
m
e
n
t
a
n
d
S
c
i
e
n
c
e
,
v
o
l
.
3
,
n
o
.
4
,
p
p
.
3
1
3
–
3
1
5
,
2
0
1
7
,
d
o
i
:
1
0
.
2
4
0
0
1
/
i
j
a
e
ms.
3
.
4
.
5
.
[
2
1
]
S
.
J
o
o
,
J.
C
h
o
i
,
N
.
K
i
m
,
a
n
d
M
.
C
.
Le
e
,
“
Ze
r
o
-
c
r
o
s
si
n
g
r
a
t
e
me
t
h
o
d
a
s
a
n
e
f
f
i
c
i
e
n
t
t
o
o
l
f
o
r
c
o
m
b
u
st
i
o
n
i
n
st
a
b
i
l
i
t
y
d
i
a
g
n
o
si
s,
”
Ex
p
e
ri
m
e
n
t
a
l
T
h
e
rm
a
l
a
n
d
F
l
u
i
d
S
c
i
e
n
c
e
,
v
o
l
.
1
2
3
,
p
.
1
1
0
3
4
0
,
M
a
y
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
x
p
t
h
e
r
mf
l
u
sc
i
.
2
0
2
0
.
1
1
0
3
4
0
.
[
2
2
]
S
.
C
u
n
n
i
n
g
h
a
m
,
H
.
R
i
d
l
e
y
,
J
.
W
e
i
n
e
l
,
a
n
d
R
.
P
i
c
k
i
n
g
,
“
S
u
p
e
r
v
i
se
d
mac
h
i
n
e
l
e
a
r
n
i
n
g
f
o
r
a
u
d
i
o
e
mo
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
:
E
n
h
a
n
c
i
n
g
f
i
l
m
s
o
u
n
d
d
e
si
g
n
u
si
n
g
a
u
d
i
o
f
e
a
t
u
r
e
s,
r
e
g
r
e
ss
i
o
n
mo
d
e
l
s
a
n
d
a
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s,”
P
e
rso
n
a
l
a
n
d
U
b
i
q
u
i
t
o
u
s
C
o
m
p
u
t
i
n
g
,
v
o
l
.
2
5
,
n
o
.
4
,
p
p
.
6
3
7
–
6
5
0
,
A
p
r
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
7
7
9
-
0
2
0
-
0
1
3
8
9
-
0.
[
2
3
]
R.
-
A
.
K
n
i
g
h
t
a
n
d
J
.
S
e
t
t
e
r
,
T
h
e
C
a
m
b
r
i
d
g
e
h
a
n
d
b
o
o
k
o
f
p
h
o
n
e
t
i
c
s
.
C
a
m
b
r
i
d
g
e
U
n
i
v
e
r
s
i
t
y
P
r
e
s
s
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
7
/
9
7
8
1
1
0
8
6
4
4
1
9
8
.
[
2
4
]
J.
M
.
K
.
K
u
a
,
T.
Th
i
r
u
v
a
r
a
n
,
M
.
N
o
s
r
a
t
i
g
h
o
d
s,
E.
A
mb
i
k
a
i
r
a
j
a
h
,
a
n
d
J.
E
p
p
s
,
“
I
n
v
e
s
t
i
g
a
t
i
o
n
o
f
sp
e
c
t
r
a
l
c
e
n
t
r
o
i
d
m
a
g
n
i
t
u
d
e
a
n
d
fr
e
q
u
e
n
c
y
f
o
r
s
p
e
a
k
e
r
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
O
d
y
ss
e
y
2
0
1
0
:
S
p
e
a
k
e
r
a
n
d
L
a
n
g
u
a
g
e
Re
c
o
g
n
i
t
i
o
n
W
o
rk
sh
o
p
,
2
0
1
0
,
p
p
.
3
4
–
3
9
.
[
2
5
]
Z.
H
u
a
n
g
,
M
.
D
o
n
g
,
Q
.
M
a
o
,
a
n
d
Y
.
Zh
a
n
,
“
S
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
u
s
i
n
g
C
N
N
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
2
2
n
d
AC
M
i
n
t
e
r
n
a
t
i
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
M
u
l
t
i
m
e
d
i
a
,
N
o
v
.
2
0
1
4
,
p
p
.
8
0
1
–
804
,
d
o
i
:
1
0
.
1
1
4
5
/
2
6
4
7
8
6
8
.
2
6
5
4
9
8
4
.
[
2
6
]
J.
L
e
e
a
n
d
I
.
Ta
s
h
e
v
,
“
H
i
g
h
-
l
e
v
e
l
f
e
a
t
u
r
e
r
e
p
r
e
se
n
t
a
t
i
o
n
u
s
i
n
g
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
sp
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
I
n
t
e
rs
p
e
e
c
h
2
0
1
5
,
S
e
p
.
2
0
1
5
,
d
o
i
:
1
0
.
2
1
4
3
7
/
i
n
t
e
r
sp
e
e
c
h
.
2
0
1
5
-
33
6.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
7
6
9
-
3778
3778
[
2
7
]
T.
A
n
v
a
r
j
o
n
,
M
u
st
a
q
e
e
m
,
a
n
d
S
.
K
w
o
n
,
“
D
e
e
p
-
N
e
t
:
a
l
i
g
h
t
w
e
i
g
h
t
C
N
N
-
b
a
se
d
s
p
e
e
c
h
e
mo
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
s
y
st
e
m
u
si
n
g
d
e
e
p
f
r
e
q
u
e
n
c
y
f
e
a
t
u
r
e
s,
”
S
e
n
so
rs
,
v
o
l
.
2
0
,
n
o
.
1
8
,
p
.
5
2
1
2
,
S
e
p
.
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
0
1
8
5
2
1
2
.
[
2
8
]
A
.
A
f
t
a
b
,
A
.
M
o
r
sa
l
i
,
S
.
G
h
a
e
mm
a
g
h
a
mi
,
a
n
d
B
.
C
h
a
m
p
a
g
n
e
,
“
LI
G
H
T
-
S
ER
N
ET
:
a
l
i
g
h
t
w
e
i
g
h
t
f
u
l
l
y
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
f
o
r
s
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
I
C
A
S
S
P
2
0
2
2
-
2
0
2
2
I
EE
E
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
Ac
o
u
s
t
i
c
s,
S
p
e
e
c
h
a
n
d
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
(
I
C
A
S
S
P)
,
M
a
y
2
0
2
2
,
p
p
.
6
9
1
2
–
6
9
1
6
,
d
o
i
:
1
0
.
1
1
0
9
/
i
c
a
ss
p
4
3
9
2
2
.
2
0
2
2
.
9
7
4
6
6
7
9
.
[
2
9
]
C
.
V
o
g
e
l
a
n
d
K
.
A
h
m
a
d
,
“
A
g
r
e
e
m
e
n
t
a
n
d
d
i
sa
g
r
e
e
me
n
t
b
e
t
w
e
e
n
m
a
j
o
r
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
s
y
st
e
ms,
”
K
n
o
w
l
e
d
g
e
-
Ba
s
e
d
S
y
s
t
e
m
s
,
v
o
l
.
2
7
6
,
p
.
1
1
0
7
5
9
,
S
e
p
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
k
n
o
s
y
s.
2
0
2
3
.
1
1
0
7
5
9
.
[
3
0
]
M
.
L.
M
c
H
u
g
h
,
“
I
n
t
e
r
r
a
t
e
r
r
e
l
i
a
b
i
l
i
t
y
:
t
h
e
K
a
p
p
a
st
a
t
i
s
t
i
c
,
”
B
i
o
c
h
e
m
i
a
Me
d
i
c
a
,
p
p
.
2
7
6
–
2
8
2
,
2
0
1
2
,
d
o
i
:
1
0
.
1
1
6
1
3
/
b
m.
2
0
1
2
.
0
3
1
.
[
3
1
]
D
.
M
.
N
o
g
u
e
i
r
a
,
C
.
A
.
F
e
r
r
e
i
r
a
,
E.
F
.
G
o
m
e
s,
a
n
d
A
.
M
.
J
o
r
g
e
,
“
C
l
a
ss
i
f
y
i
n
g
h
e
a
r
t
s
o
u
n
d
s
u
s
i
n
g
i
ma
g
e
s
o
f
m
o
t
i
f
s,
M
F
C
C
a
n
d
t
e
m
p
o
r
a
l
f
e
a
t
u
r
e
s
,
”
J
o
u
r
n
a
l
o
f
M
e
d
i
c
a
l
S
y
st
e
m
s
,
v
o
l
.
4
3
,
n
o
.
6
,
M
a
y
2
0
1
9
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
9
1
6
-
0
1
9
-
1
2
8
6
-
5.
[
3
2
]
S
.
D
.
W
a
g
h
mare
,
R
.
R
.
D
e
s
h
mu
k
h
,
P
.
P
.
S
h
r
i
s
h
r
i
ma
l
,
V
.
B
.
W
a
g
h
m
a
r
e
,
a
n
d
G
.
B
.
Ja
n
v
a
l
e
,
“
S
t
u
t
t
e
r
e
d
i
s
o
l
a
t
e
d
s
p
o
k
e
n
M
a
r
a
t
h
i
sp
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
b
y
u
s
i
n
g
M
F
C
C
a
n
d
LPC
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
n
o
v
a
t
i
o
n
s i
n
E
n
g
i
n
e
e
ri
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
8
,
n
o
.
3
,
2
0
1
7
,
d
o
i
:
1
0
.
2
1
1
7
2
/
i
j
i
e
t
.
8
3
.
0
1
8
.
[
3
3
]
M
.
M
.
R
e
z
a
p
o
u
r
M
a
s
h
h
a
d
i
a
n
d
K
.
O
sei
-
B
o
n
s
u
,
“
S
p
e
e
c
h
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s:
F
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
a
n
d
c
o
mp
a
r
i
s
o
n
o
f
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
a
n
d
r
a
n
d
o
m
f
o
r
e
st
,
”
PLO
S
O
N
E
,
v
o
l
.
1
8
,
n
o
.
1
1
,
p
.
e
0
2
9
1
5
0
0
,
N
o
v
.
2
0
2
3
,
d
o
i
:
1
0
.
1
3
7
1
/
j
o
u
r
n
a
l
.
p
o
n
e
.
0
2
9
1
5
0
0
.
[
3
4
]
H
.
K
.
P
a
l
o
,
“
T
h
e
e
f
f
e
c
t
o
f
a
g
e
,
g
e
n
d
e
r
,
a
n
d
a
r
o
u
s
a
l
l
e
v
e
l
o
n
c
a
t
e
g
o
r
i
z
i
n
g
h
u
ma
n
a
f
f
e
c
t
i
v
e
s
t
a
t
e
s,
”
i
n
Em
o
t
i
o
n
a
n
d
I
n
f
o
rm
a
t
i
o
n
Pro
c
e
ssi
n
g
,
S
p
r
i
n
g
e
r
I
n
t
e
r
n
a
t
i
o
n
a
l
P
u
b
l
i
sh
i
n
g
,
2
0
2
0
,
p
p
.
9
7
–
1
2
4
,
doi
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
0
-
4
8
8
4
9
-
9
_
7
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Iz
z
a
Nur
Afifa
h
re
c
e
iv
e
d
h
e
r
Ba
c
h
e
lo
r
o
f
Ap
p
li
e
d
S
c
ien
c
e
d
e
g
re
e
in
tele
c
o
m
m
u
n
ica
ti
o
n
e
n
g
i
n
e
e
rin
g
f
ro
m
P
o
l
it
e
k
n
ik
El
e
k
tr
o
n
i
k
a
Ne
g
e
ri
S
u
ra
b
a
y
a
,
In
d
o
n
e
sia
,
in
2
0
2
2
.
C
u
rre
n
tl
y
,
sh
e
is
p
u
rsu
in
g
a
M
a
ste
r
o
f
Ap
p
l
ied
S
c
ien
c
e
d
e
g
re
e
in
c
o
m
p
u
ter
a
n
d
in
fo
rm
a
ti
c
s
e
n
g
in
e
e
ri
n
g
a
t
th
e
sa
m
e
in
stit
u
ti
o
n
.
He
r
re
se
a
rc
h
i
n
ter
e
sts
d
u
ri
n
g
h
e
r
st
u
d
y
i
n
c
lu
d
e
sp
e
e
c
h
e
m
o
ti
o
n
re
c
o
g
n
it
io
n
a
n
d
m
a
c
h
in
e
lea
r
n
i
n
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
izz
a
n
u
ra
fifah
@g
m
a
il
.
c
o
m
.
Tr
i
Bu
d
i
S
a
n
to
s
o
re
c
e
iv
e
d
h
is
B.
E
n
g
.
d
e
g
re
e
in
e
n
g
in
e
e
rin
g
p
h
y
sic
s
,
h
is
M
.
T.
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
,
a
n
d
h
is
Dr.
d
e
g
re
e
fro
m
th
e
In
stit
u
t
Tek
n
o
lo
g
i
S
e
p
u
l
u
h
No
p
e
m
b
e
r,
S
u
ra
b
a
y
a
,
In
d
o
n
e
sia
,
in
1
9
9
4
,
1
9
9
9
,
a
n
d
2
0
1
6
,
re
sp
e
c
ti
v
e
ly
,
wit
h
re
se
a
rc
h
fo
c
u
se
d
o
n
sig
n
a
l
p
ro
c
e
ss
in
g
.
He
h
a
s
b
e
e
n
with
t
h
e
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
En
g
i
n
e
e
rin
g
a
t
th
e
El
e
c
tro
n
ic
En
g
in
e
e
rin
g
P
o
ly
tec
h
n
ic
In
stit
u
te
o
f
S
u
ra
b
a
y
a
(P
ENS
)
sin
c
e
1
9
9
5
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
tele
c
o
m
m
u
n
ica
ti
o
n
s
a
n
d
a
c
o
u
stic
sig
n
a
l
p
r
o
c
e
ss
in
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
tri
b
u
d
i@p
e
n
s.a
c
.
i
d
.
Tito
n
Duto
n
o
wa
s
b
o
rn
in
S
u
ra
b
a
y
a
,
I
n
d
o
n
e
sia
,
in
1
9
6
0
.
H
e
re
c
e
iv
e
d
a
B.
S
.
d
e
g
re
e
in
tele
c
o
m
m
u
n
ica
ti
o
n
e
n
g
in
e
e
rin
g
fro
m
S
e
p
u
l
u
h
No
p
e
m
b
e
r
In
stit
u
te
o
f
Tec
h
n
o
l
o
g
y
,
S
u
ra
b
a
y
a
,
I
n
d
o
n
e
sia
,
i
n
1
9
8
5
,
a
n
d
b
o
t
h
h
is
M
a
ste
r’s
a
n
d
Do
c
to
r
o
f
En
g
i
n
e
e
rin
g
d
e
g
re
e
s
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
a
n
d
c
o
m
p
u
ter
sc
ien
c
e
fro
m
Ku
m
a
m
o
to
Un
iv
e
rsity
,
Ku
m
a
m
o
t
o
,
Ja
p
a
n
,
in
1
9
9
4
a
n
d
1
9
9
7
,
re
sp
e
c
ti
v
e
ly
.
F
ro
m
2
0
0
2
t
o
2
0
0
8
,
h
e
se
rv
e
d
a
s
th
e
P
ri
n
c
ip
a
l
o
f
th
e
El
e
c
tro
n
ics
En
g
in
e
e
ri
n
g
P
o
ly
tec
h
n
ic
In
st
it
u
te
o
f
S
u
ra
b
a
y
a
(P
EN
S
).
F
ro
m
2
0
0
8
t
o
2
0
1
6
,
h
e
wa
s
a
p
p
o
i
n
ted
De
p
u
t
y
Dire
c
to
r
-
G
e
n
e
r
a
l
fo
r
S
p
e
c
tru
m
P
o
li
c
y
a
n
d
P
lan
n
in
g
a
t
th
e
M
in
istr
y
o
f
Co
m
m
u
n
ica
ti
o
n
a
n
d
In
f
o
rm
a
ti
o
n
Tec
h
n
o
l
o
g
y
,
Re
p
u
b
li
c
o
f
In
d
o
n
e
sia
.
He
wa
s
a
lso
in
c
h
a
rg
e
o
f
lea
d
in
g
th
e
In
d
o
n
e
sia
n
d
e
leg
a
ti
o
n
d
u
rin
g
ITU
re
g
u
lato
ry
m
e
e
ti
n
g
s
in
G
e
n
e
v
a
a
n
d
o
th
e
r
v
e
n
u
e
s.
S
in
c
e
2
0
1
7
,
h
e
h
a
s
re
t
u
rn
e
d
t
o
c
a
m
p
u
s
a
s
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
i
n
t
h
e
El
e
c
tri
c
a
l
En
g
i
n
e
e
rin
g
De
p
a
rtme
n
t
o
f
EE
P
IS
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
sig
n
a
l
p
r
o
c
e
ss
in
g
,
ra
d
io
c
o
m
m
u
n
ica
ti
o
n
,
tele
c
o
m
m
u
n
ica
ti
o
n
re
g
u
latio
n
,
a
n
d
tea
c
h
in
g
m
e
th
o
d
o
lo
g
y
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
ti
t
o
n
@
p
e
n
s.a
c
.
id
.
Evaluation Warning : The document was created with Spire.PDF for Python.