I
nte
rna
t
io
na
l
J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
201
8
,
p
p
.
3
7
2
~3
7
8
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v8
i
1
.
p
p
3
7
2
-
3
7
8
372
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
Dev
elo
p
m
e
nt
o
f
Q
ura
nic
Reci
ter
I
dentif
i
ca
tion Sy
st
e
m
using
MFCC
and
G
M
M
Clas
sifier
T
eddy
Su
ry
a
G
un
a
w
a
n
1
,
Nur
At
ik
a
h M
uh
a
m
a
t
Sa
leh
2
,
M
ir
a
K
a
rt
i
w
i
3
1
,2
De
p
a
rtm
e
n
t
o
f
El
e
c
tri
c
a
l
a
n
d
Co
m
p
u
ter E
n
g
in
e
e
rin
g
,
I
n
tern
a
ti
o
n
a
l
Isla
m
ic Un
iv
e
rsit
y
M
a
la
y
sia
,
M
a
la
y
sia
3
De
p
a
rtme
n
t
o
f
In
f
o
rm
a
ti
o
n
S
y
ste
m
s,
In
tern
a
ti
o
n
a
l
Isla
m
ic Un
iv
e
rs
it
y
M
a
la
y
sia
,
M
a
la
y
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
2
,
2
0
1
7
R
ev
i
s
ed
Dec
2
,
2
0
1
7
A
cc
ep
ted
Dec
1
6
,
2
0
1
7
No
w
a
d
a
y
s,
th
e
re
a
re
m
a
n
y
b
e
a
u
ti
f
u
l
re
c
it
a
ti
o
n
o
f
A
l
-
Qu
ra
n
a
v
a
il
a
b
le.
Qu
ra
n
ic
re
c
it
a
ti
o
n
h
a
s
it
s
o
w
n
c
h
a
ra
c
teristics
,
a
n
d
th
e
p
r
o
b
lem
to
i
d
e
n
ti
f
y
th
e
re
c
it
e
r
is
si
m
il
a
r
to
th
e
sp
e
a
k
e
r
re
c
o
g
n
it
io
n
/i
d
e
n
ti
f
ica
ti
o
n
p
r
o
b
lem
.
T
h
e
o
b
jec
ti
v
e
o
f
th
is
p
a
p
e
r
is
to
d
e
v
e
l
o
p
Qu
ra
n
re
c
it
e
r
id
e
n
ti
f
ica
ti
o
n
sy
ste
m
u
sin
g
M
e
l
-
f
re
q
u
e
n
c
y
Ce
p
stra
l
Co
e
ff
ici
e
n
t
(M
F
CC)
a
n
d
G
a
u
ss
ian
M
ix
t
u
re
M
o
d
e
l
(G
M
M
).
In
t
h
is
p
a
p
e
r,
a
d
a
ta
b
a
se
o
f
f
iv
e
Qu
ra
n
ic
re
c
it
e
rs
is
d
e
v
e
lo
p
e
d
a
n
d
u
se
d
in
train
in
g
a
n
d
tes
ti
n
g
p
h
a
se
s.
We
c
a
r
e
f
u
ll
y
r
a
n
d
o
m
ize
d
th
e
d
a
tab
a
se
f
ro
m
v
a
rio
u
s su
ra
h
in
th
e
Qu
ra
n
s
o
th
a
t
th
e
p
r
o
p
o
se
d
sy
st
e
m
w
il
l
n
o
t
p
ro
n
e
to
th
e
re
c
it
e
d
v
e
r
se
s
b
u
t
o
n
ly
to
th
e
re
c
it
e
r.
A
ro
u
n
d
1
5
Qu
ra
n
ic
a
u
d
i
o
sa
m
p
les
f
ro
m
5
re
c
it
e
rs
w
e
r
e
c
o
ll
e
c
ted
a
n
d
ra
n
d
o
m
ize
d
,
in
w
h
ich
1
0
sa
m
p
les
w
e
r
e
u
se
d
f
o
r
train
in
g
th
e
G
M
M
a
n
d
5
sa
m
p
les
we
re
u
se
d
f
o
r
tes
ti
n
g
.
Re
su
lt
s
sh
o
w
e
d
th
a
t
o
u
r
p
ro
p
o
se
d
sy
ste
m
h
a
s
1
0
0
%
re
c
o
g
n
it
io
n
ra
te
f
o
r
th
e
f
iv
e
re
c
it
e
rs
tes
ted
.
Ev
e
n
w
h
e
n
tes
t
e
d
w
it
h
u
n
k
n
o
w
n
sa
m
p
les
,
th
e
p
ro
p
o
se
d
s
y
ste
m
is
a
b
le t
o
re
jec
t
it
.
K
ey
w
o
r
d
:
GM
M
MFC
C
Qu
r
an
ic
r
ec
itat
io
n
Sp
ea
k
er
id
en
ti
f
icatio
n
Co
p
y
rig
h
t
©
2
0
1
8
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
T
ed
d
y
S
u
r
y
a
G
u
n
a
w
an
,
Dep
ar
t
m
en
t o
f
E
lectr
ical
an
d
C
o
m
p
u
ter
E
n
g
in
ee
r
i
n
g
,
I
n
ter
n
atio
n
al
I
s
la
m
ic
U
n
iv
er
s
it
y
Ma
la
y
s
ia,
J
alan
Go
m
b
ak
,
5
3
1
0
0
Ku
ala
L
u
m
p
u
r
,
Ma
la
y
s
ia
.
E
m
ail:
t
s
g
u
n
a
w
an
@
i
iu
m
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
Al
-
Q
u
r
an
i
s
th
e
h
o
l
y
b
o
o
k
o
f
Mu
s
l
i
m
s
w
h
ich
i
s
w
r
i
tten
an
d
r
ec
ited
in
A
r
ab
ic
lan
g
u
a
g
e.
I
n
ter
esti
n
g
l
y
,
A
l
-
Qu
r
a
n
is
th
e
m
o
s
t
p
o
p
u
lar
an
d
m
o
s
t r
ec
ited
b
o
o
k
o
f
all
ti
m
e
[1
]
,
[
2]
.
Mu
s
li
m
s
h
o
u
ld
tr
y
th
eir
b
est
to
av
o
id
m
i
s
tak
e
s
in
r
ec
itin
g
th
e
Q
u
r
an
,
s
u
ch
as
r
ec
itin
g
r
u
les
(
ta
jw
id
)
,
m
i
s
s
i
n
g
w
o
r
d
s
,
v
er
s
es,
m
is
r
ea
d
i
n
g
v
o
w
el
p
r
o
n
o
u
n
c
i
atio
n
s
,
p
u
n
ct
u
atio
n
s
,
a
n
d
ac
c
en
ts
[
3
]
.
R
ec
itatio
n
s
h
o
u
ld
f
o
llo
w
t
h
e
r
u
les
o
f
p
r
o
n
o
u
n
ciatio
n
,
in
to
n
atio
n
,
a
n
d
ca
es
u
r
as
e
s
tab
lis
h
ed
b
y
t
h
e
th
e
I
s
la
m
ic
p
r
o
p
h
et
Mu
h
a
m
m
ad
(
P
B
UH)
.
T
h
e
r
u
les
an
d
g
u
id
an
ce
to
r
ea
d
Qu
r
an
is
p
r
o
p
ag
ated
f
r
o
m
th
e
p
r
o
p
h
et
Mu
h
a
m
m
ad
u
n
til
th
e
Qu
r
a
n
ic
r
ec
iter
th
r
o
u
g
h
a
v
er
if
ied
c
h
ai
n
o
f
tr
an
s
m
i
s
s
io
n
(
s
a
n
a
d
)
.
Ma
n
y
n
o
n
-
A
r
ab
ic
p
eo
p
le
s
tu
d
ied
a
n
d
l
ea
r
n
t
A
l
-
Q
u
r
an
b
y
lis
te
n
in
g
to
th
e
w
ell
k
n
o
w
n
Qu
r
an
ic
r
ec
iter
s
(
q
a
r
i
)
.
Alth
o
u
g
h
e
ac
h
r
ec
iter
r
ec
ited
th
e
s
a
m
e
Qu
r
a
n
ic
v
er
s
es,
b
u
t
it
h
a
s
d
if
f
er
e
n
ce
s
d
u
e
to
th
eir
u
n
iq
u
e
v
o
ice
a
n
d
ch
ar
ac
ter
is
tics
.
T
o
id
en
tify
t
h
e
Q
u
r
an
ic
r
ec
iter
,
t
h
e
p
r
o
b
lem
is
s
i
m
ilar
to
th
e
s
p
ea
k
er
r
ec
o
g
n
itio
n
[
4
]
.
T
y
p
ical
s
p
ea
k
er
r
ec
o
g
n
it
io
n
s
y
s
te
m
i
n
cl
u
d
es
p
r
e
-
p
r
o
ce
s
s
i
n
g
,
f
ea
tu
r
e
e
x
tr
ac
tio
n
,
a
n
d
cla
s
s
i
f
icatio
n
[
5
]
.
Ma
n
y
f
ea
tu
r
e
s
an
d
c
lass
if
ier
s
h
a
v
e
b
ee
n
u
s
ed
in
th
e
s
p
e
ak
er
r
ec
o
g
n
itio
n
r
e
s
ea
r
ch
.
Au
d
io
f
ea
tu
r
es
s
u
ch
as
Mel
-
f
r
eq
u
en
c
y
C
ep
s
tr
al
C
o
e
f
f
icien
ts
[
4
]
,
[
6]
,
lin
ea
r
-
f
r
eq
u
en
c
y
ce
p
s
tr
al
co
ef
f
icie
n
ts
(
L
F
C
C
)
,
a
n
d
li
n
ea
r
p
r
ed
ictiv
e
co
ef
f
icien
ts
(
L
P
C
)
.
L
F
C
C
is
s
i
m
ilar
to
MF
C
C
ex
c
ep
t th
at
t
h
eir
f
r
eq
u
en
c
ies i
s
n
o
t
w
ar
p
ed
b
y
a
n
o
n
-
lin
ea
r
f
r
eq
u
e
n
c
y
s
ca
le
an
d
it
h
as
b
ee
n
f
o
u
n
d
th
at
L
FC
C
p
er
f
o
r
m
ed
b
etter
th
an
MF
C
C
i
n
f
e
m
ale
tr
ials
[
7
]
.
As s
tated
b
y
[
8
]
is
th
e
m
o
s
t c
o
m
m
o
n
l
y
u
s
ed
f
ea
t
u
r
es i
n
s
p
ea
k
er
r
ec
o
g
n
itio
n
.
Giv
e
n
a
s
e
t
o
f
f
ea
t
u
r
e
v
ec
to
r
s
,
ea
ch
s
p
ea
k
er
m
o
d
el
w
ill
b
e
b
u
ilt
s
o
th
a
t
a
v
ec
to
r
f
r
o
m
th
e
s
a
m
e
s
p
ea
k
er
h
a
s
h
i
g
h
er
p
r
o
b
ab
ilit
y
co
m
p
ar
ed
to
an
y
o
t
h
er
m
o
d
els.
Se
v
er
al
clas
s
i
f
er
s
h
av
e
b
ee
n
u
s
ed
,
s
u
c
h
as
k
-
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:
2
0
8
8
-
8708
Dev
elo
p
men
t o
f Q
u
r
a
n
ic
R
ec
it
er I
d
en
tifi
ca
tio
n
S
ystem
u
s
in
g
MFC
C
a
n
d
GMM …
.
(
Ted
d
y
S
u
r
ya
Gu
n
a
w
a
n
)
373
n
ea
r
est
n
ei
g
h
b
o
r
s
,
v
ec
to
r
q
u
an
tizatio
n
[
6
]
,
h
id
d
en
Ma
r
k
o
v
m
o
d
el
(
HM
M)
[
9
]
,
Gau
s
s
i
an
m
i
x
tu
r
e
m
o
d
el
(
GM
M)
[
1
0
]
,
ar
tif
icial
n
e
u
r
al
n
et
w
o
r
k
[
4
]
,
an
d
d
ee
p
n
eu
r
al
n
et
w
o
r
k
(
DNN)
[
1
1
]
.
Of
th
e
v
ar
io
u
s
cla
s
s
i
f
ier
s
av
ailab
le,
in
t
h
i
s
r
esear
ch
w
e
s
elec
ted
GM
M
as o
u
r
b
aselin
e
f
o
r
s
p
ea
k
er
r
ec
o
g
n
it
io
n
.
A
lt
h
o
u
g
h
m
an
y
r
esear
ch
e
s
h
a
v
e
b
ee
n
co
n
d
u
c
ted
o
n
s
p
ea
k
er
r
ec
o
g
n
itio
n
,
b
u
t
v
er
y
li
m
ited
ar
e
tar
g
eted
o
n
th
e
Qu
r
an
ic
r
ec
iter
r
ec
o
g
n
itio
n
.
R
ec
en
t
r
esear
ch
co
n
d
u
cted
b
y
[
1
2
]
s
tated
th
at
th
e
Qu
r
a
n
ic
r
ec
itatio
n
h
a
s
d
if
f
er
en
t
ch
ar
ac
ter
is
tic
s
co
m
p
ar
ed
to
t
h
e
E
n
g
lis
h
s
p
o
k
en
la
n
g
u
a
g
e.
T
h
e
Qu
r
a
n
ic
r
ec
itat
io
n
is
p
r
ed
o
m
i
n
a
n
tl
y
v
o
iced
s
p
ee
ch
,
i
n
w
h
ic
h
it
co
u
ld
p
o
ten
tiall
y
b
e
ex
p
lo
ited
to
b
u
il
d
m
o
r
e
e
f
f
ic
ien
t
s
p
ea
k
er
m
o
d
els.
T
h
er
ef
o
r
e,
th
e
o
b
j
ec
tiv
e
o
f
th
is
r
eseac
h
is
t
o
d
ev
elo
p
a
Qu
r
an
ic
r
ec
itatio
n
id
en
ti
f
icatio
n
u
s
i
n
g
MFC
C
an
d
GM
M,
an
d
to
ev
alu
a
te
its
p
er
f
o
r
m
an
ce
.
T
h
e
r
est
o
f
th
e
p
ap
er
is
o
r
g
an
ized
as
f
o
llo
w
s
:
Sectio
n
2
d
e
s
cr
ib
es
th
e
t
y
p
ica
l
co
m
p
o
n
e
n
t
s
i
n
a
s
p
ea
k
er
r
ec
o
g
n
i
tio
n
s
y
s
te
m
.
Sectio
n
3
ex
p
lain
s
t
h
e
p
r
o
p
o
s
ed
Qu
r
an
ic
r
ec
iter
id
en
ti
f
icatio
n
s
y
s
te
m
.
Sect
io
n
4
ev
alu
a
tes
its
p
er
f
o
r
m
an
ce
i
n
ter
m
s
o
f
r
ec
o
g
n
itio
n
r
ate,
w
h
ile
S
ec
tio
n
5
co
n
clu
d
es
th
is
p
ap
er
.
2.
SPEAK
E
R
RE
CO
G
N
I
T
I
O
N
T
h
e
f
lo
w
c
h
ar
t o
f
b
asic
m
o
d
el
f
o
r
r
ec
o
g
n
itio
n
s
p
ea
k
er
as
s
h
o
w
n
i
n
Fig
u
r
e
1
.
F
ir
s
t,
t
h
e
a
u
d
io
s
i
g
n
al
is
g
o
in
g
th
r
o
u
g
h
th
e
f
r
o
n
t
-
e
n
d
p
r
o
ce
s
s
in
g
,
i
n
w
h
ic
h
th
e
f
ea
t
u
r
es
th
at
co
u
ld
u
n
iq
u
el
y
r
ep
r
esen
t
th
e
s
p
ea
k
er
in
f
o
r
m
atio
n
ar
e
ex
tr
ac
ted
.
T
h
e
s
h
o
r
t
-
ti
m
e
s
p
ec
tr
al
is
th
e
m
o
s
t
-
f
r
eq
u
e
n
tl
y
u
s
ed
t
y
p
e
d
o
f
f
ea
tu
r
es
[
5
]
.
T
h
e
f
r
o
n
t
-
e
n
d
m
a
y
al
s
o
in
cl
u
d
e
p
r
e
-
p
r
o
ce
s
s
in
g
m
o
d
u
les,
s
u
ch
as
v
o
ice
ac
tiv
it
y
d
etec
tio
n
to
r
em
o
v
e
s
ile
n
ce
f
r
o
m
t
h
e
in
p
u
t,
o
r
a
ch
an
n
el
c
o
m
p
e
n
s
at
io
n
m
o
d
u
le
to
n
o
r
m
a
lize
th
e
e
f
f
ec
t o
f
th
e
r
ec
o
r
d
in
g
ch
an
n
el
[5
]
,
[
1
3
]
.
Fig
u
r
e
1
.
T
y
p
ical
Sp
ea
k
er
R
ec
o
g
n
itio
n
S
y
s
te
m
C
u
r
r
en
tl
y
,
t
h
er
e
ar
e
m
an
y
m
e
th
o
d
s
t
h
at
ca
n
b
e
u
s
ed
to
v
er
i
f
y
a
s
p
ea
k
er
id
e
n
tit
y
a
n
d
th
e
m
o
s
t
t
w
o
k
n
o
w
n
m
et
h
o
d
s
ar
e
lin
ea
r
p
r
ed
ictiv
e
co
d
in
g
(
L
P
C
)
a
n
d
Me
l
f
r
eq
u
e
n
c
y
ce
p
s
tr
u
m
(
MFC
C
)
[4
]
,
[
6]
.
Ho
w
e
v
er
,
in
t
h
is
p
ap
er
MFC
C
m
et
h
o
d
s
is
ch
o
o
s
en
a
s
th
e
f
ea
t
u
r
e
ex
t
r
ac
tio
n
s
in
ce
t
h
e
s
y
s
te
m
g
i
v
e
h
ig
h
er
ac
cu
r
a
n
c
y
.
MFC
C
i
s
t
h
e
m
o
s
t
p
o
p
u
lar
m
et
h
o
d
d
u
e
to
it
i
s
ea
s
y
to
m
o
d
er
ate
a
n
d
ca
n
h
an
d
le
m
u
ltip
le
s
p
ea
k
er
s
o
r
m
u
ltip
le
la
n
g
u
a
g
es.
A
v
ec
to
r
o
f
f
ea
t
u
r
es
ac
q
u
ir
ed
f
r
o
m
t
h
e
p
r
ev
io
u
s
s
tep
is
th
en
co
m
p
ar
ed
ag
ai
n
s
a
s
et
o
f
s
p
ea
k
e
r
m
o
d
el
s
.
T
h
e
id
en
tit
y
o
f
th
e
test
s
p
ea
k
er
is
a
s
s
o
ciate
d
w
it
h
th
e
id
e
n
tit
y
o
f
t
h
e
h
i
g
h
e
s
t
s
co
r
in
g
m
o
d
el.
A
s
p
ea
k
er
m
o
d
el
i
s
a
s
tati
s
tical
m
o
d
el
th
at
r
ep
r
esen
t
s
s
p
ea
k
er
-
d
ep
en
d
en
t i
n
f
o
r
m
a
tio
n
,
a
n
d
ca
n
b
e
u
s
ed
to
p
r
ed
ict
n
e
w
d
ata.
An
y
m
o
d
eli
n
g
tec
h
n
iq
u
es
ca
n
b
e
u
s
ed
,
b
u
t
t
h
e
m
o
s
t
p
o
p
u
lar
tech
n
iq
u
e
s
ar
e
:
clu
s
ter
in
g
,
h
id
d
en
Ma
r
k
o
v
m
o
d
el,
ar
tific
ial
n
e
u
r
al
n
et
w
o
r
k
,
an
d
Gau
s
s
ia
n
m
i
x
t
u
r
e
m
o
d
el
[4
]
,
[
5
]
,
[
14]
.
I
n
th
i
s
r
esear
ch
,
w
e
u
s
ed
GM
M
as
it
i
s
o
n
e
o
f
t
h
e
m
o
s
t
ef
f
ec
ti
v
e
tech
n
iq
u
es
i
n
s
p
ea
k
er
r
ec
o
g
n
itio
n
[
5
]
.
GM
M
u
s
ed
est
i
m
a
tio
n
m
ax
i
m
u
m
lo
g
-
l
ik
eli
h
o
o
d
alg
o
r
ith
m
to
f
in
d
t
h
e
p
atter
n
m
atc
h
i
n
g
a
n
d
is
ab
le
to
f
o
r
m
s
m
o
o
th
ap
p
r
o
x
im
a
tio
n
f
o
r
ar
b
itra
r
ily
s
h
ap
ed
d
en
s
it
ies.
3.
P
RO
P
O
SE
D
Q
URAN
I
C
RE
CIT
E
R
I
D
E
N
T
I
F
I
CA
T
I
O
N
SYST
E
M
Fig
u
r
e
2
illu
s
tr
ates
o
u
r
p
o
r
p
o
s
ed
s
y
s
te
m
f
o
r
Qu
r
a
n
ic
r
ec
ita
tio
n
id
en
ti
f
icatio
n
.
W
e
u
s
ed
MFC
C
f
o
r
f
ea
t
u
r
e
ex
tr
ac
tio
n
a
n
d
GM
M
f
o
r
class
if
ier
d
u
e
to
its
p
o
p
u
lar
i
t
y
an
d
e
f
f
ec
t
iv
e
n
es
s
f
o
r
s
p
ea
k
er
r
ec
o
g
n
itio
n
.
3
.
1
.
M
el
-
F
re
q
uency
Cepstr
a
l C
o
ef
f
icient
s
(
M
F
CCs)
MFC
C
s
u
s
e
a
n
o
n
-
li
n
ea
r
f
r
eq
u
en
c
y
s
ca
le,
i.e
.
m
e
l
s
ca
le,
b
ase
d
o
n
th
e
a
u
d
ito
r
y
p
er
ce
p
tio
n
.
A
me
l
is
a
u
n
i
t
o
f
m
ea
s
u
r
e
o
f
p
er
ce
i
v
ed
p
i
tch
o
r
f
r
eq
u
e
n
c
y
o
f
a
to
n
e
.
E
q
u
atio
n
(
1
)
ca
n
b
e
u
s
ed
to
co
n
v
er
t
f
r
eq
u
en
c
y
s
ca
le
to
m
el
s
ca
le.
700
1
ln
117
Hz
m
e
l
f
f
(
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
8
:
3
7
2
–
3
7
8
374
w
h
er
e
m
e
l
f
is
t
h
e
f
r
eq
u
e
n
c
y
in
m
el
s
an
d
Hz
f
is
th
e
n
o
r
m
al
f
r
eq
u
e
n
c
y
i
n
Hz.
MFC
C
s
ar
e
o
f
te
n
ca
lc
u
l
ated
u
s
i
n
g
a
f
il
ter
b
an
k
o
f
M
f
il
ter
s
,
i
n
wh
ich
ea
ch
f
ilter
h
as
a
tr
ian
g
u
la
r
s
h
ap
e
a
n
d
is
s
p
ac
ed
u
n
i
f
o
r
m
l
y
o
n
t
h
e
m
el
s
ca
le
as sh
o
w
n
i
n
E
q
u
atio
n
(
2
)
.
1
0
1
1
1
1
1
1
1
0
m
f
k
m
f
k
m
f
m
f
m
f
k
m
f
m
f
k
m
f
m
f
m
f
m
f
k
m
f
k
k
H
m
(
2
)
w
h
er
e
1
,
,
1
,
0
M
m
.
T
h
e
lo
g
-
en
er
g
y
m
el
s
p
ec
tr
u
m
i
s
th
e
n
ca
lc
u
lated
as
f
o
llo
w
s
:
1
,
,
1
,
0
ln
1
0
2
M
m
k
H
k
X
m
S
N
k
m
(
3
)
w
h
er
e
k
X
is
t
h
e
d
is
cr
ete
Fo
u
r
ier
t
r
an
s
f
o
r
m
(
DFT
)
o
f
a
s
p
ee
ch
in
p
u
t
n
x
.
A
lt
h
o
u
g
h
tr
ad
itio
n
al
ce
p
s
tr
u
m
u
s
es
i
n
v
er
s
e
d
is
cr
ete
Fo
u
r
ier
tr
an
s
f
o
r
m
(
I
DFT
)
,
m
el
f
r
eq
u
en
c
y
ce
p
s
tr
u
m
is
n
o
r
m
all
y
i
m
p
le
m
en
ted
u
s
in
g
d
is
cr
ete
co
s
i
n
e
tr
an
s
f
o
r
m
(
D
C
T
)
s
in
ce
m
S
is
ev
e
n
as
s
h
o
w
n
i
n
Eq
u
atio
n
(
4
)
,
as f
o
llo
w
s
:
1
,
,
1
,
0
2
1
c
o
s
ˆ
1
0
M
n
M
n
m
m
S
n
x
M
m
(
4
)
T
y
p
icall
y
,
t
h
e
n
u
m
b
er
o
f
f
ilte
r
s
M
r
an
g
es
f
r
o
m
2
0
to
4
0
,
an
d
th
e
n
u
m
b
er
o
f
k
ep
t
co
ef
f
i
ci
en
ts
i
s
1
3
.
So
m
e
r
esear
ch
r
ep
o
r
ted
th
at
th
e
p
er
f
o
r
m
a
n
ce
o
f
s
p
ee
ch
r
ec
o
g
n
itio
n
an
d
s
p
ea
k
er
id
en
ti
f
icatio
n
s
y
s
te
m
s
r
ea
ch
ed
p
ea
k
w
i
th
3
2
-
3
5
f
il
ter
s
[
8
]
.
Fig
u
r
e
2
.
T
h
e
p
r
o
p
o
s
ed
Qu
r
an
ic
r
ec
iter
id
en
tif
icat
io
n
s
y
s
te
m
3
.
2
.
G
a
us
s
ia
n M
ix
t
ure
M
o
del (
G
M
M
)
GM
M
p
r
o
v
id
es
a
p
r
o
b
ab
ilis
ti
c
m
o
d
el
o
f
a
s
p
ea
k
er
’
s
v
o
ice.
A
Gau
s
s
ian
m
i
x
t
u
r
e
d
is
tr
ib
u
tio
n
is
a
w
ei
g
h
ted
s
u
m
o
f
M
d
en
s
ities
:
M
i
i
i
x
b
p
x
p
1
|
(
5
)
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:
2
0
8
8
-
8708
Dev
elo
p
men
t o
f Q
u
r
a
n
ic
R
ec
it
er I
d
en
tifi
ca
tio
n
S
ystem
u
s
in
g
MFC
C
a
n
d
GMM …
.
(
Ted
d
y
S
u
r
ya
Gu
n
a
w
a
n
)
375
w
h
er
e
x
is
a
D
-
d
i
m
en
s
io
n
al
v
ec
to
r
,
x
b
i
is
t
h
e
i
-
th
co
m
p
o
n
en
t
d
en
s
it
y
,
an
d
i
p
is
th
e
w
ei
g
h
t
o
f
th
e
i
-
th
co
m
p
o
n
e
n
t.
T
h
e
m
ix
t
u
r
e
w
ei
g
h
ts
s
atis
f
y
1
1
M
i
i
p
.
E
ac
h
m
i
x
tu
r
e
co
m
p
o
n
en
t
is
a
D
-
v
ar
iate
Gau
s
s
ia
n
d
en
s
it
y
f
u
n
ctio
n
:
i
i
T
i
i
i
x
x
x
b
D
1
2
1
e
x
p
2
1
2
1
2
(
6
)
w
h
er
e
i
is
th
e
m
ea
n
v
ec
to
r
,
an
d
i
is
t
h
e
co
v
ar
ia
n
ce
m
atr
ix
.
A
GM
M
is
c
h
ar
ac
ter
ized
b
y
t
h
e
m
ea
n
v
ec
to
r
,
co
v
ar
ian
ce
m
atr
ix
,
a
n
d
w
ei
g
h
t
f
r
o
m
all
co
m
p
o
n
e
n
t
s
.
So
,
w
e
ca
n
r
ep
r
esen
t it
i
n
a
co
m
p
ac
t n
o
tatio
n
as f
o
llo
w
s
:
M
i
p
i
i
i
,
,
2
,
1
,
,
(
7
)
3
.
3
.
M
a
x
i
m
u
m
L
i
k
el
iho
o
d E
s
t
i
ma
t
io
n
Giv
e
n
a
s
et
o
f
tr
ai
n
i
n
g
s
a
m
p
le
s
X
,
t
h
e
m
o
s
t
p
o
p
u
lar
m
et
h
o
d
to
tr
ain
a
GM
M
i
s
m
a
x
i
m
u
m
l
ik
eli
h
o
o
d
esti
m
atio
n
.
T
h
e
lik
el
ih
o
o
d
o
f
a
GM
M
ca
n
b
e
d
ef
in
ed
as:
T
t
t
x
p
X
p
1
|
|
(
8
)
Ma
x
i
m
u
m
li
k
eli
h
o
o
d
p
ar
am
et
er
s
ar
e
n
o
r
m
all
y
es
ti
m
ated
u
s
in
g
t
h
e
e
x
p
ec
tatio
n
m
a
x
i
m
iza
tio
n
(
E
M)
alg
o
r
ith
m
.
Am
o
n
g
a
s
et
o
f
s
p
ea
k
er
s
ch
ar
ac
ter
ized
b
y
p
ar
a
m
eter
s
n
,
,
,
2
1
,
a
GM
M
s
y
s
te
m
m
ak
es
it
p
r
ed
ictio
n
b
y
r
etu
r
n
in
g
t
h
e
s
p
ea
k
er
th
at
m
a
x
i
m
izes
t
h
e
a
p
o
s
teri
o
r
i
p
r
o
b
ab
ilit
y
g
i
v
e
n
an
u
tter
an
ce
X
as
f
o
llo
w
s
:
X
P
P
X
P
X
P
s
k
k
k
n
k
|
|
ma
x
a
r
g
ˆ
1
(
9
)
I
f
p
r
io
r
p
r
o
b
ab
ilit
ies
o
f
all
s
p
ea
k
er
s
ar
e
eq
u
al,
k
P
n
k
,
1
,
s
in
ce
X
P
is
th
e
s
a
m
e
f
o
r
all
s
p
ea
k
er
s
,
w
e
ca
n
r
e
w
r
ite
E
q
u
a
tio
n
(
1
0
)
as f
o
llo
w
s
:
T
t
k
t
n
k
x
p
s
1
1
|
l
o
g
m
a
x
a
r
g
ˆ
(
1
0
)
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
I
n
th
is
s
ec
tio
n
,
w
e
w
ill
p
r
esen
t
th
e
ex
p
er
i
m
e
n
tal
s
et
u
p
an
d
Qu
r
an
ic
a
u
d
io
d
atab
ase,
ex
p
er
i
m
e
n
t
w
it
h
tr
ain
i
n
g
s
a
m
p
les,
ex
p
er
i
m
e
n
t
w
it
h
tes
tin
g
s
a
m
p
les,
ex
p
er
i
m
en
t
w
it
h
u
n
k
n
o
w
n
s
a
m
p
le
s
,
an
d
r
ec
o
g
n
itio
n
r
ate
ev
alu
a
tio
n
.
W
e
ca
r
ef
u
ll
y
r
a
n
d
o
m
ized
t
h
e
au
d
io
d
atab
ase
s
o
th
at
t
h
e
p
r
o
p
o
s
ed
s
y
s
te
m
is
n
o
t
p
r
o
n
e
to
th
e
r
ec
ited
v
er
s
es
b
u
t to
th
e
r
ec
ite
r
v
o
ice
o
n
l
y
.
4
.
1
.
E
x
peri
m
e
nta
l Set
up
a
nd
Q
u
ra
nic A
ud
io
Da
t
a
ba
s
e
A
h
ig
h
p
er
f
o
r
m
an
ce
s
y
s
te
m
was
u
s
ed
f
o
r
p
r
o
ce
s
s
in
g
,
i.e
.
a
m
u
ltico
r
e
s
y
s
te
m
w
it
h
I
n
tel
C
o
r
e
i7
6
7
0
0
K
4
.
0
0
GHz
(
4
co
r
es
w
it
h
8
th
r
ea
d
s
)
,
3
2
GB
y
tes
R
AM
,
2
5
6
GB
y
tes
SS
D
a
n
d
2
T
B
y
tes
h
ar
d
d
is
k
,
in
s
talled
w
it
h
t
h
e
lates
t v
er
s
io
n
o
f
W
in
d
o
w
s
1
0
6
4
-
b
its
o
p
er
atin
g
s
y
s
t
e
m
an
d
Ma
tlab
2
0
1
7
b
w
it
h
Si
g
n
al
P
r
o
ce
s
s
i
n
g
a
n
d
Neu
r
al
Net
w
o
r
k
T
o
o
lb
o
x
es.
T
h
e
au
d
io
d
atab
ase
i
n
t
h
is
r
es
ea
r
ch
w
er
e
d
o
w
n
lo
ad
ed
i
n
f
o
r
m
o
f
MP
3
f
r
o
m
t
h
e
i
n
ter
n
e
t
o
r
o
r
ig
in
a
l
l
y
f
r
o
m
C
D
au
d
io
f
o
r
f
i
v
e
r
ec
iter
s
,
i.e
.
A
b
d
u
l
B
asit
(
R
e
citer
A
)
,
A
b
d
u
r
ah
m
a
n
A
s
-
S
u
d
ais
(
R
ec
iter
B
)
,
Sau
d
Ash
-
S
h
u
r
a
y
m
(
R
ec
iter
C
)
,
S
h
eik
h
A
li
A
b
d
u
lr
ah
m
a
n
(
R
ec
iter
D)
a
n
d
S
h
ei
k
h
Said
al
-
G
h
a
m
d
i
(
R
ec
ite
r
E
)
.
Usi
n
g
Au
d
ac
it
y
s
o
f
t
w
ar
e
,
th
e
au
d
io
f
i
les
w
er
e
co
n
v
er
ted
f
r
o
m
MP
3
to
W
A
V
f
iles
w
it
h
8
0
0
0
Hz
s
a
m
p
lin
g
f
r
eq
u
en
c
y
,
m
o
n
o
,
an
d
w
er
e
cu
t
in
to
d
u
r
atio
n
o
f
6
0
s
ec
o
n
d
s
f
o
r
ea
ch
s
am
p
le.
T
h
e
d
atab
ase
is
d
iv
id
ed
in
to
t
w
o
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
8
:
3
7
2
–
3
7
8
376
p
ar
ts
,
in
cl
u
d
in
g
te
s
ti
n
g
a
n
d
tr
ain
i
n
g
s
a
m
p
les.
T
h
e
tr
ai
n
i
n
g
s
a
m
p
les
s
tar
ti
n
g
f
r
o
m
s
a
m
p
le
s
0
1
u
n
til
1
0
,
w
h
ile
test
i
n
g
s
a
m
p
le
s
f
r
o
m
n
u
m
b
er
1
1
u
n
til 1
5
.
T
a
b
le
1
s
h
o
w
s
t
h
e
Qu
r
an
ic
a
u
d
io
d
atab
ase
f
o
r
f
i
v
e
r
ec
iter
s
.
T
ab
le
1
.
Qu
r
an
ic
Au
d
io
Data
b
ase
f
o
r
Fiv
e
R
ec
iter
s
S
a
mp
l
e
s
R
e
c
i
t
e
r
A
R
e
c
i
t
e
r
B
R
e
c
i
t
e
r
C
R
e
c
i
t
e
r
D
R
e
c
i
t
e
r
E
01
Al
-
F
a
t
i
h
a
h
Al
-
M
u
n
a
f
i
q
u
n
Al
-
B
a
q
a
r
a
h
An
-
N
a
b
a
’
Al
-
F
a
t
i
h
a
h
02
Al
-
B
a
q
a
r
a
h
At
-
T
a
l
a
q
Al
-
F
a
j
r
An
-
N
a
z
i
’
a
t
Al
-
L
a
i
l
03
An
-
N
i
sa’
At
-
T
a
h
r
i
m
Al
-
G
h
a
sy
i
y
a
h
‘
A
b
a
sa
At
-
T
a
k
w
i
r
04
Al
-
I
mr
a
n
Al
-
M
u
r
sal
a
t
Al
-
B
u
r
u
j
At
-
T
a
k
w
i
r
Al
-
I
n
f
i
t
a
r
05
Al
-
M
a
’
i
d
a
h
Al
-
I
n
san
Al
-
I
n
s
y
i
q
a
q
Al
-
I
n
f
i
t
a
r
Al
-
M
u
t
a
f
f
i
f
i
n
06
Al
-
A
n
’
a
m
Al
-
M
u
d
d
a
t
h
t
h
i
r
Al
-
I
n
f
i
t
a
r
Al
-
M
u
mt
a
h
a
n
a
h
Al
-
I
n
s
y
i
q
a
q
07
Ar
-
R
a
’
d
Al
-
M
u
z
z
a
mi
l
‘
A
b
a
sa
Al
-
I
n
s
y
i
q
a
q
Al
-
B
u
r
u
j
08
I
b
r
a
h
i
m
Ji
n
An
-
N
a
z
i
’
a
t
Al
-
B
u
r
u
j
At
-
T
a
r
i
q
09
Al
-
H
i
j
r
N
u
h
N
u
h
At
-
T
a
r
i
q
N
u
h
10
M
a
r
y
a
m
Al
-
H
a
q
q
a
h
Al
-
H
a
q
q
a
h
Al
-
A
’
l
a
‘
A
b
a
sa
11
Al
-
G
h
a
sh
i
y
a
h
Al
-
Q
a
l
a
m
Al
-
Q
a
l
a
m
Al
-
G
h
a
sh
i
y
a
h
Al
-
M
u
z
a
mm
i
l
12
Al
-
K
a
h
f
i
At
-
T
a
g
h
a
b
u
n
At
-
T
a
h
r
i
m
Al
-
F
a
j
r
Al
-
Q
i
y
a
mah
13
Y
a
si
n
Al
-
Ju
mu
’
a
h
At
-
T
a
l
a
q
Al
-
B
a
l
a
d
Al
-
I
n
san
14
Al
-
M
u
l
k
As
-
S
a
f
f
At
-
T
a
g
h
a
b
u
n
Ad
-
D
h
u
h
a
An
-
N
a
b
a
’
15
Y
u
n
u
s
Al
-
M
u
l
k
Al
-
M
u
l
k
A
sh
-
S
h
a
ms
An
-
N
a
z
i
’
a
t
4
.
2
.
E
x
peri
m
e
nt
w
it
h t
he
T
ra
ini
ng
Sa
m
ples
I
n
th
is
e
x
p
er
i
m
e
n
t,
th
e
s
a
m
p
l
es
0
1
u
n
til
1
0
f
o
r
ea
ch
r
ec
iter
w
er
e
u
s
ed
to
tr
ain
th
e
GM
M.
T
h
e
s
a
m
e
s
a
m
p
les
w
er
e
th
e
n
w
er
e
u
s
ed
to
ev
alu
ate
t
h
e
r
ec
o
g
n
itio
n
r
ate
o
f
th
e
tr
ai
n
ed
GM
M.
T
h
e
lo
g
-
li
k
eli
h
o
o
d
w
as
ca
lcu
lated
f
o
r
ea
ch
r
ec
iter
a
n
d
ea
ch
tr
ain
in
g
s
a
m
p
le
s
,
i
n
w
h
ic
h
th
e
h
i
g
h
est
li
k
eli
h
o
o
d
as
s
elec
ted
as
th
e
r
ec
o
g
n
ized
r
ec
iter
.
T
ab
le
2
s
h
o
w
s
th
e
r
ec
o
g
n
i
tio
n
r
ate
o
f
t
h
e
tr
ain
i
n
g
s
a
m
p
les.
I
t
s
h
o
w
s
t
h
at
ea
c
h
s
a
m
p
les
f
o
r
ea
ch
r
ec
iter
s
w
er
e
id
en
ti
f
ied
co
r
r
ec
tly
.
I
t
m
ea
n
s
t
h
at
t
h
e
r
ec
o
g
n
i
tio
n
r
ate
f
o
r
t
h
e
tr
ain
i
n
g
p
h
ase
w
a
s
1
0
0
%.
T
a
b
le
2.
R
e
c
o
g
n
it
i
o
n
R
at
e
o
f
T
r
a
i
n
i
n
g
S
am
p
l
es
R
e
c
i
t
e
r
S
a
mp
l
e
s
01
02
03
04
05
06
07
08
09
10
A
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
B
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
C
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
D
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
E
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
*Y
=
m
a
t
c
h
e
d
,
N=
u
n
m
a
t
c
h
e
d
4
.
3
.
E
x
peri
m
e
nt
w
it
h t
he
T
esting
Sa
m
ples
I
n
th
is
e
x
p
er
i
m
e
n
t,
th
e
p
r
ev
i
o
u
s
tr
ain
ed
GM
M
in
s
ec
tio
n
4
.
2
w
as
u
s
ed
to
test
d
if
f
er
e
n
t
s
a
m
p
le
s
.
T
h
e
s
a
m
p
les
1
1
u
n
til
1
5
w
er
e
u
s
ed
to
ev
a
lu
ate
t
h
e
r
ec
o
g
n
itio
n
r
ate
o
f
t
h
e
tr
ain
ed
G
M
M
w
h
e
n
test
ed
w
i
th
d
if
f
er
e
n
t
s
a
m
p
le
s
f
r
o
m
t
h
e
s
a
m
e
r
ec
iter
s
.
T
h
e
lo
g
-
lik
e
lih
o
o
d
w
a
s
ca
lc
u
lated
f
o
r
ea
c
h
r
ec
i
ter
an
d
ea
ch
test
i
n
g
s
a
m
p
les,
i
n
w
h
ic
h
t
h
e
h
i
g
h
e
s
t
lik
eli
h
o
o
d
as
s
elec
ted
as
th
e
r
ec
o
g
n
ized
r
ec
iter
.
T
a
b
le
3
s
h
o
w
s
t
h
e
r
ec
o
g
n
itio
n
r
ate
o
f
t
h
e
te
s
ti
n
g
s
a
m
p
les.
I
t
s
h
o
w
s
th
at
ea
ch
s
a
m
p
l
es
f
o
r
ea
ch
r
ec
iter
s
w
er
e
i
d
en
tifie
d
co
r
r
ec
tl
y
.
I
t
m
ea
n
s
th
at
t
h
e
r
ec
o
g
n
itio
n
r
ate
f
o
r
th
e
test
in
g
p
h
ase
w
a
s
1
0
0
% a
s
w
ell.
T
a
b
le
3.
R
e
c
o
g
n
it
i
o
n
R
at
e
o
f
T
esti
n
g
S
am
p
les
R
e
c
i
t
e
r
S
a
m
p
e
l
s
1
1
1
2
1
3
1
4
1
5
A
Y
Y
Y
Y
Y
B
Y
Y
Y
Y
Y
C
Y
Y
Y
Y
Y
D
Y
Y
Y
Y
Y
E
Y
Y
Y
Y
Y
*Y
=
m
a
t
c
h
e
d
,
N=
u
n
m
a
t
c
h
e
d
4
.
4
.
E
x
peri
m
e
nt
w
it
h Unkn
o
w
n Sa
m
ples
T
h
e
last
ex
p
er
i
m
en
t
is
co
n
d
u
cte
d
to
e
v
a
l
u
a
t
e
th
e
p
r
o
p
o
s
e
d
s
y
s
te
m
w
h
eth
er
it
can
d
ete
ct
u
n
k
n
o
w
n
s
p
e
ak
er
w
h
ich
is
n
o
t
in
th
e
d
ata
b
as
e.
T
h
e
d
atab
ase
ar
e
co
l
l
ec
t
ed
f
o
r
f
i
v
e
r
ec
it
er
o
n
l
y
w
h
i
ch
ar
e
A
b
d
u
l
B
asit
,
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:
2
0
8
8
-
8708
Dev
elo
p
men
t o
f Q
u
r
a
n
ic
R
ec
it
er I
d
en
tifi
ca
tio
n
S
ystem
u
s
in
g
MFC
C
a
n
d
GMM …
.
(
Ted
d
y
S
u
r
ya
Gu
n
a
w
a
n
)
377
A
b
d
u
r
ah
m
an
As
-
S
u
d
ais,
Sau
d
As
h
-
Sh
u
r
ay
m
,
Sh
eik
h
A
li
Ab
d
u
lr
ah
m
an
an
d
Sh
eik
h
Said
al
-
Gh
am
d
i.
F
o
r
t
h
i
s
p
u
r
p
s
e
,
o
ne
u
n
k
n
o
w
n
r
e
c
i
t
e
r
n
a
m
ed
Fati
h
is
test
ed
u
s
i
n
g
t
h
e
t
r
a
i
n
e
d
G
M
M
a
n
d
t
h
e
t
r
a
i
n
i
n
g
s
a
m
p
l
e
s
.
T
h
e
m
a
x
i
m
u
m
esti
m
atio
n
l
o
g
-
lik
elih
o
o
d
of
u
n
k
n
o
w
n
s
p
ea
k
e
r
d
o
es
n
o
t
m
atch
ed
w
it
h
a
n
y
p
ar
am
eter
of
th
e
r
ec
ite
r
in
th
e
d
at
a
b
ase
.
B
ec
au
s
e
of
th
at,
t
h
e
r
esu
lt
s
h
o
w
s
t
h
at
Fatih
is
r
e
co
g
n
iz
e
d
as
an
u
n
k
n
o
w
n
s
p
e
ak
er
as
s
o
w
n
in
Fig
u
r
e
4
.
T
a
b
l
e
4
.
R
e
s
u
l
t
o
f
te
s
t
in
g
w
i
th
u
n
k
n
o
w
n
r
ec
i
t
e
r
S
p
e
a
k
e
r
S
a
mp
l
e
s
01
02
03
04
05
06
07
08
09
10
A
N
N
N
N
N
N
N
N
N
N
B
N
N
N
N
N
N
N
N
N
N
C
N
N
N
N
N
N
N
N
N
N
D
N
N
N
N
N
N
N
N
N
N
E
N
N
N
N
N
N
N
N
N
N
*Y
=
m
a
t
c
h
e
d
,
N=
u
n
m
a
t
c
h
e
d
4
.
5
.
Rec
o
g
nitio
n Ra
t
e
E
v
a
lua
t
io
n
It
can
be
co
n
clu
d
ed
t
h
at,
th
r
e
e
of
t
h
e
ex
p
e
r
im
en
ts
ar
e
s
u
cc
ess
f
u
l
l
y
co
n
d
u
cte
d
.
T
h
is
r
e
s
u
l
t
s
s
h
o
w
ed
th
at
th
e
p
r
o
p
o
s
e
d
s
y
s
te
m
w
a
s
ab
l
e
to
v
er
if
y
an
d
id
en
tify
t
h
e
t
r
a
i
n
e
d
r
ec
iter
s
.
A
lth
o
u
g
h
t
h
e
r
ec
it
er
w
a
s
r
a
n
d
o
m
l
y
r
ec
itin
g
d
if
f
er
en
t
Su
r
ah
in
ea
ch
s
am
p
les
,
b
u
t
th
e
s
y
s
te
m
s
t
ill
can
r
e
co
g
n
i
ze
th
e
p
att
er
n
of
th
e
r
ec
i
te
r
’
s
r
ec
i
tat
io
n
.
F
u
r
t
h
e
m
o
r
e
,
t
h
e
p
r
o
p
o
s
e
d
s
y
s
t
e
m
wa
s
a
l
s
o
a
b
l
e
t
o
r
e
j
e
c
t
t
h
e
u
n
k
n
o
w
n
r
e
c
i
t
e
r
t
e
s
t
e
d
.
T
a
b
le
5
s
h
o
w
t
h
e
r
ec
o
g
n
iti
o
n
r
a
te
f
o
r
ea
ch
ex
p
er
im
en
ts
.
T
h
is
r
esu
lt
is
b
ett
er
th
an
th
e
r
esu
lt
r
e
p
o
r
te
d
in
[
4
]
,
in
w
h
ich
th
ey
o
b
tain
e
d
a
r
o
u
n
d
9
1
%
ac
cu
r
a
cy
u
s
in
g
MFC
C
an
d
A
NN.
B
e
tte
r
r
esu
lts
ac
h
iev
e
d
in
th
is
p
a
p
e
r
co
u
l
d
b
e
d
u
e
t
o
th
e
u
s
e
o
f
G
MM
in
s
tea
d
o
f
A
NN,
an
d
th
e
r
an
d
o
m
ize
Qu
r
an
ic
au
d
i
o
d
ata
b
as
e
s
o
th
at
th
e
p
r
o
p
o
s
e
d
s
y
s
tem
is
n
o
t
p
r
o
n
e
t
o
th
e
r
e
cit
e
d
Qu
r
an
i
c
v
e
r
s
es
b
u
t
ab
le
to
d
is
tin
g
u
is
h
th
e
ch
ar
ac
t
er
is
ti
cs
o
f
in
d
iv
i
d
u
al
r
ec
ite
r
.
T
a
b
l
e
5
.
Pe
r
f
o
r
m
an
c
e
o
f
th
e
s
y
s
t
em
w
ith
d
if
f
e
r
en
t
s
am
p
l
es
Ex
p
e
r
i
me
n
t
R
e
c
o
g
n
i
t
i
o
n
R
a
t
e
T
e
st
i
n
g
w
i
t
h
t
r
a
i
n
i
n
g
s
a
m
p
l
e
s
1
0
0
%
m
a
t
c
h
,
0
%
m
i
s
m
a
t
c
h
T
e
st
i
n
g
w
i
t
h
t
e
s
t
i
n
g
s
a
m
p
l
e
s
1
0
0
%
m
a
t
c
h
,
0
%
m
i
s
m
a
t
c
h
T
e
st
i
n
g
w
i
t
h
u
n
k
n
o
w
n
r
e
c
i
t
e
r
0
%
m
a
t
c
h
,
1
0
0
%
m
i
sm
a
t
c
h
5.
CO
NCLU
SI
O
NS A
ND
F
UT
URE WO
RK
S
T
h
is
p
ap
er
h
as
p
r
esen
ted
th
e
d
ev
elo
p
m
en
t
o
f
Q
u
r
an
ic
r
ec
iter
id
en
tif
icatio
n
s
y
s
te
m
u
s
in
g
MFC
C
an
d
GM
M.
MFC
C
w
as
s
elec
ted
as
th
e
f
ea
tu
r
es,
w
h
ile
GM
M
is
s
elec
ted
as
th
e
clas
s
i
f
er
.
First,
w
e
b
u
ild
a
Qu
r
a
n
ic
au
d
io
d
atab
ase
f
r
o
m
f
i
v
e
r
ec
it
er
s
,
in
w
h
ic
h
th
e
y
r
ec
ite
d
i
f
f
er
en
t s
u
r
ah
f
r
o
m
Al
-
Qu
r
a
n
.
A
lto
g
eth
er
,
t
h
er
e
ar
e
1
5
s
a
m
p
les
co
llected
f
o
r
ea
ch
r
ec
iter
,
in
w
h
ic
h
1
0
s
a
m
p
le
s
w
er
e
u
s
ed
to
tr
ai
n
t
h
e
GM
M
a
n
d
5
s
a
m
p
les
w
er
e
u
s
ed
f
o
r
test
i
n
g
.
Fu
r
t
h
er
m
o
r
e,
w
e
u
s
e
a
n
o
th
er
u
n
k
n
o
w
n
r
ec
ite
r
to
ev
alu
ate
th
e
p
er
f
o
r
m
an
c
e
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
te
m
.
R
es
u
lts
s
h
o
w
ed
th
a
t
o
u
r
p
r
o
p
o
s
ed
s
y
s
te
m
ac
h
ie
v
ed
1
0
0
%
ac
cu
r
ac
y
i
n
th
e
tr
ai
n
i
n
g
an
d
test
in
g
p
h
ase
.
T
h
e
u
n
k
n
o
w
n
s
a
m
p
le
s
w
er
e
also
ac
h
iev
ed
1
0
0
%
r
ej
ec
tio
n
r
ate.
Fu
r
th
er
r
esear
ch
i
n
cl
u
d
es
v
ar
iatio
n
o
f
s
h
o
r
ter
u
tter
an
ce
o
f
th
e
r
ec
ited
Qu
r
a
n
i
c
v
er
s
es,
d
if
f
er
en
t r
ec
iter
s
,
d
if
f
er
en
t f
ea
t
u
r
es,
o
r
d
if
f
er
e
n
t c
las
s
if
ier
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
e
au
th
o
r
s
w
o
u
ld
li
k
e
to
e
x
p
r
ess
th
eir
g
r
atit
u
d
e
to
t
h
e
M
ala
y
s
ia
n
Mi
n
i
s
tr
y
o
f
Hi
g
h
er
E
d
u
ca
tio
n
(
MO
HE
)
,
w
h
ic
h
h
a
s
p
r
o
v
id
ed
f
u
n
d
in
g
f
o
r
th
e
r
e
s
ea
r
ch
t
h
r
o
u
g
h
th
e
Fu
n
d
a
m
en
ta
l
R
e
s
e
ar
ch
Gr
an
t
Sch
e
m
e,
FR
GS1
5
-
194
-
0435.
RE
F
E
R
E
NC
E
S
[1
]
B.
L
a
w
re
n
c
e
,
T
h
e
Qu
ra
n
:
A
Bi
o
g
ra
p
h
y
(
A
Bo
o
k
t
h
a
t
S
h
o
o
k
th
e
W
o
rld
)
,
A
tl
a
n
ti
c
Bo
o
k
s L
td
,
2
0
1
4
.
[2
]
T
.
Eco
n
o
m
ist,
"
T
h
e
Bib
le
v
s
th
e
Ko
ra
n
:
T
h
e
b
a
tt
le
o
f
th
e
b
o
o
k
s,"
[
h
tt
p
:/
/www
.
e
c
o
n
o
m
ist.
c
o
m
/n
o
d
e
/1
0
3
1
1
3
1
7
],
Re
tri
e
v
e
d
o
n
:
1
3
N
o
v
2
0
1
7
.
[3
]
I.
A
ls
m
a
d
i
a
n
d
M
.
Zaro
u
r
,
"
On
li
n
e
in
teg
rit
y
a
n
d
a
u
t
h
e
n
ti
c
a
ti
o
n
c
h
e
c
k
in
g
f
o
r
Qu
ra
n
e
lec
tro
n
ic
v
e
rsio
n
s,"
Ap
p
li
e
d
Co
mp
u
t
in
g
a
n
d
In
f
o
rm
a
ti
c
s
,
v
o
l
.
1
3
,
p
p
.
3
8
-
4
6
,
2
0
1
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
8
:
3
7
2
–
3
7
8
378
[4
]
T
.
M
.
H.
A
sd
a
,
T
.
S
.
G
u
n
a
w
a
n
,
M
.
Ka
rti
w
i,
a
n
d
H.
M
a
n
s
o
r,
"
De
v
e
lo
p
m
e
n
t
o
f
Qu
ra
n
Re
c
it
e
r
Id
e
n
ti
f
ica
ti
o
n
S
y
ste
m
Us
in
g
M
F
CC
a
n
d
Ne
u
ra
l
Ne
tw
o
rk
,
"
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
E
lec
trica
l
En
g
in
e
e
rin
g
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
1
,
p
p
.
1
6
8
-
1
7
5
,
2
0
1
6
.
[5
]
H.
Be
ig
i,
Fu
n
d
a
me
n
ta
ls
o
f
s
p
e
a
k
e
r re
c
o
g
n
it
i
o
n
,
S
p
rin
g
e
r
S
c
ien
c
e
&
Bu
sin
e
ss
M
e
d
ia,
2
0
1
1
.
[6
]
P
.
Ba
n
sa
l,
S
.
A
.
I
m
a
m
,
a
n
d
R.
Bh
a
rti
,
"
S
p
e
a
k
e
r
re
c
o
g
n
it
i
o
n
u
sin
g
M
FCC,
sh
if
ted
M
FCC
wit
h
v
e
c
to
r
q
u
a
n
ti
za
ti
o
n
a
n
d
f
u
zz
y
,
"
i
n
S
o
f
t
Co
m
p
u
ti
n
g
T
e
c
h
n
iq
u
e
s
a
n
d
Im
p
le
m
e
n
tatio
n
s
(
ICS
CT
I),
2
0
1
5
In
tern
a
ti
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
,
p
p
.
41
-
4
4
,
2
0
1
5
.
[7
]
X
.
Zh
o
u
,
D.
G
a
rc
i
a
-
Ro
m
e
ro
,
R.
Du
ra
iswa
m
i,
C.
Esp
y
-
W
il
so
n
,
a
n
d
S
.
S
h
a
m
m
a
,
"
L
in
e
a
r
v
e
rs
u
s
me
l
fre
q
u
e
n
c
y
c
e
p
stra
l
c
o
e
ff
icie
n
ts
fo
r
sp
e
a
k
e
r
r
e
c
o
g
n
it
i
o
n
,
"
in
A
u
to
m
a
ti
c
S
p
e
e
c
h
Re
c
o
g
n
it
io
n
a
n
d
Un
d
e
rsta
n
d
i
n
g
(A
S
RU),
2
0
1
1
IEE
E
W
o
rk
sh
o
p
o
n
,
p
p
.
5
5
9
-
5
6
4
,
2
0
1
1
.
[8
]
V
.
T
iw
a
ri,
"
M
F
CC
a
n
d
it
s
a
p
p
li
c
a
ti
o
n
s
in
sp
e
a
k
e
r
re
c
o
g
n
it
io
n
,
"
In
ter
n
a
t
io
n
a
l
jo
u
rn
a
l
o
n
e
me
rg
in
g
tec
h
n
o
lo
g
ies
,
v
o
l.
1
,
p
p
.
1
9
-
2
2
,
2
0
1
0
.
[9
]
A
.
P
o
rit
z
,
"
L
in
e
a
r
p
re
d
ictive
h
i
d
d
e
n
M
a
rk
o
v
mo
d
e
ls
a
n
d
th
e
sp
e
e
c
h
sig
n
a
l,
"
in
A
c
o
u
stics
,
S
p
e
e
c
h
,
a
n
d
S
ig
n
a
l
P
r
o
c
e
ss
in
g
,
IEE
E
I
n
tern
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
ICA
S
S
P
'
8
2
.
,
p
p
.
1
2
9
1
-
1
2
9
4
,
1
9
8
2
.
[1
0
]
L
.
L
i,
D.
W
a
n
g
,
C.
Zh
a
n
g
,
a
n
d
T
.
F
.
Zh
e
n
g
,
"
Im
p
ro
v
in
g
sh
o
rt
u
t
tera
n
c
e
sp
e
a
k
e
r
re
c
o
g
n
it
io
n
b
y
m
o
d
e
li
n
g
sp
e
e
c
h
u
n
it
c
las
se
s,"
IEE
E/
ACM
T
ra
n
sa
c
ti
o
n
s
o
n
Au
d
io
,
S
p
e
e
c
h
a
n
d
L
a
n
g
u
a
g
e
Pro
c
e
ss
in
g
(
T
AS
L
P)
,
v
o
l
.
2
4
,
p
p
.
1
1
2
9
-
1
1
3
9
,
2
0
1
6
.
[1
1
]
F
.
Rich
a
rd
s
o
n
,
D.
Re
y
n
o
ld
s,
a
n
d
N.
De
h
a
k
,
"
De
e
p
n
e
u
ra
l
n
e
tw
o
rk
a
p
p
ro
a
c
h
e
s
to
sp
e
a
k
e
r
a
n
d
lan
g
u
a
g
e
re
c
o
g
n
it
io
n
,
"
IEE
E
S
ig
n
a
l
Pro
c
e
ss
in
g
L
e
t
ter
s
,
v
o
l.
2
2
,
p
p
.
1
6
7
1
-
1
6
7
5
,
2
0
1
5
.
[1
2
]
T
.
S
.
G
u
n
a
w
a
n
a
n
d
M
.
Ka
rti
w
i,
"
On
th
e
Ch
a
ra
c
ter
isti
c
s
o
f
Va
ri
o
u
s
Qu
ra
n
ic
Rec
it
a
ti
o
n
fo
r
L
o
ss
les
s
Au
d
io
Co
d
in
g
Ap
p
li
c
a
ti
o
n
,
"
i
n
Co
m
p
u
ter
a
n
d
C
o
m
m
u
n
ica
ti
o
n
E
n
g
in
e
e
rin
g
(ICCCE),
2
0
1
6
In
ter
n
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
,
p
p
.
1
2
1
-
1
2
5
,
2
0
1
6
.
[1
3
]
N.
De
h
a
k
,
P
.
J.
Ke
n
n
y
,
R.
De
h
a
k
,
P
.
Du
m
o
u
c
h
e
l,
a
n
d
P
.
Ou
e
ll
e
t,
"
F
ro
n
t
-
e
n
d
f
a
c
to
r
a
n
a
ly
sis
f
o
r
sp
e
a
k
e
r
v
e
ri
f
ica
ti
o
n
,
"
IEE
E
T
r
a
n
sa
c
ti
o
n
s
o
n
A
u
d
io
,
S
p
e
e
c
h
,
a
n
d
L
a
n
g
u
a
g
e
Pro
c
e
ss
in
g
,
v
o
l
.
1
9
,
p
p
.
7
8
8
-
7
9
8
,
2
0
1
1
.
[1
4
]
J.
P
.
Ca
m
p
b
e
ll
,
"
S
p
e
a
k
e
r
re
c
o
g
n
it
io
n
:
A
tu
to
rial,
"
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
IEE
E
,
v
o
l.
8
5
,
p
p
.
1
4
3
7
-
1
4
6
2
,
1
9
9
7
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
Te
d
d
y
S
u
r
y
a
G
u
n
a
w
a
n
re
c
e
i
v
e
d
h
is
BEn
g
d
e
g
re
e
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
w
it
h
c
u
m
lau
d
e
a
wa
rd
f
ro
m
In
stit
u
t
T
e
k
n
o
lo
g
i
Ba
n
d
u
n
g
(IT
B),
In
d
o
n
e
sia
in
1
9
9
8
.
He
o
b
tain
e
d
h
is
M
.
E
n
g
d
e
g
re
e
in
2
0
0
1
f
ro
m
th
e
S
c
h
o
o
l
o
f
C
o
m
p
u
ter
En
g
in
e
e
rin
g
a
t
Na
n
y
a
n
g
T
e
c
h
n
o
lo
g
ica
l
Un
iv
e
rsity
,
S
in
g
a
p
o
re
,
a
n
d
P
h
D
d
e
g
re
e
i
n
2
0
0
7
f
ro
m
th
e
S
c
h
o
o
l
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
a
n
d
T
e
le
c
o
m
m
u
n
ica
ti
o
n
s,
T
h
e
Un
iv
e
r
sity
o
f
N
e
w
S
o
u
th
W
a
les
,
A
u
stra
li
a
.
His
re
se
a
rc
h
in
tere
sts
a
re
in
sp
e
e
c
h
a
n
d
a
u
d
i
o
p
r
o
c
e
ss
in
g
,
b
io
m
e
d
ica
l
sig
n
a
l
p
ro
c
e
ss
in
g
a
n
d
in
stru
m
e
n
t
a
ti
o
n
,
im
a
g
e
a
n
d
v
id
e
o
p
ro
c
e
ss
in
g
,
a
n
d
p
a
ra
ll
e
l
c
o
m
p
u
ti
n
g
.
He
is
c
u
rre
n
tl
y
a
n
IEE
E
S
e
n
io
r
M
e
m
b
e
r
(sin
c
e
2
0
1
2
)
,
w
a
s
c
h
a
ir
m
a
n
o
f
IEE
E
In
stru
m
e
n
tatio
n
a
n
d
M
e
a
su
re
m
e
n
t
S
o
c
iety
–
M
a
lay
sia
S
e
c
ti
o
n
(2
0
1
3
a
n
d
2
0
1
4
),
A
ss
o
c
iate
P
ro
f
e
ss
o
r
(sin
c
e
2
0
1
2
),
He
a
d
o
f
De
p
a
rt
m
e
n
t
(
2
0
1
5
-
2
0
1
6
)
a
t
De
p
a
rtm
e
n
t
o
f
El
e
c
tri
c
a
l
a
n
d
Co
m
p
u
ter
En
g
in
e
e
rin
g
,
a
n
d
He
a
d
o
f
P
r
o
g
ra
m
m
e
Ac
c
re
d
it
a
ti
o
n
a
n
d
Q
u
a
li
ty
A
s
su
ra
n
c
e
f
o
r
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
(sin
c
e
2
0
1
7
),
I
n
tern
a
ti
o
n
a
l
Isla
m
ic
Un
iv
e
rsit
y
M
a
l
a
y
sia
.
He
is
Ch
a
rtere
d
En
g
in
e
e
r
(IE
T
,
UK
)
a
n
d
In
si
n
y
u
r
P
r
o
f
e
sio
n
a
l
M
a
d
y
a
(P
II,
In
d
o
n
e
sia
)
sin
c
e
2
0
1
6
.
Nu
r
‘Ati
k
a
h
M
u
h
a
m
a
t
S
a
le
h
re
c
e
iv
e
d
h
e
r
BEn
g
d
e
g
re
e
in
C
o
m
m
u
n
ica
ti
o
n
En
g
in
e
e
rin
g
in
2
0
1
7
f
ro
m
th
e
De
p
a
rtm
e
n
t
o
f
El
c
tri
c
a
l
a
n
d
C
o
m
p
u
ter
En
g
i
n
e
e
rin
g
,
In
ter
n
a
ti
o
n
a
l
Isla
m
ic
Un
iv
e
rsit
y
M
a
la
y
sia
.
He
r
re
se
a
r
c
h
in
tere
sts
i
n
c
lu
d
e
sp
e
e
c
h
p
ro
c
e
ss
in
g
a
n
d
sig
n
a
l
p
ro
c
e
ss
in
g
.
Cu
rre
n
tl
y
,
sh
e
is
p
u
rsu
in
g
M
a
ste
r
o
f
En
g
in
e
e
rin
g
in
Co
m
m
u
n
ica
ti
o
n
a
t
th
e
sa
m
e
u
n
iv
e
rsit
y
w
o
rk
in
g
o
n
sp
e
e
c
h
e
m
o
ti
o
n
re
c
o
g
n
it
io
n
.
M
ira
K
a
r
ti
w
i
c
o
m
p
lete
d
h
e
r
stu
d
ies
a
t
th
e
Un
iv
e
rsity
o
f
W
o
ll
o
n
g
o
n
g
,
A
u
stra
li
a
re
su
lt
in
g
in
t
h
e
f
o
ll
o
w
in
g
d
e
g
re
e
s
b
e
in
g
c
o
n
f
e
rr
e
d
:
Ba
c
h
e
lo
r
o
f
Co
m
m
e
rc
e
in
B
u
sin
e
ss
In
f
o
rm
a
ti
o
n
S
y
ste
m
s,
M
a
ste
r
in
In
f
o
r
m
a
ti
o
n
S
y
ste
m
s
i
n
2
0
0
1
a
n
d
h
e
r
Do
c
to
r
o
f
P
h
il
o
s
o
p
h
y
in
2
0
0
9
.
S
h
e
is
c
u
rre
n
tl
y
a
n
A
s
so
c
iate
P
ro
f
e
ss
o
r
in
De
p
a
rtm
e
n
t
o
f
In
f
o
r
m
a
ti
o
n
S
y
ste
m
s,
Ku
li
y
y
a
h
o
f
In
f
o
r
m
a
ti
o
n
a
n
d
Co
m
m
u
n
ica
ti
o
n
T
e
c
h
n
o
lo
g
y
,
In
tern
a
ti
o
n
a
l
Isla
m
ic
Un
iv
e
rsit
y
M
a
la
y
sia
.
H
e
r
re
s
e
a
rc
h
in
tere
sts
in
c
lu
d
e
e
lec
tro
n
ic co
m
m
e
rc
e
,
d
a
ta m
in
in
g
,
e
-
h
e
a
lt
h
a
n
d
m
o
b
il
e
a
p
p
li
c
a
ti
o
n
s d
e
v
e
lo
p
m
e
n
t.
Evaluation Warning : The document was created with Spire.PDF for Python.