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
.
2
,
A
p
r
il
20
25
,
p
p
.
2
1
5
0
~
2
1
6
0
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
2
.
pp
2
1
5
0
-
2
1
6
0
2150
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
H
o
mo
mo
rphic en
cryp
tion, pr
iv
a
cy
-
preserv
ing
f
ea
tu
r
e
ex
trac
tion, and
d
ecentr
a
lize
d archi
tect
u
re f
o
r
enhan
cing
priv
a
cy
in
v
o
ice authent
ica
tion
K
a
t
hiresh
M
ur
ug
esa
n
1
,
L
a
v
a
ny
a
Su
bb
a
ra
y
a
lu Ra
m
a
m
u
rt
hy
2
,
B
o
o
pa
t
hi
P
a
la
nis
a
m
y
3
,
Ya
m
ini
Cha
nd
ra
s
ek
a
r
4
,
K
a
v
it
ha
M
a
s
a
g
o
un
da
np
u
dh
ur
S
ha
nm
ug
a
m
5
,
B
a
llu
ru
T
ha
m
m
a
ia
hs
het
t
y
A
dis
ha
nk
a
r
Nit
hy
a
6
,
Velum
a
ni
T
hiy
a
g
a
ra
j
a
n
1
,
Ra
m
a
ra
j
M
un
ia
pp
a
n
1
1
D
e
p
a
r
t
e
m
e
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
R
a
t
h
i
n
a
m
C
o
l
l
e
g
e
o
f
A
r
t
s a
n
d
S
c
i
e
n
c
e
,
C
o
i
mb
a
t
o
r
e
,
I
n
d
i
a
2
D
e
p
a
r
t
e
m
e
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
K
P
R
C
o
l
l
e
g
e
o
f
A
r
t
s Sc
i
e
n
c
e
a
n
d
R
e
s
e
a
r
c
h
,
C
o
i
m
b
a
t
o
r
e
,
I
n
d
i
a
3
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
A
p
p
l
i
c
a
t
i
o
n
s,
N
e
h
r
u
A
r
t
s a
n
d
S
c
i
e
n
c
e
C
o
l
l
e
g
e
,
C
o
i
m
b
a
t
o
r
e
,
I
n
d
i
a
4
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
A
p
p
l
i
c
a
t
i
o
n
s,
S
r
i
R
a
m
a
k
r
i
sh
n
a
C
A
S
f
o
r
W
o
me
n
,
C
o
i
mb
a
t
o
r
e
,
I
n
d
i
a
5
D
e
p
a
r
t
e
m
e
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
A
k
sh
a
y
a
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
Te
c
h
n
o
l
o
g
y
,
C
o
i
m
b
a
t
o
r
e
,
I
n
d
i
a
6
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
P
r
e
si
d
e
n
c
y
U
n
i
v
e
r
si
t
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 3
1
,
2
0
2
4
R
ev
is
ed
No
v
2
2
,
2
0
2
4
Acc
ep
ted
Dec
2
,
2
0
2
4
Th
is
p
a
p
e
r
in
tr
o
d
u
c
e
s
a
n
o
v
e
l
fra
m
e
wo
rk
d
e
sig
n
e
d
t
o
b
o
lste
r
p
riv
a
c
y
p
ro
tec
ti
o
n
s
wi
th
i
n
a
u
to
m
a
ted
v
o
ice
a
u
th
e
n
t
ica
ti
o
n
sy
ste
m
s,
a
d
d
re
ss
in
g
m
o
u
n
ti
n
g
c
o
n
c
e
rn
s
a
s
v
o
ice
-
b
a
se
d
a
u
th
e
n
ti
c
a
ti
o
n
g
ro
ws
in
p
ro
m
i
n
e
n
c
e
.
Th
e
wid
e
sp
re
a
d
a
d
o
p
ti
o
n
o
f
t
h
e
se
sy
ste
m
s
h
a
s
u
n
d
e
rsc
o
re
d
a
p
p
re
h
e
n
sio
n
s
re
g
a
rd
in
g
t
h
e
sto
ra
g
e
a
n
d
p
ro
c
e
ss
in
g
o
f
se
n
siti
v
e
v
o
ice
b
io
m
e
tri
c
d
a
ta
with
o
u
t
a
d
e
q
u
a
te
sa
fe
g
u
a
rd
s.
To
m
it
ig
a
te
th
e
se
risk
s,
a
m
o
d
i
fied
f
ra
m
e
wo
rk
is
p
ro
p
o
se
d
,
a
imin
g
to
e
n
h
a
n
c
e
p
ri
v
a
c
y
with
o
u
t
c
o
m
p
ro
m
isin
g
a
u
th
e
n
ti
c
a
ti
o
n
a
c
c
u
ra
c
y
a
n
d
e
fficie
n
c
y
.
Th
re
e
k
e
y
tec
h
n
i
q
u
e
s
a
re
imp
lem
e
n
ted
to
a
d
d
re
ss
th
e
se
c
h
a
ll
e
n
g
e
s.
F
irstl
y
,
a
d
v
a
n
c
e
d
e
n
c
ry
p
t
io
n
m
e
th
o
d
s
a
re
e
m
p
lo
y
e
d
f
o
r
se
c
u
re
v
o
ice
d
a
ta
sto
ra
g
e
a
n
d
tra
n
sm
issio
n
,
th
ro
u
g
h
th
e
h
o
m
o
m
o
rp
h
ic
e
n
c
ry
p
ti
o
n
to
e
n
a
b
le
a
u
t
h
e
n
ti
c
a
ti
o
n
p
ro
c
e
ss
in
g
o
n
e
n
c
ry
p
ted
d
a
ta.
S
e
c
o
n
d
ly
,
a
p
ri
v
a
c
y
-
p
re
se
rv
in
g
fe
a
tu
re
e
x
trac
ti
o
n
m
e
th
o
d
is
in
tro
d
u
c
e
d
,
tran
sfo
rm
i
n
g
ra
w
v
o
i
c
e
in
p
u
ts
i
n
to
irre
v
e
rsib
le
re
p
re
se
n
tatio
n
s
to
sh
ield
o
rig
i
n
a
l
b
io
m
e
tri
c
in
f
o
rm
a
ti
o
n
.
Ad
d
it
io
n
a
ll
y
,
t
h
e
f
ra
m
e
wo
rk
in
c
o
rp
o
ra
tes
d
iffere
n
ti
a
l
p
riv
a
c
y
m
e
c
h
a
n
ism
s,
a
d
d
in
g
c
o
n
tr
o
ll
e
d
n
o
ise
to
a
g
g
re
g
a
ted
v
o
ice
d
a
ta
to
p
re
v
e
n
t
i
n
d
i
v
id
u
a
l
id
e
n
ti
fica
ti
o
n
wi
th
in
larg
e
d
a
tas
e
ts.
A
u
se
r
-
c
e
n
tri
c
c
o
n
se
n
t
a
n
d
c
o
n
tr
o
l
m
o
d
e
l
is
p
ro
p
o
se
d
,
e
m
p
o
we
rin
g
in
d
i
v
id
u
a
ls
t
o
m
a
n
a
g
e
t
h
e
ir
v
o
ice
p
ro
f
il
e
s
a
n
d
a
u
t
h
e
n
ti
c
a
ti
o
n
se
tt
in
g
s.
Ex
p
e
rime
n
tal
fin
d
i
n
g
s
d
e
m
o
n
str
a
te
th
a
t
th
e
fra
m
e
wo
rk
a
c
h
iev
e
s
e
n
h
a
n
c
e
d
a
u
th
e
n
ti
c
a
ti
o
n
a
c
c
u
ra
c
y
wh
il
e
m
a
rk
e
d
ly
re
d
u
c
i
n
g
p
riv
a
c
y
ris
k
s
c
o
m
p
a
re
d
to
c
o
n
v
e
n
ti
o
n
a
l
sy
ste
m
s.
T
h
is
c
o
n
tr
ib
u
ti
o
n
a
d
d
re
ss
e
s
th
e
o
n
g
o
i
n
g
c
h
a
ll
e
n
g
e
o
f
b
a
lan
c
in
g
se
c
u
rit
y
a
n
d
p
ri
v
a
c
y
in
b
io
m
e
tri
c
a
u
t
h
e
n
ti
c
a
ti
o
n
tec
h
n
o
l
o
g
ies
.
K
ey
w
o
r
d
s
:
C
h
af
f
g
en
er
atio
n
C
o
n
v
o
lu
tio
n
n
eu
r
al
n
etwo
r
k
Ho
m
o
m
o
r
p
h
ic
e
n
c
r
y
p
tio
n
R
iv
est Sh
am
ir
Ad
lem
an
Vo
ice
a
u
th
en
ticatio
n
Vo
ice
s
ig
n
al
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
:
Kath
ir
esh
Mu
r
u
g
esan
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
R
ath
in
am
C
o
lleg
e
o
f
A
r
ts
an
d
Scien
ce
C
o
im
b
ato
r
e,
T
am
il Na
d
u
6
4
1
0
2
1
,
I
n
d
ia
E
m
ail:
k
ath
ir
esh
cs8
3
@
g
m
a
il.c
o
m
1.
I
NT
RO
D
UCT
I
O
N
Mo
s
t
o
f
th
e
cu
r
r
en
t
g
en
er
atio
n
s
m
ar
t
d
ev
ices
an
d
s
er
v
ices
p
r
o
v
id
e
v
o
ice
-
d
r
iv
en
in
ter
ac
ti
o
n
s
u
p
p
o
r
t.
An
in
cr
ea
s
in
g
n
u
m
b
er
o
f
s
p
ee
ch
r
ec
o
r
d
in
g
s
ar
e
g
ath
er
e
d
b
y
telev
is
io
n
s
,
s
m
ar
t
p
h
o
n
es,
lo
u
d
s
p
ea
k
er
s
,
watc
h
es,
s
m
ar
t
d
ig
ital
ass
is
tan
ts
,
an
d
also
s
m
ar
t
v
eh
icles
[
1
]
.
T
h
e
ch
an
ce
s
f
o
r
s
p
ee
ch
d
ata
to
b
e
d
is
clo
s
ed
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Ho
mo
mo
r
p
h
ic
en
cryp
tio
n
,
p
r
iva
cy
-
p
r
eser
vin
g
fea
tu
r
e
ex
tr
a
ctio
n
,
a
n
d
…
(
K
a
th
ir
esh
Mu
r
u
g
esa
n
)
2151
co
r
r
u
p
te
d
b
y
m
alicio
u
s
u
s
er
s
o
r
h
ac
k
e
r
s
g
iv
e
r
is
e
to
s
er
io
u
s
p
r
iv
ac
y
c
h
allen
g
es.
Sp
ee
ch
r
ec
o
r
d
in
g
s
h
av
e
ab
u
n
d
a
n
ce
o
f
p
er
s
o
n
al,
c
o
n
f
i
d
en
tial
d
ata,
wh
ich
ca
n
b
e
h
elp
f
u
l
in
s
u
p
p
o
r
tin
g
d
if
f
er
e
n
t
ap
p
licatio
n
s
,
r
an
g
in
g
f
r
o
m
h
ea
lth
p
r
o
f
ilin
g
to
b
io
m
etr
ic
r
ec
o
g
n
itio
n
[
2
]
.
Hen
ce
,
it
is
im
p
o
r
tan
t
th
at
s
p
ee
ch
r
ec
o
r
d
in
g
s
ar
e
s
u
f
f
icien
tly
s
ec
u
r
ed
s
o
th
at
t
h
eir
m
is
u
s
e
is
n
o
t
p
o
s
s
ib
le.
I
n
f
ac
t,
m
u
ch
o
f
th
e
p
r
iv
ac
y
-
o
r
ien
ted
in
f
o
r
m
atio
n
p
r
esen
t
in
s
p
ee
ch
s
ig
n
als
is
t
h
at
ass
o
ciate
d
with
in
d
iv
id
u
a
l
id
en
tity
,
i
n
f
o
r
m
atio
n
d
is
clo
s
ed
v
ia
a
u
to
m
ate
d
s
p
ea
k
er
an
d
s
p
ee
ch
ch
a
r
ac
ter
izatio
n
tech
n
iq
u
es
[
3
]
.
T
h
e
h
u
m
an
v
o
ice
b
elo
n
g
s
to
t
h
e
ca
teg
o
r
y
o
f
t
h
e
m
o
s
t
ca
s
u
al,
n
o
n
-
in
v
asiv
e
an
d
ea
s
iest
o
f
a
ll
f
ea
tu
r
es.
Au
to
m
ated
s
p
ea
k
er
ch
ar
ac
ter
izatio
n
s
h
av
e
p
r
o
g
r
ess
ed
ex
tr
em
ely
well
in
th
e
p
ast
two
d
ec
ad
es.
Priv
ac
y
co
n
ce
r
n
e
d
v
o
ice
au
th
e
n
ticatio
n
s
y
s
tem
(
PVAS)
is
b
asically
a
p
r
o
to
co
l,
w
h
ich
em
p
lo
y
s
a
f
u
s
io
n
o
f
Gau
s
s
ian
m
ix
tu
r
e
m
o
d
els
(
GM
Ms
)
g
en
er
ated
f
r
o
m
th
e
au
d
io
r
ec
o
r
d
i
n
g
s
o
f
a
u
s
er
an
d
c
h
af
f
GM
Ms
g
en
er
ated
f
r
o
m
o
th
er
u
s
er
s
o
r
f
r
o
m
ch
an
g
in
g
t
h
e
av
ailab
le
GM
Ms
to
estab
lis
h
a
ch
allen
g
e
r
esp
o
n
s
e
p
r
o
to
c
o
l
th
at
h
as
v
o
ice
as
th
e
b
aselin
e
[
4
]
.
Pre
s
en
tly
,
th
e
tech
n
o
lo
g
y
o
f
s
p
ea
k
er
ch
ar
ac
ter
izatio
n
is
g
ettin
g
h
u
g
e
ly
p
o
p
u
lar
,
with
its
ap
p
licatio
n
i
n
th
e
p
er
s
o
n
al
a
u
th
en
ticatio
n
an
d
ac
ce
s
s
co
n
tr
o
l
ac
r
o
s
s
a
p
leth
o
r
a
o
f
v
ar
io
u
s
s
er
v
ices
an
d
d
ev
ices,
f
o
r
in
s
tan
ce
,
telep
h
o
n
e
b
an
k
in
g
s
er
v
ices
an
d
s
m
ar
t
d
ev
ices,
wh
ich
eith
er
h
av
e
o
r
r
en
d
e
r
ac
ce
s
s
to
p
er
s
o
n
al
o
r
co
n
f
i
d
en
tial
d
ata
[
5
]
,
[
6
]
.
R
iv
est
Sh
am
ir
A
d
lem
an
(
R
SA
)
en
c
r
y
p
tio
n
is
em
p
lo
y
ed
f
o
r
t
h
e
g
en
er
atio
n
o
f
th
e
s
er
v
er
an
d
t
h
e
u
s
er
k
ey
p
ai
r
s
in
b
o
th
th
e
ac
tu
al
an
d
th
e
m
o
d
if
ied
f
o
r
m
o
f
th
e
p
r
o
to
co
l
f
o
r
b
o
th
1
0
3
r
eg
is
tr
atio
n
an
d
v
e
r
if
icatio
n
[
7
]
.
I
n
s
p
ite
o
f
th
e
clea
r
b
en
e
f
its
an
d
u
b
i
q
u
ity
o
f
b
io
m
etr
ics
m
ec
h
an
is
m
,
d
o
u
b
ts
with
r
esp
ec
t
to
th
r
ea
t
s
to
p
r
iv
ac
y
h
a
v
e
cr
u
s
h
ed
th
e
tr
u
s
t
o
f
th
e
p
u
b
lic.
No
t
ju
s
t,
th
e
id
en
tity
o
f
a
s
p
ea
k
er
is
in
f
o
r
m
atio
n
th
at
is
h
ig
h
ly
co
n
f
id
en
tial,
t
h
e
co
n
ten
t
s
p
elled
o
u
t
m
ig
h
t
also
b
e
s
en
s
itiv
e.
T
h
e
ch
allen
g
es
with
r
esp
ec
t
to
p
r
i
v
ac
y
ar
e
ass
o
ciate
d
with
t
h
e
p
r
o
b
ab
le
i
n
ter
f
er
e
n
ce
a
n
d
illeg
al
u
s
ag
e
o
f
b
io
m
etr
ic
an
d
n
o
n
-
b
i
o
m
etr
ic
s
p
ee
ch
d
ata
[
8
]
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
ao
an
d
B
u
s
s
o
[
9
]
,
h
as
p
r
o
p
o
s
ed
au
d
io
-
v
is
u
al
au
to
m
atic
s
p
ee
ch
r
ec
o
g
n
itio
n
(
AV
-
ASR
)
s
y
s
tem
b
ased
o
n
m
u
ltit
ask
lear
n
in
g
(
MTL
)
wh
er
e
au
d
io
v
is
u
al
v
o
ic
e
ac
tiv
ity
d
etec
tio
n
s
(
AV
-
VA
D
)
wer
e
a
s
ec
o
n
d
ar
y
g
o
al.
T
h
e
s
tu
d
y
ac
co
m
p
lis
h
ed
a
g
en
er
aliza
b
le
an
d
d
ep
e
n
d
a
b
le
au
d
io
-
v
is
u
al
(
AV
)
s
y
s
tem
th
at
was
ac
cu
r
ate.
AV
-
ASR
p
er
f
o
r
m
an
ce
im
p
r
o
v
ed
wh
en
s
p
ee
ch
ac
tiv
ities
wer
e
d
etec
ted
in
s
eg
m
en
ts
.
T
h
e
s
tu
d
y
[
1
0
]
c
o
n
s
id
er
s
th
e
tem
p
o
r
al
d
y
n
am
ics
th
at
o
cc
u
r
b
etwe
en
th
e
m
o
d
alities
,
r
esu
ltin
g
in
an
ap
p
ea
lin
g
a
n
d
p
r
ac
tical
f
u
s
io
n
ap
p
r
o
ac
h
.
T
h
e
r
esear
ch
er
s
c
o
m
p
ar
ed
t
h
eir
ap
p
r
o
ac
h
es
u
s
in
g
a
lar
g
e
a
u
d
io
v
is
u
al
c
o
r
p
u
s
(
n
ea
r
ly
6
0
h
o
u
r
s
)
with
d
if
f
er
en
t
ch
an
n
els
an
d
s
in
g
l
e
an
d
m
u
ltip
le
jo
b
s
.
G
u
g
lan
i
an
d
Mish
r
a
[
1
1
]
,
en
h
an
ce
d
au
to
m
atic
s
p
ee
ch
r
ec
o
g
n
itio
n
(
ASR
)
s
y
s
tem
p
er
f
o
r
m
an
ce
s
wh
e
r
e
p
itch
an
d
v
o
ice
q
u
alities
wer
e
ex
am
in
ed
.
B
ec
au
s
e
o
f
th
e
to
n
al
f
o
u
n
d
atio
n
o
f
Pu
n
jab
i,
t
h
eir
A
SR
s
y
s
tem
s
b
ased
o
n
p
itch
ch
ar
ac
ter
is
tics
wer
e
in
v
esti
g
ated
.
W
h
en
ass
ess
ed
in
ter
m
s
o
f
wo
r
d
m
is
tak
e
r
ates
,
th
eir
s
y
s
tem
p
er
f
o
r
m
e
d
we
ll,
with
im
p
r
o
v
em
en
ts
d
u
e
t
o
p
itch
a
n
d
v
o
ice
d
ep
en
d
e
n
t
f
ea
tu
r
es.
W
in
u
r
s
ito
et
a
l.
[
1
2
]
h
as
co
m
b
in
ed
m
el
-
f
r
eq
u
e
n
cy
ce
p
s
tr
al
co
ef
f
icien
ts
(
MFC
C
)
f
ea
tu
r
e
ex
tr
ac
tio
n
s
with
p
r
in
cip
al
co
m
p
o
n
e
n
t
an
a
ly
s
is
(
PC
A
)
f
o
r
im
p
r
o
v
ed
ac
c
u
r
ac
y
o
f
I
n
d
o
n
esian
s
p
ee
c
h
r
e
co
g
n
itio
n
s
y
s
tem
s
.
T
h
eir
s
y
s
tem
’
s
ac
cu
r
ac
y
in
c
r
ea
s
ed
an
d
d
im
e
n
s
io
n
s
r
ed
u
ce
d
d
u
e
to
MFC
C
an
d
PC
A.
T
r
an
et
a
l
.
[
1
3
]
h
as
p
r
esen
ted
in
th
e
MFC
C
-
b
ased
f
ea
tu
r
e
ex
tr
a
ctio
n
s
an
d
d
elt
a
co
ef
f
icien
ts
wer
e
u
s
ed
to
b
u
ild
m
atr
ices,
wh
ile
PC
A
m
in
im
ized
d
im
en
s
io
n
ali
ties
.
Data
r
ed
u
ctio
n
s
wer
e
ac
co
m
p
lis
h
ed
b
y
t
h
e
u
s
ag
e
o
f
PC
A
v
ar
iatio
n
s
an
d
s
u
b
s
eq
u
en
tly
class
if
ied
u
s
in
g
K
-
n
ea
r
est
n
eig
h
b
o
r
(
KNN
)
.
C
elin
et
a
l.
[
1
4
]
h
ad
s
u
g
g
ested
m
u
lti
-
r
eso
lu
tio
n
f
ea
tu
r
e
e
x
tr
ac
tio
n
s
i
n
af
te
r
d
u
al
d
ata
a
u
g
m
e
n
tatio
n
s
o
n
d
y
s
ar
th
r
ic
s
p
ee
ch
es
u
tili
zin
g
m
icr
o
p
h
o
n
e
a
r
r
ay
b
ased
v
ir
tu
al
lin
ea
r
s
y
n
th
eses
.
T
h
eir
ASR
s
y
s
tem
was
d
esig
n
ed
f
o
r
lo
w
an
d
ex
tr
em
el
y
lo
w
in
tellig
ib
le
s
p
ea
k
er
s
with
d
y
s
a
r
th
r
ia
an
d
th
eir
r
esu
lts
s
h
o
wed
lo
wer
w
o
r
d
er
r
o
r
r
ates
(
W
E
R
s
)
o
f
u
p
to
3
2
.
7
9
%
ag
ain
s
t
3
5
.
7
5
%
wh
en
co
m
p
ar
ed
to
cu
r
r
en
t
d
y
s
ar
th
r
ic
s
p
ee
ch
r
ec
o
g
n
itio
n
d
ata
au
g
m
en
tatio
n
s
.
Aid
a
-
Z
ad
e
et
a
l.
[
1
5
]
,
u
tili
ze
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
(
SVMs
)
to
d
ev
elo
p
ac
o
u
s
tic
m
o
d
els
o
f
s
p
ee
ch
r
ec
o
g
n
itio
n
s
b
ased
o
n
MFC
C
an
d
lin
ea
r
p
r
ed
ictiv
e
co
d
in
g
(
L
PC
)
ch
ar
ac
ter
is
tics
,
ev
alu
ated
o
n
Aze
r
b
aijan
i
d
ata.
A
co
n
v
o
lu
tio
n
n
eu
r
al
n
etwo
r
k
-
b
id
ir
ec
tio
n
al
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
C
NN
-
B
L
STM
)
h
y
b
r
i
d
ar
ch
itectu
r
e
was
d
ev
elo
p
ed
b
y
Pas
s
r
ich
a
an
d
Ag
g
ar
wal
[
1
6
]
,
t
o
ef
f
ec
tiv
ely
m
ak
e
u
s
e
o
f
th
ese
ch
ar
ac
ter
is
tics
an
d
en
h
an
ce
co
n
tin
u
o
u
s
v
o
ice
r
ec
o
g
n
itio
n
s
.
T
h
e
s
tu
d
y
f
o
cu
s
ed
o
n
n
u
m
b
er
o
f
b
id
ir
e
ctio
n
al
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
B
L
STM
)
lay
er
s
th
at
wer
e
u
s
ef
u
l.
I
n
co
m
p
ar
is
o
n
t
o
co
n
v
o
l
u
tio
n
n
eu
r
al
n
etwo
r
k
(
C
NN
)
an
d
d
ee
p
n
e
u
r
al
n
etwo
r
k
(
DNN
)
s
y
s
tem
s
,
th
eir
ex
p
er
im
en
ts
s
h
o
w
th
at
h
y
b
r
id
ar
ch
itectu
r
es
with
s
p
e
ec
h
f
ea
tu
r
es
an
d
n
o
n
-
lin
ea
r
iti
es
with
d
r
o
p
o
u
ts
,
d
ec
r
ea
s
ed
W
E
R
b
y
5
.
8
%
an
d
1
0
%,
r
esp
ec
tiv
ely
[
1
7
]
.
T
h
e
s
tu
d
y
u
s
ed
d
ata
cr
ea
ted
f
r
o
m
1
4
0
v
o
ice
r
ec
o
r
d
in
g
s
f
r
o
m
2
8
d
if
f
er
en
t
s
p
ea
k
er
s
.
I
n
th
eir
ex
p
e
r
im
en
tatio
n
s
,
t
h
eir
f
ir
s
t
v
ar
iatio
n
o
f
PC
A
r
ed
u
ce
d
f
ea
tu
r
es
f
r
o
m
2
6
to
1
2
wh
ile
m
ain
tain
in
g
s
p
ee
c
h
r
ec
o
g
n
itio
n
ac
c
u
r
ac
ies
at
8
6
.
4
3
%
an
d
eq
u
al
to
co
n
v
e
n
tio
n
al
MFC
C
ap
p
r
o
ac
h
es
with
o
u
t
PC
A
[
1
8
]
.
T
h
eir
s
ec
o
n
d
v
ar
iatio
n
o
f
PC
A
r
ed
u
ce
d
f
ea
tu
r
e
co
u
n
ts
f
r
o
m
2
6
to
1
0
with
e
n
h
an
ce
d
r
ec
o
g
n
itio
n
ac
cu
r
ac
ies
o
f
8
9
.
2
9
% f
r
o
m
8
6
.
4
3
% o
v
e
r
M
FC
C
with
o
u
t PC
A
b
aselin
e
[
1
9
]
.
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
.
2
,
Ap
r
il
20
25
:
2
1
5
0
-
2
1
6
0
2152
3.
M
E
T
H
O
D
A
n
o
v
el
f
r
am
ewo
r
k
d
esig
n
e
d
to
en
h
an
ce
p
r
iv
ac
y
in
v
o
i
ce
au
th
en
ticatio
n
s
y
s
tem
s
th
r
o
u
g
h
th
e
in
teg
r
atio
n
o
f
th
r
ee
k
ey
co
m
p
o
n
en
ts
:
h
o
m
o
m
o
r
p
h
ic
en
cr
y
p
t
io
n
,
p
r
iv
ac
y
-
p
r
eser
v
in
g
f
ea
tu
r
e
ex
tr
ac
tio
n
,
an
d
a
d
ec
en
tr
alize
d
ar
c
h
itectu
r
e.
Ho
m
o
m
o
r
p
h
ic
en
cr
y
p
tio
n
en
s
u
r
e
s
th
at
v
o
ice
d
ata
r
em
ain
s
en
c
r
y
p
ted
th
r
o
u
g
h
o
u
t
p
r
o
ce
s
s
in
g
,
p
r
ev
en
tin
g
u
n
au
th
o
r
ized
ac
ce
s
s
to
s
en
s
itiv
e
in
f
o
r
m
atio
n
[
2
0
]
.
Priv
ac
y
-
p
r
eser
v
in
g
f
ea
t
u
r
e
ex
tr
ac
tio
n
tech
n
i
q
u
es
ar
e
em
p
lo
y
ed
to
ex
t
r
ac
t
ess
en
tial
v
o
ice
ch
ar
ac
ter
is
tics
wh
ile
p
r
eser
v
in
g
th
e
an
o
n
y
m
ity
o
f
th
e
u
s
er
.
3
.
1
.
P
ri
v
a
cy
c
o
ncer
ned v
o
ice
a
uthent
ica
t
io
n sy
s
t
em
Priv
ac
y
b
ased
v
o
ice
au
th
e
n
ticatio
n
is
r
eg
a
r
d
ed
to
b
e
t
h
e
m
o
s
t
im
p
o
r
ta
n
t
co
n
ce
p
t
in
th
e
d
ef
en
s
e
s
ec
to
r
to
g
u
ar
an
tee
th
at
v
o
ic
e
co
m
m
u
n
icatio
n
an
d
au
th
en
ticatio
n
is
p
r
o
v
id
ed
u
tm
o
s
t
s
ec
u
r
ity
[
2
1
]
.
I
t
is
ac
co
m
p
lis
h
ed
th
r
o
u
g
h
t
h
e
g
e
n
er
atio
n
o
f
th
e
ch
af
f
m
o
d
el
th
a
t
will
b
e
co
m
m
u
n
icate
d
to
t
h
e
s
er
v
er
r
ath
er
th
an
s
h
ar
in
g
th
e
ac
tu
al
v
o
ice
m
o
d
e
l.
I
n
th
is
,
at
f
ir
s
t,
th
e
ch
af
f
m
o
d
el
o
f
v
o
ice
s
ig
n
al
is
cr
ea
ted
with
th
e
h
elp
o
f
th
e
Gau
s
s
ian
m
ix
tu
r
e
m
o
d
el.
T
h
e
en
tire
p
r
o
ce
s
s
in
g
f
lo
w
o
f
th
e
n
ewly
in
tr
o
d
u
ce
d
r
esear
c
h
ap
p
r
o
ac
h
is
illu
s
tr
ated
in
th
e
Fig
u
r
e
1
.
T
h
e
ch
af
f
m
o
d
el
c
r
ea
ted
will
b
e
p
r
esen
ted
as
in
p
u
t
to
t
h
e
R
SA
m
o
d
el
f
o
r
th
e
s
ec
r
et
k
ey
p
air
g
e
n
er
atio
n
.
Dep
en
d
in
g
o
n
th
e
cr
ea
ted
k
ey
p
air
,
v
o
ice
ch
af
f
m
o
d
el
will
b
e
en
co
d
ed
an
d
later
s
h
ar
ed
with
th
e
s
er
v
er
.
I
n
th
e
s
er
v
er
en
d
,
t
h
e
ac
tu
al
v
o
ice
s
ig
n
al
will
b
e
o
b
tain
ed
wi
th
th
e
h
elp
o
f
th
e
r
ig
h
t
k
e
y
s
an
d
th
er
ea
f
te
r
,
co
n
v
o
l
u
tio
n
al
n
e
u
r
al
n
etwo
r
k
will
h
elp
in
th
e
a
u
th
en
ti
ca
tio
n
p
r
o
ce
s
s
.
T
h
e
ch
a
r
ac
t
er
is
tics
f
o
r
v
o
ice
au
th
en
ticatio
n
will b
e
o
b
tain
e
d
b
ef
o
r
e
u
s
in
g
g
en
etic
b
ased
C
NN
th
at
u
s
es th
e
Gab
o
r
f
ilter
.
Fig
u
r
e
1
.
Pro
p
o
s
ed
wo
r
k
f
lo
w
d
iag
r
am
3
.
2
.
Cha
f
f
g
ener
a
t
i
o
n us
ing
G
a
us
s
ia
n m
ix
t
ure
m
o
del
I
n
th
is
tech
n
ical
wo
r
k
,
Gau
s
s
ian
m
ix
tu
r
e
m
o
d
el
is
u
s
ed
f
o
r
th
e
g
en
er
atio
n
o
f
t
h
e
v
o
ice
s
i
g
n
al
ch
af
f
m
o
d
el.
On
e
s
ig
n
if
ican
t
s
tag
e
wh
ile
im
p
lem
en
tin
g
th
e
ab
o
v
e
-
m
en
tio
n
ed
lik
elih
o
o
d
r
atio
d
etec
to
r
is
ch
o
o
s
in
g
th
e
r
ea
l
lik
eli
h
o
o
d
f
u
n
ctio
n
,
(
|
)
[
2
2
]
.
I
n
tex
t
-
b
ased
a
p
p
licatio
n
s
,
wh
er
e
a
g
o
o
d
p
r
io
r
in
f
o
r
m
atio
n
o
f
th
e
s
p
o
k
en
tex
t
is
av
ailab
le,
m
o
r
e
tem
p
o
r
al
k
n
o
wled
g
e
ca
n
b
e
in
clu
d
ed
ap
p
ly
in
g
th
e
h
id
d
en
Ma
r
k
o
v
m
o
d
els
(
HM
Ms)
to
f
o
r
m
t
h
e
lik
elih
o
o
d
f
u
n
ctio
n
’
s
b
aselin
e.
p
(
x
|
λ
)
=
∑
w
i
p
i
(
x
)
M
i
=
1
(
1
)
T
h
e
d
en
s
ity
is
d
ef
in
ed
as
a
we
ig
h
ted
lin
ea
r
co
m
b
in
atio
n
o
f
M
u
n
im
o
d
al
Gau
s
s
ian
d
en
s
ities
,
(
)
,
with
ea
ch
o
n
e
p
ar
a
m
eter
ized
u
s
in
g
a
m
e
an
D×
1
v
ec
to
r
,
µ
,
an
d
a
D×
D
c
o
v
ar
ian
ce
m
at
r
ix
,
Σ
;
p
i
(
x
)
=
1
(
2π
)
D
/
2
|
Σ
i
|
1
/
2
e
xp
{
−
1
2
(
x
−
μ
i
)
′
(
Σ
i
)
−
1
(
x
−
μ
i
)
}
(
2
)
T
h
e
m
ix
tu
r
e
weig
h
ts
,
,
also
m
ee
t
th
e
co
n
d
itio
n
∑
w
i
M
i
=
1
=
1
.
J
o
in
tly
,
t
h
e
p
ar
am
eter
s
o
f
th
e
d
en
s
ity
m
o
d
el
ar
e
r
ep
r
esen
ted
as
λ
=
{wi,
µi,
Σ
i
},
wh
er
e
=
1
.
.
.
,
.
T
h
e
f
ir
s
t
o
n
e,
th
e
d
en
s
ity
m
o
d
elin
g
o
f
an
th
o
r
d
e
r
f
u
ll
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
Ho
mo
mo
r
p
h
ic
en
cryp
tio
n
,
p
r
iva
cy
-
p
r
eser
vin
g
fea
tu
r
e
ex
tr
a
ctio
n
,
a
n
d
…
(
K
a
th
ir
esh
Mu
r
u
g
esa
n
)
2153
co
v
ar
ian
ce
GM
M
ca
n
b
e
v
e
r
y
well
co
n
s
id
er
ed
to
b
e
eq
u
i
v
alen
t
to
th
at
o
b
tain
ed
a
p
p
ly
in
g
a
b
ig
g
er
o
r
d
e
r
d
iag
o
n
al
c
o
v
ar
ian
ce
GM
M
[
2
3
]
.
l
og
p
(
X
|
λ
)
=
∑
l
og
p
(
x
t
|
λ
)
T
t
=
1
(
3
)
wh
er
e
(
|
)
is
ca
lcu
lated
.
Fre
q
u
en
tly
,
th
e
m
ea
n
l
o
g
-
lik
elih
o
o
d
v
alu
e
is
u
tili
ze
d
b
y
d
iv
id
in
g
t
h
e
v
alu
e
o
f
(
|
)
b
y
.
T
h
is
is
ca
r
r
ied
o
u
t
s
o
th
a
t
th
e
d
u
r
atio
n
ef
f
ec
ts
ar
is
in
g
f
r
o
m
th
e
lo
g
-
lik
elih
o
o
d
v
al
u
e
ca
n
b
e
n
o
r
m
alize
d
o
u
t.
3
.
3
.
User
re
g
is
t
ra
t
i
o
n e
m
plo
y
ing
RSA
encr
y
ptio
n
T
h
e
n
am
in
g
o
f
th
e
R
SA a
lg
o
r
ith
m
g
o
es a
f
ter
th
e
n
am
es o
f
it
s
in
v
en
to
r
s
in
1
9
7
8
,
wh
o
ar
e
R
o
n
R
iv
est,
Ad
i
Sh
am
ir
,
an
d
L
eo
n
ar
d
Ad
lem
an
[
2
4
]
.
I
t
is
es
s
en
tial
to
b
e
awa
r
e
o
f
th
e
n
u
m
b
er
o
f
s
tep
s
th
at
a
co
m
p
u
ter
wo
u
ld
tak
e
f
o
r
th
e
e
n
cr
y
p
tio
n
o
f
th
e
m
ess
ag
e
s
o
th
at
it
ca
n
b
e
f
o
u
n
d
wh
eth
er
a
tech
n
iq
u
e
is
q
u
ick
an
d
ef
f
ec
tiv
e.
gc
d
(
a
,
b
)
=
1
a
n
d
J
(
a
,
b
)
=
a
(
b
−
1
)
÷
2
(
mod
b
)
(
4
)
wh
er
e
(
,
)
r
ef
er
s
to
th
e
J
ac
o
b
i sy
m
b
o
l,
wh
ich
ca
n
also
b
e
d
en
o
t
ed
as
(
5
)
.
(
a
b
)
=
(
a
p1
)
α1
(
α
p2
)
α1
…
…
(
α
pk
)
α
k
(
5
)
wh
er
e
b
=
p
1
α1
p
2
α2
…
.
p
k
α
k
r
ef
er
s
t
o
th
e
p
r
im
al
f
ac
to
r
izatio
n
o
f
b
,
a
n
d
th
e
L
e
g
en
d
r
e
s
y
m
b
o
l
is
e
x
p
r
ess
ed
f
o
r
all
in
teg
er
s
a
an
d
all
o
d
d
p
r
im
es p
d
ef
in
ed
b
y
(
6
)
.
(
a
p
)
=
{
0
if
a
≡
0
(
mod
p
)
+
1
if
a
≠
0
(
mod
p
)
a
n
d
for
s
ome
in
te
ge
r
x
,
a
≡
x
2
(
mod
p
)
−
1
if
the
r
e
is
no
s
uc
h
a
≡
0
(
mod
p
)
(
6
)
(
a
1
)
≡
1
T
h
e
J
ac
o
b
i sy
m
b
o
l is o
n
ly
s
p
e
cif
ied
if
a
is
an
in
teg
er
an
d
b
i
s
a
p
o
s
itiv
e
o
d
d
in
te
g
er
.
I
n
ad
d
itio
n
,
(
,
)
is
0
if
(
,
)
≠
1
an
d
±
1
if
(
,
)
=
1
.
I
f
h
o
ld
s
tr
u
e
alwa
y
s
if
is
p
r
im
e,
else
(
i
f
is
co
m
p
o
s
ite)
,
it
will
in
clu
d
e
a
p
r
o
b
a
b
ilit
y
o
f
b
ei
n
g
f
alse
b
y
m
o
r
e
th
a
n
5
0
%.
I
f
h
o
l
es
tr
u
e
1
0
0
tim
es
f
o
r
r
an
d
o
m
l
y
s
elec
ted
’
s
,
th
en
b
is
r
o
u
g
h
ly
to
o
ce
r
tain
to
b
e
p
r
im
e,
with
a
p
r
o
b
ab
ilit
y
o
f
b
ei
n
g
a
1
’
s
co
m
p
o
s
ite
in
2
1
0
0
.
3
.
4
.
G
a
bo
r
f
ilte
r
-
ba
s
ed
f
ea
t
ure
ex
t
ra
ct
io
n
A
Gab
o
r
f
ilter
is
b
asically
a
lin
ea
r
b
an
d
p
ass
f
ilter
u
tili
ze
d
i
n
im
ag
e
p
r
o
ce
s
s
in
g
f
o
r
f
ea
tu
r
e
d
etec
tio
n
an
d
tex
tu
r
e
a
n
aly
s
is
[
2
5
]
.
T
h
e
r
ep
r
esen
tatio
n
s
o
f
Gab
o
r
f
i
lter
s
in
ter
m
s
o
f
f
r
eq
u
e
n
cy
a
n
d
o
r
ien
tatio
n
ar
e
id
en
tical
to
th
o
s
e
p
er
tain
in
g
to
th
e
h
u
m
an
v
is
u
al
s
y
s
tem
,
an
d
th
ey
h
av
e
b
ee
n
o
b
s
er
v
ed
to
b
e
esp
ec
ially
s
u
itab
le
f
o
r
tex
tu
r
e
r
e
p
r
esen
t
atio
n
an
d
d
if
f
er
en
tiatio
n
[
2
6
]
.
I
n
th
e
s
p
atial
d
o
m
ain
,
a
2
D
Gab
o
r
f
ilter
is
b
asically
a
Gau
s
s
ian
k
er
n
el
f
u
n
ctio
n
th
at
is
m
o
d
u
lated
u
s
in
g
a
s
in
u
s
o
id
al
p
lan
e
wav
e
.
(
,
)
=
(
−
(
1
2
+
2
1
2
/
2
2
)
)
(
(
2
1
+
)
)
(
7
)
wh
er
e
x
1
=
x
c
os
θ
+
y
s
in
θ
an
d
y
1
=
−
x
s
in
θ
+
y
c
os
θ
.
T
h
e
in
p
u
t
p
a
r
a
m
eter
s
tan
d
s
f
o
r
th
e
s
tan
d
ar
d
d
ev
iatio
n
o
f
th
e
Ga
u
s
s
ian
f
u
n
ctio
n
,
in
d
icate
s
th
e
wav
elen
g
th
o
f
h
a
r
m
o
n
ic
f
u
n
ctio
n
,
r
ef
er
s
to
th
e
o
r
ien
tatio
n
,
in
d
icate
s
th
e
s
p
atial
asp
ec
t
r
atio
with
a
f
ix
ed
v
alu
e
o
f
0
.
5
,
an
d
r
ef
er
s
to
th
e
p
h
ase
s
h
if
t
o
f
h
ar
m
o
n
ic
f
u
n
ctio
n
.
I
n
th
e
ab
o
v
e
Fig
u
r
e
2
is
r
ep
r
esen
ted
b
y
th
e
co
n
v
o
l
u
tio
n
al
n
etwo
r
k
s
tak
e
t
h
eir
i
n
s
p
ir
atio
n
f
r
o
m
b
i
o
lo
g
ical
p
r
o
ce
s
s
es,
wh
er
e
th
e
co
n
n
ec
tiv
ity
p
atter
n
b
etwe
en
n
eu
r
o
n
s
is
id
en
tical
t
o
th
e
way
in
wh
ich
th
e
an
im
al
v
is
u
al
c
o
r
tex
is
o
r
g
an
ized
.
I
t
h
as
f
iv
e
la
y
er
s
,
w
h
ich
in
clu
d
e,
two
co
n
v
o
lu
tio
n
lay
er
s
,
C
1
a
n
d
C
2
,
an
d
two
p
o
o
lin
g
lay
er
s
,
P1
an
d
P2
,
an
d
an
o
u
t
p
u
t la
y
er
.
T
h
e
s
ize
o
f
th
e
f
ilter
u
tili
ze
d
in
co
n
v
o
lu
tio
n
lay
er
s
C
1
an
d
C
2
was
5
×5
wh
ile
th
e
f
ilter
s
ize
u
tili
ze
d
in
p
o
o
lin
g
lay
er
s
P1
an
d
P2
was
2
×2
.
2
8
×2
8
g
r
ay
s
ca
le
im
ag
es
f
o
r
m
t
h
e
in
p
u
t
to
th
e
C
NN.
T
h
e
o
u
t
p
u
t
o
f
C
1
a
n
d
C
2
p
r
o
v
i
d
es
a
f
ea
tu
r
e
m
ap
o
f
3
2
s
ets
f
o
r
b
o
th
.
T
h
e
o
u
tp
u
ts
f
r
o
m
P1
an
d
P2
co
n
tain
th
e
s
am
e
n
u
m
b
er
o
f
f
ea
tu
r
e
m
a
p
s
ets
lik
e
C
1
an
d
C
2
.
T
h
e
C
N
N
o
u
tp
u
t
was
a
co
lu
m
n
v
ec
to
r
with
2
5
6
×1
d
im
e
n
s
io
n
,
u
tili
ze
d
b
y
th
e
So
f
tMa
x
class
if
ier
f
o
r
class
if
y
in
g
th
e
in
p
u
t
im
ag
es
in
to
o
n
e
o
u
t
o
f
th
e
two
o
u
tp
u
t c
lass
es.
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
.
2
,
Ap
r
il
20
25
:
2
1
5
0
-
2
1
6
0
2154
Fig
u
r
e
2
.
E
n
h
an
ce
d
C
NN
a
r
ch
itectu
r
e
3
.
5
.
H
y
brid
g
enet
ic
a
nd
CNN
ba
s
ed
v
o
ice
a
uthent
ica
t
io
n
Hy
b
r
id
g
en
etic
alg
o
r
ith
m
(
GA)
an
d
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
(
C
NN)
b
ased
v
o
ice
au
th
en
ticatio
n
lev
er
ag
es
th
e
o
p
tim
izatio
n
ca
p
ab
ilit
ies
o
f
GAs
to
d
esig
n
an
d
en
h
an
ce
th
e
ar
c
h
itectu
r
e
o
f
C
NNs
f
o
r
r
o
b
u
s
t
v
o
ice
r
ec
o
g
n
itio
n
.
Ma
p
p
in
g
th
e
d
o
m
ain
o
f
ar
tific
ial
n
eu
r
al
n
etwo
r
k
s
(
ANN)
to
GA
in
v
o
lv
es
u
s
in
g
ev
o
lu
tio
n
ar
y
p
r
in
cip
les
t
o
o
p
t
im
ize
h
y
p
er
p
ar
a
m
eter
s
,
s
u
ch
as
n
etwo
r
k
lay
er
s
,
lear
n
in
g
r
a
tes,
an
d
ac
tiv
atio
n
f
u
n
ctio
n
s
[
2
7
]
.
T
h
is
in
te
g
r
atio
n
ad
d
r
ess
es
ch
allen
g
es
in
n
etwo
r
k
d
esig
n
b
y
ex
p
lo
r
in
g
a
v
ast
s
ea
r
ch
s
p
ac
e
ef
f
icien
tly
.
GA
aid
s
in
d
ev
elo
p
in
g
d
ee
p
C
NNs
b
y
au
to
m
atin
g
th
e
s
elec
tio
n
o
f
o
p
ti
m
al
co
n
f
ig
u
r
atio
n
s
,
im
p
r
o
v
in
g
ac
cu
r
ac
y
,
an
d
r
ed
u
cin
g
tr
ain
in
g
tim
e.
T
h
is
h
y
b
r
i
d
ap
p
r
o
ac
h
en
a
b
les
ad
ap
tiv
e
an
d
ef
f
icien
t
v
o
ice
au
th
en
ticatio
n
s
y
s
tem
s
.
Fig
u
r
e
3
illu
s
tr
ates
th
e
en
h
a
n
c
ed
g
en
etic
an
d
C
NN
-
b
ased
v
o
ice
au
th
en
ticatio
n
p
r
o
ce
s
s
,
h
ig
h
lig
h
tin
g
th
e
u
s
e
o
f
GA
to
d
ev
elo
p
a
d
ee
p
C
NN.
I
n
th
is
p
r
o
ce
s
s
,
GA
tr
ea
ts
n
eu
r
al
n
etwo
r
k
s
as
in
d
iv
id
u
als
in
a
p
o
p
u
latio
n
.
T
h
e
p
e
r
f
o
r
m
an
ce
o
f
ea
ch
n
etwo
r
k
r
ep
r
esen
ts
it
s
f
itn
ess
,
g
u
id
in
g
th
e
s
elec
tio
n
o
f
ca
n
d
id
ates
f
o
r
s
u
b
s
eq
u
en
t
g
en
e
r
atio
n
s
.
I
ter
a
tiv
e
s
tep
s
in
th
e
alg
o
r
ith
m
f
o
cu
s
o
n
ev
o
lv
in
g
t
h
e
p
o
p
u
la
tio
n
b
y
o
p
tim
izin
g
n
etwo
r
k
p
ar
a
m
eter
s
an
d
ar
ch
i
tectu
r
e,
aim
in
g
f
o
r
im
p
r
o
v
ed
ac
cu
r
ac
y
.
T
h
is
s
y
n
er
g
y
b
etw
ee
n
GA
an
d
C
NN
d
em
o
n
s
tr
ates
an
ef
f
ec
tiv
e
m
ea
n
s
o
f
au
to
m
atin
g
a
n
d
en
h
an
cin
g
n
eu
r
al
n
etwo
r
k
d
esig
n
f
o
r
r
o
b
u
s
t
v
o
ice
au
th
en
ticatio
n
s
y
s
tem
s
.
Fig
u
r
e
3
.
E
n
h
an
ce
d
g
e
n
etic
an
d
C
NN
b
ased
v
o
ice
au
th
e
n
ticatio
n
p
r
o
ce
s
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Ho
mo
mo
r
p
h
ic
en
cryp
tio
n
,
p
r
iva
cy
-
p
r
eser
vin
g
fea
tu
r
e
ex
tr
a
ctio
n
,
a
n
d
…
(
K
a
th
ir
esh
Mu
r
u
g
esa
n
)
2155
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
d
is
cu
s
s
ed
two
-
lev
el
v
o
ic
e
au
th
en
ticatio
n
s
y
s
tem
(
T
L
V
AS
)
ass
e
s
s
m
en
t
is
ca
r
r
ied
o
u
t
u
s
in
g
th
e
MA
T
L
AB
s
im
u
latio
n
en
v
ir
o
n
m
en
t.
T
h
e
v
o
icep
r
in
t
is
th
en
co
m
p
ar
ed
ag
ain
s
t
ea
ch
o
th
er
f
o
r
im
p
r
o
v
in
g
th
e
p
er
f
o
r
m
an
ce
.
T
h
e
ch
af
f
m
o
d
e
l
o
f
th
e
in
p
u
t
s
ig
n
al
will
b
e
p
r
o
d
u
ce
d
,
wh
o
s
e
en
c
r
y
p
tio
n
will
b
e
d
o
n
e
with
th
e
s
ec
r
et
k
ey
p
air
.
T
h
en
t
h
is
en
cr
y
p
ted
s
ig
n
al
will
b
e
r
elay
e
d
to
th
e
r
ec
eiv
ed
wh
e
r
e
it
is
d
ec
r
y
p
ted
a
n
d
th
en
f
ea
tu
r
e
ex
tr
a
ctio
n
will
b
e
ca
r
r
ied
o
u
t.
T
h
e
f
ea
tu
r
es
e
x
tr
ac
t
ed
will
b
e
lear
n
ed
em
p
lo
y
in
g
th
e
g
e
n
etic
b
ased
C
NN
alg
o
r
ith
m
f
o
r
th
e
au
th
e
n
ticatio
n
p
r
o
ce
s
s
.
T
h
e
p
r
esen
ted
f
ig
u
r
es
illu
s
tr
ate
v
ar
io
u
s
s
tag
es
o
f
th
e
v
o
ice
au
th
en
ticatio
n
p
r
o
ce
s
s
.
Fig
u
r
e
4
d
em
o
n
s
tr
ates
th
e
in
p
u
t
h
u
m
a
n
v
o
ice
p
r
o
ce
s
s
in
g
,
wh
er
e
t
h
e
in
itial
v
o
ice
s
ig
n
al
is
ca
p
tu
r
e
d
an
d
p
r
ep
ar
e
d
f
o
r
f
u
r
th
er
a
n
aly
s
is
.
T
h
is
is
a
cr
u
cial
s
tep
in
en
s
u
r
in
g
th
at
th
e
r
a
w
au
d
io
d
ata
is
s
u
itab
le
f
o
r
s
u
b
s
eq
u
en
t p
r
o
ce
s
s
in
g
s
tag
es.
Fig
u
r
e
5
d
ep
icts
th
e
ch
af
f
m
o
d
el
p
r
o
ce
s
s
,
wh
ich
i
s
d
esig
n
ed
to
en
h
an
ce
th
e
s
e
cu
r
ity
o
f
th
e
v
o
ice
au
th
en
ticatio
n
s
y
s
tem
.
T
h
e
ch
af
f
m
o
d
el
in
tr
o
d
u
ce
s
d
ec
o
y
o
r
“c
h
af
f
”
f
ea
tu
r
es
in
to
th
e
d
ata,
m
ak
in
g
it
r
esil
ien
t
ag
ain
s
t
u
n
au
th
o
r
ized
attem
p
ts
an
d
tam
p
er
in
g
.
T
h
is
s
tep
p
lay
s
a
s
ig
n
if
ican
t
r
o
le
in
p
r
eser
v
in
g
th
e
in
teg
r
ity
an
d
co
n
f
id
en
tiality
o
f
th
e
a
u
th
en
t
icatio
n
p
r
o
ce
s
s
.
Fig
u
r
e
6
o
u
t
lin
es
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
p
r
o
ce
s
s
,
wh
er
e
th
e
ess
en
tial
ch
ar
ac
ter
is
tic
s
o
f
t
h
e
v
o
ice
s
ig
n
al
ar
e
id
en
tifi
ed
an
d
ex
tr
ac
ted
.
T
h
ese
f
e
atu
r
es
in
clu
d
e
k
ey
p
ar
am
eter
s
s
u
ch
as
p
itch
,
f
r
e
q
u
en
cy
,
an
d
am
p
litu
d
e,
wh
ic
h
ar
e
u
n
iq
u
e
t
o
ea
ch
i
n
d
iv
id
u
al
an
d
c
r
itical
f
o
r
ac
cu
r
ate
v
o
ice
r
ec
o
g
n
itio
n
.
T
o
g
eth
er
,
th
ese
f
ig
u
r
es
r
e
p
r
esen
t
a
s
y
s
tem
atic
ap
p
r
o
ac
h
to
a
ch
iev
in
g
s
ec
u
r
e
a
n
d
r
eliab
le
v
o
ice
au
th
en
ticatio
n
.
T
h
e
i
n
p
u
t
v
o
ice
s
ig
n
al
u
n
d
e
r
g
o
es
m
u
ltip
le
lay
er
s
o
f
p
r
o
c
ess
in
g
,
f
r
o
m
in
itial
ca
p
tu
r
e
to
f
ea
tu
r
e
ex
tr
ac
tio
n
,
with
s
ec
u
r
ity
en
h
an
ce
m
e
n
t
s
p
r
o
v
id
ed
b
y
th
e
ch
a
f
f
m
o
d
el.
T
h
ese
s
tag
es
co
llectiv
ely
f
o
r
m
th
e
f
o
u
n
d
atio
n
o
f
a
r
o
b
u
s
t a
n
d
e
f
f
ec
tiv
e
v
o
ice
au
th
en
ticatio
n
s
y
s
tem
.
Fig
u
r
e
4
.
I
n
p
u
t
h
u
m
a
n
v
o
ice
p
r
o
ce
s
s
in
g
Fig
u
r
e
5
.
C
h
af
f
m
o
d
el
p
r
o
ce
s
s
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
.
2
,
Ap
r
il
20
25
:
2
1
5
0
-
2
1
6
0
2156
Fig
u
r
e
6
s
h
o
ws
th
e
f
ea
tu
r
e
e
x
tr
ac
tio
n
p
r
o
ce
s
s
ca
r
r
ied
o
u
t
o
n
th
e
ch
af
f
m
o
d
el
o
f
in
p
u
t
s
ig
n
al.
T
h
e
lear
n
in
g
o
f
th
e
ex
tr
a
cted
f
ea
t
u
r
es
will
b
e
d
o
n
e
em
p
lo
y
i
n
g
th
e
g
en
etic
b
ased
C
NN
alg
o
r
ith
m
f
o
r
th
e
v
o
ice
au
th
en
ticatio
n
r
esu
lts
.
Fig
u
r
e
6
o
u
tlin
es
th
e
f
ea
tu
r
e
ex
tr
ac
ti
o
n
p
r
o
ce
s
s
,
wh
er
e
th
e
ess
en
tial
ch
ar
ac
ter
is
tics
o
f
th
e
v
o
ice
s
ig
n
al
ar
e
id
e
n
tifie
d
an
d
ex
tr
ac
te
d
.
T
h
ese
f
ea
tu
r
es
in
clu
d
e
k
e
y
p
a
r
am
eter
s
s
u
ch
a
s
p
itch
,
f
r
e
q
u
en
c
y
,
an
d
am
p
litu
d
e,
wh
ic
h
ar
e
u
n
i
q
u
e
to
ea
ch
in
d
iv
id
u
al
an
d
c
r
itical
f
o
r
ac
c
u
r
ate
v
o
ice
r
ec
o
g
n
itio
n
.
T
o
g
et
h
er
,
th
ese
f
ig
u
r
es
r
ep
r
esen
t
a
s
y
s
te
m
atic
ap
p
r
o
ac
h
to
ac
h
iev
in
g
s
ec
u
r
e
a
n
d
r
eliab
le
v
o
ice
a
u
th
e
n
ticatio
n
.
T
h
e
in
p
u
t
v
o
ice
s
ig
n
al
u
n
d
er
g
o
es
m
u
ltip
le
lay
er
s
o
f
p
r
o
ce
s
s
in
g
,
f
r
o
m
i
n
itial
ca
p
tu
r
e
to
f
ea
tu
r
e
ex
tr
ac
tio
n
,
with
s
ec
u
r
ity
en
h
an
ce
m
e
n
ts
p
r
o
v
id
ed
b
y
th
e
ch
af
f
m
o
d
el
.
T
h
ese
s
tag
es
co
llectiv
ely
f
o
r
m
t
h
e
f
o
u
n
d
atio
n
o
f
a
r
o
b
u
s
t
an
d
ef
f
ec
tiv
e
v
o
ice
a
u
th
en
ticatio
n
s
y
s
tem
.
Fig
u
r
e
6
.
E
x
tr
ac
tio
n
f
ea
tu
r
e
p
r
o
ce
s
s
4
.
1
.
Sens
it
iv
it
y
Sen
s
itiv
ity
is
d
ef
in
ed
as th
e
ev
alu
atio
n
o
f
r
atio
co
n
s
is
tin
g
o
f
ac
tu
al
p
o
s
itiv
e
f
o
r
ca
te
g
o
r
izin
g
th
e
r
ea
l
v
o
ice
to
b
e
th
e
o
r
i
g
in
al.
T
h
e
s
en
s
itiv
ity
is
d
ef
in
ed
as
(
8
)
:
=
+
(
8
)
wh
er
e
r
ef
er
s
to
th
e
ac
tu
al
v
o
ice
id
en
tifie
d
r
ig
h
t
t
o
b
e
th
e
au
th
en
ticated
v
o
ice.
in
d
icat
es
th
e
in
tr
u
d
e
r
v
o
ice,
wh
ic
h
is
wr
o
n
g
ly
au
th
e
n
ticated
to
b
e
in
t
r
u
d
er
v
o
ice.
r
ef
er
s
to
th
e
in
t
r
u
d
er
v
o
ice
is
wr
o
n
g
ly
d
etec
ted
as th
e
au
th
en
ticated
v
o
ice.
T
h
e
in
tr
u
d
er
v
o
ice
is
id
en
tifie
d
c
o
r
r
ec
tly
to
b
e
th
e
in
tr
u
d
er
v
o
ic
e.
4
.
2
.
P
re
cisi
o
n
Pre
cisi
o
n
ca
n
b
e
co
m
p
u
ted
b
y
th
e
r
atio
o
f
tr
u
e
p
o
s
itiv
es
ag
ain
s
t
th
e
tr
u
e
p
o
s
itiv
es
ad
d
e
d
with
f
alse
p
o
s
itiv
es r
esu
lts
o
f
th
e
v
o
ice
s
ig
n
al.
T
h
e
p
r
ec
is
io
n
is
d
ef
in
e
d
as
(
9
)
.
=
+
(
9
)
4
.
3
.
Acc
ura
cy
Acc
u
r
ac
y
d
ef
i
n
es
th
e
o
v
e
r
all
co
r
r
ec
tn
ess
o
f
th
e
m
o
d
el
.
Acc
u
r
ac
y
is
co
m
p
u
te
d
as
f
o
llo
ws:
th
e
r
atio
o
f
th
e
s
u
m
o
f
ac
tu
al
class
if
icati
o
n
p
ar
am
ete
r
s
(
T
p
+
T
n
)
to
th
e
s
u
m
o
f
n
u
m
b
er
o
f
class
if
icatio
n
p
ar
am
eter
s
(
T
p
+
T
n
+
F
p
+
F
n
)
.
=
+
+
+
+
(
1
0
)
4.
4
.
Sig
na
l
t
o
no
is
e
ra
t
io
I
n
s
cien
ce
an
d
en
g
in
ee
r
in
g
,
s
i
g
n
al
-
to
-
n
o
is
e
r
atio
m
ea
s
u
r
es
t
h
e
r
atio
b
etwe
en
th
e
d
eg
r
e
e
o
f
a
g
iv
en
s
ig
n
al
to
th
e
lev
el
o
f
b
ac
k
g
r
o
u
n
d
n
o
is
e.
SNR
is
ex
p
r
ess
ed
as
th
e
r
atio
o
f
s
ig
n
al
p
o
wer
to
t
h
e
n
o
is
e
p
o
wer
,
an
d
its
u
n
it is
g
en
er
ally
d
ec
ib
els.
A
r
atio
g
r
ea
ter
th
a
n
1
:1
im
p
lie
s
m
o
r
e
s
ig
n
al
co
m
p
ar
ed
t
o
n
o
i
s
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
Ho
mo
mo
r
p
h
ic
en
cryp
tio
n
,
p
r
iva
cy
-
p
r
eser
vin
g
fea
tu
r
e
ex
tr
a
ctio
n
,
a
n
d
…
(
K
a
th
ir
esh
Mu
r
u
g
esa
n
)
2157
=
(
1
1
)
4.
5
.
Ro
o
t
m
e
a
n sq
ua
re
dev
ia
t
io
n
Fo
r
th
e
co
m
p
a
r
is
o
n
o
f
th
e
e
n
cr
y
p
tio
n
r
esu
lts
with
less
f
lu
c
tu
atio
n
in
th
e
p
ar
am
ete
r
,
th
e
r
o
o
t
m
ea
n
s
s
q
u
ar
e
d
if
f
e
r
en
ce
(
R
MSD)
is
ca
lcu
lated
.
L
et
µ
1
,
1
(
0
)
r
ef
er
to
th
e
en
cr
y
p
tio
n
r
esu
lt
o
f
th
e
s
ig
n
al
an
d
let
µ
2
,
2
(
0
)
b
e
th
e
n
o
n
-
e
n
cr
y
p
ted
s
ig
n
al
.
T
h
e
R
MSD
b
etwe
en
µ
1
,
1
(
0
)
an
d
µ
2
,
2
(
0
)
is
ex
p
r
ess
ed
as
(
1
2
)
.
R
M
SD
=
(
1
L
×
P
∑
∑
(
f
μ
1
x
1
(
0
)
′
(
i
,
j
)
−
f
μ
2
x
2
(
0
)
′
(
i
,
j
)
)
2
P
−
1
j
=
0
L
−
1
i
=
0
)
1
/
2
(
1
2
)
4.
6
.
No
inv
er
t
ibi
lity
ind
ex
T
h
e
n
o
in
v
er
tib
ilit
y
in
d
ex
(
NI
)
m
ea
s
u
r
es
th
e
r
esil
ien
ce
o
f
b
io
m
et
r
ic
s
y
s
tem
s
ag
ain
s
t
in
v
er
s
io
n
attac
k
s
.
T
o
s
af
eg
u
ar
d
co
n
f
id
e
n
tial
b
io
m
etr
ic
d
ata,
s
o
m
e
alg
o
r
ith
m
s
m
o
d
if
y
b
io
m
etr
ic
in
f
o
r
m
atio
n
to
m
a
k
e
it
ir
r
ev
er
s
ib
le,
en
s
u
r
in
g
th
at
attac
k
er
s
ca
n
n
o
t
r
ec
o
n
s
tr
u
ct
th
e
o
r
ig
in
al
b
io
m
etr
ic
f
ea
tu
r
es
ev
en
if
th
ey
g
ain
ac
ce
s
s
to
th
e
d
atab
ase.
T
h
ese
tr
a
n
s
f
o
r
m
atio
n
s
aim
to
p
r
o
v
id
e
r
o
b
u
s
t
s
ec
u
r
ity
b
y
r
e
n
d
er
i
n
g
s
to
r
e
d
d
ata
u
n
u
s
ab
le
f
o
r
id
en
tity
th
ef
t
o
r
u
n
au
th
o
r
ized
ac
ce
s
s
.
Hig
h
NI
v
alu
es
in
d
icate
b
etter
r
esis
tan
ce
ag
ain
s
t
r
e
v
er
s
e
en
g
in
ee
r
in
g
o
f
b
io
m
etr
ic
tr
aits
,
en
h
a
n
cin
g
o
v
er
all
s
ec
u
r
ity
m
ea
s
u
r
es.
I
n
th
e
Fig
u
r
e
7
s
h
o
ws
th
e
s
e
n
s
itiv
ity
ev
alu
atio
n
is
ca
r
r
ie
d
o
u
t
an
d
it
is
o
b
s
er
v
e
d
th
at
t
h
e
n
ewly
in
tr
o
d
u
ce
d
tech
n
i
q
u
e
y
ield
s
b
etter
p
er
f
o
r
m
an
ce
co
m
p
a
r
ed
t
o
th
e
a
v
ailab
le
tech
n
i
q
u
es
in
th
e
T
ab
le
1
.
T
h
e
r
esu
lts
o
f
ev
alu
atio
n
ar
e
g
iv
e
n
as
b
elo
w:
f
o
r
MFC
C
i
t
is
8
7
.
5
%,
r
etin
a
an
d
f
in
g
er
p
r
i
n
t
b
ased
m
u
lti
–
b
io
m
etr
ic
s
y
s
tem
(
R
FB
M
B
S
)
tech
n
iq
u
e
y
ield
s
8
7
.
9
%,
n
o
is
e
co
n
ce
r
n
e
d
ac
cu
r
ate
v
o
ice
a
u
th
en
ticatio
n
s
y
s
tem
(
NAVA
S)
ac
h
iev
es
9
1
.
5
%,
Su
p
r
av
alv
u
l
ar
ao
r
tic
s
ten
o
s
is
(
SVAS
)
y
ield
s
9
2
.
4
%
an
d
p
ed
iatr
ic
v
asc
u
liti
s
ac
tiv
ity
s
co
r
e
(
PVAS
)
ac
h
iev
es
9
5
.
6
%.
T
h
e
p
er
f
o
r
m
an
ce
n
ewly
in
tr
o
d
u
ce
d
tech
n
iq
u
es
in
ter
m
s
o
f
th
is
p
r
ec
is
io
n
ass
es
s
m
en
t
o
u
tp
er
f
o
r
m
s
th
e
av
ailab
le
t
ec
h
n
iq
u
es.
T
h
e
MFC
C
y
ield
s
8
9
.
7
%,
R
FB
M
B
S
y
ield
s
9
3
.
5
%,
NAVA
S
r
en
d
er
s
9
3
.
8
%.
9
4
.
6
%
ac
h
iev
e
s
SVAS
an
d
PVAS
ac
h
iev
es
9
6
.
2
%.
T
h
e
ev
alu
atio
n
o
f
t
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
d
is
cu
s
s
ed
tech
n
iq
u
e
in
ter
m
s
o
f
ac
cu
r
ac
y
is
m
u
ch
b
etter
th
an
th
e
av
ailab
le
tech
n
iq
u
es
.
MFC
C
tech
n
iq
u
e
y
ield
s
8
2
.
0
0
%,
8
5
.
0
0
%
y
ield
s
R
F
B
MB
S,
NAV
AS
g
iv
es
8
8
%,
SVA
S
g
iv
es
9
1
%
an
d
PV
AS
r
en
d
er
s
9
7
.
5
%.
W
h
en
ev
alu
ated
i
n
ter
m
s
o
f
S
VAS,
th
e
d
is
cu
s
s
ed
tech
n
iq
u
e
PVAS
ac
h
iev
es
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
w
h
er
ea
s
it
is
1
2
.
9
9
%
m
o
r
e
th
a
n
SVAS,
2
5
.
2
8
%
m
o
r
e
th
a
n
NAVA
S,
8
2
.
9
1
%
m
o
r
e
th
a
n
R
FB
MBS
an
d
1
1
4
.
6
5
%
m
o
r
e
th
an
MFC
C
.
I
n
ter
m
s
o
f
R
MSD
m
etr
ic,
PVAS
d
em
o
n
s
tr
ates
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
wh
e
r
e
i
t
is
3
3
.
8
4
%
less
er
co
m
p
ar
ed
t
o
SVAS,
3
9
.
4
3
%
c
o
m
p
ar
ed
t
o
NAVA
S,
5
0
.
5
7
%
co
m
p
ar
ed
t
o
R
FB
M
B
S
an
d
5
4
.
2
5
%
co
m
p
a
r
ed
to
MFC
C
.
Fig
u
r
e
7
.
Ov
e
r
all
p
er
f
o
r
m
an
ce
T
ab
le
1
.
Per
f
o
r
m
an
ce
m
etr
ic
c
o
m
p
ar
is
o
n
v
alu
es
M
e
t
h
o
d
s
S
e
n
s
i
t
i
v
i
t
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
A
c
c
u
r
a
c
y
(
%)
S
N
R
(
d
B
)
R
M
S
D
M
F
C
C
8
7
.
5
0
8
9
.
7
4
8
2
.
0
0
7
.
7
8
0
.
9
4
R
F
B
M
B
S
8
7
.
9
0
9
3
.
5
9
8
5
.
0
0
9
.
1
3
0
.
8
7
NAVAS
9
1
.
5
0
9
3
.
8
3
8
8
.
0
0
1
3
.
3
3
0
.
7
1
S
V
A
S
9
2
.
4
0
9
4
.
6
0
9
1
.
0
0
1
4
.
7
8
0
.
6
5
P
V
A
S
9
5
.
6
0
9
6
.
2
0
9
7
.
5
0
1
6
.
7
0
0
.
4
3
0
10
20
30
40
50
60
70
80
90
100
M
FC
C
R
FBM
BS
N
A
V
A
S
S
V
A
S
PV
A
S
A
c
c
u
ra
c
y
A
lg
o
rit
h
m
Ov
e
ra
l
l
Proc
e
s
s
Sen
s
itivi
t
y
P
re
cision
A
ccu
ra
cy
SN
R
R
MS
D
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
.
2
,
Ap
r
il
20
25
:
2
1
5
0
-
2
1
6
0
2158
5.
CO
NCLU
SI
O
N
I
n
th
is
p
a
p
er
p
r
esen
ts
a
p
io
n
ee
r
in
g
f
r
am
ewo
r
k
th
at
s
ig
n
if
i
ca
n
tly
en
h
a
n
ce
s
p
r
iv
ac
y
p
r
o
t
ec
tio
n
s
in
au
to
m
ated
v
o
ice
au
th
en
ticati
o
n
s
y
s
tem
s
,
ef
f
ec
tiv
ely
ad
d
r
ess
in
g
th
e
in
cr
ea
s
in
g
co
n
ce
r
n
s
s
u
r
r
o
u
n
d
in
g
th
e
s
to
r
ag
e
an
d
p
r
o
ce
s
s
in
g
o
f
s
en
s
itiv
e
v
o
ice
b
io
m
etr
ic
d
ata.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
s
h
av
e
im
p
lem
en
tin
g
ad
v
an
ce
d
en
cr
y
p
tio
n
tech
n
iq
u
es,
p
r
iv
ac
y
-
p
r
eser
v
in
g
f
ea
tu
r
e
e
x
tr
ac
tio
n
,
an
d
a
d
ec
e
n
tr
alize
d
ar
ch
ite
ctu
r
e,
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
en
s
u
r
es
r
o
b
u
s
t
p
r
iv
ac
y
s
af
eg
u
ar
d
s
with
o
u
t
s
ac
r
if
icin
g
au
th
en
ticatio
n
ac
cu
r
a
cy
an
d
ef
f
icien
cy
.
Ad
d
itio
n
ally
,
th
e
in
teg
r
atio
n
o
f
d
if
f
er
en
tial
p
r
iv
ac
y
m
ec
h
a
n
is
m
s
an
d
a
u
s
er
-
ce
n
tr
ic
c
o
n
s
en
t
m
o
d
el
f
u
r
th
er
s
tr
en
g
th
en
s
th
e
f
r
am
ewo
r
k
'
s
ab
ilit
y
to
p
r
o
tect
in
d
iv
id
u
al
p
r
iv
ac
y
.
E
x
p
er
im
en
tal
r
e
s
u
lts
v
alid
ate
th
e
f
r
am
ewo
r
k
'
s
ef
f
ec
tiv
en
ess
,
d
em
o
n
s
tr
atin
g
s
u
p
er
io
r
au
th
en
ticatio
n
ac
cu
r
ac
y
an
d
a
s
u
b
s
tan
tial
r
ed
u
ctio
n
in
p
r
iv
ac
y
r
is
k
s
co
m
p
ar
e
d
to
tr
a
d
itio
n
al
s
y
s
tem
s
.
T
h
is
co
n
tr
ib
u
tio
n
r
e
p
r
esen
ts
a
v
ital
ad
v
an
ce
m
en
t
in
th
e
q
u
est
to
b
alan
ce
s
ec
u
r
ity
a
n
d
p
r
iv
ac
y
with
in
b
io
m
etr
ic
a
u
th
en
ticat
io
n
tech
n
o
lo
g
ies.
RE
F
E
R
E
NC
E
S
[
1
]
L.
S
c
h
ö
n
h
e
r
r
,
T.
Ei
sen
h
o
f
e
r
,
S
.
Ze
i
l
e
r
,
T.
H
o
l
z
,
a
n
d
D
.
K
o
l
o
ssa,
“
I
m
p
e
r
i
o
:
r
o
b
u
st
o
v
e
r
-
t
h
e
-
a
i
r
a
d
v
e
r
sa
r
i
a
l
e
x
a
mp
l
e
s
f
o
r
a
u
t
o
mat
i
c
sp
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
sy
st
e
ms,”
AC
M
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
ren
c
e
Pro
c
e
e
d
i
n
g
S
e
ri
e
s
,
p
p
.
8
4
3
–
8
5
5
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
4
5
/
3
4
2
7
2
2
8
.
3
4
2
7
2
7
6
.
[
2
]
A
.
M
.
A
v
r
a
m,
V
.
P
a
i
ş,
a
n
d
D
.
T
u
f
i
ş,
“
To
w
a
r
d
s
a
R
o
ma
n
i
a
n
e
n
d
-
to
-
e
n
d
a
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
b
a
se
d
o
n
d
e
e
p
sp
e
e
c
h
2
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
R
o
m
a
n
i
a
n
Ac
a
d
e
m
y
S
e
ri
e
s
A
-
M
a
t
h
e
m
a
t
i
c
s
P
h
y
s
i
c
s
T
e
c
h
n
i
c
a
l
S
c
i
e
n
c
e
s
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
,
2
0
2
0
,
v
o
l
.
2
0
,
n
o
.
4
,
p
p
.
3
9
5
–
4
0
2
.
[
3
]
D
.
S
.
P
a
r
k
e
t
a
l
.
,
“
S
p
e
c
a
u
g
me
n
t
:
A
s
i
mp
l
e
d
a
t
a
a
u
g
m
e
n
t
a
t
i
o
n
met
h
o
d
f
o
r
a
u
t
o
m
a
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
P
r
o
c
e
e
d
i
n
g
s
o
f
t
h
e
An
n
u
a
l
C
o
n
f
e
r
e
n
c
e
o
f
t
h
e
I
n
t
e
r
n
a
t
i
o
n
a
l
S
p
e
e
c
h
C
o
m
m
u
n
i
c
a
t
i
o
n
Ass
o
c
i
a
t
i
o
n
,
I
N
T
ER
S
P
EEC
H
,
2
0
1
9
,
p
p
.
2
6
1
3
–
2
6
1
7
,
d
o
i
:
1
0
.
2
1
4
3
7
/
I
n
t
e
r
sp
e
e
c
h
.
2
0
1
9
-
2
6
8
0
.
[
4
]
M
.
Ta
m
a
z
i
n
,
A
.
G
o
u
d
a
,
a
n
d
M
.
K
h
e
d
r
,
“
E
n
h
a
n
c
e
d
a
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
sy
s
t
e
m
b
a
se
d
o
n
e
n
h
a
n
c
i
n
g
p
o
w
e
r
-
n
o
r
ma
l
i
z
e
d
c
e
p
st
r
a
l
c
o
e
f
f
i
c
i
e
n
t
s,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s (S
w
i
t
zer
l
a
n
d
)
,
v
o
l
.
9
,
n
o
.
1
0
,
2
0
1
9
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
9
1
0
2
1
6
6
.
[
5
]
J.
Li
,
R
.
G
a
d
d
e
,
B
.
G
i
n
s
b
u
r
g
,
a
n
d
V
.
L
a
v
r
u
k
h
i
n
,
“
Tr
a
i
n
i
n
g
n
e
u
r
a
l
sp
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
s
y
s
t
e
ms
w
i
t
h
sy
n
t
h
e
t
i
c
sp
e
e
c
h
a
u
g
me
n
t
a
t
i
o
n
,
”
a
rX
i
v
p
r
e
p
r
i
n
t
a
rX
i
v
:
1
8
1
1
.
0
0
7
0
7
,
2
0
1
8
.
[
6
]
H
.
G
u
p
t
a
a
n
d
D
.
G
u
p
t
a
,
“
LPC
a
n
d
L
P
C
C
m
e
t
h
o
d
o
f
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
i
n
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
sy
s
t
e
m
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
1
6
6
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
-
C
l
o
u
d
S
y
st
e
m
a
n
d
B
i
g
D
a
t
a
En
g
i
n
e
e
ri
n
g
,
C
o
n
f
l
u
e
n
c
e
2
0
1
6
,
2
0
1
6
,
p
p
.
4
9
8
–
5
0
2
,
d
o
i
:
1
0
.
1
1
0
9
/
C
O
N
F
LU
EN
C
E.
2
0
1
6
.
7
5
0
8
1
7
1
.
[
7
]
C
.
H
.
H
.
Y
a
n
g
e
t
a
l
.
,
“
D
e
c
e
n
t
r
a
l
i
z
i
n
g
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
w
i
t
h
q
u
a
n
t
u
m
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
a
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
I
C
A
S
S
P,
I
E
EE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Ac
o
u
s
t
i
c
s,
S
p
e
e
c
h
a
n
d
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
-
Pr
o
c
e
e
d
i
n
g
s
,
2
0
2
1
,
p
p
.
6
5
2
3
–
6
5
2
7
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
S
S
P
3
9
7
2
8
.
2
0
2
1
.
9
4
1
3
4
5
3
.
[
8
]
D
.
P
a
l
a
z
,
M
.
M
a
g
i
mai
-
D
o
ss,
a
n
d
R
.
C
o
l
l
o
b
e
r
t
,
“
E
n
d
-
to
-
e
n
d
a
c
o
u
st
i
c
mo
d
e
l
i
n
g
u
si
n
g
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
f
o
r
H
M
M
-
b
a
s
e
d
a
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
,
”
S
p
e
e
c
h
C
o
m
m
u
n
i
c
a
t
i
o
n
,
v
o
l
.
1
0
8
,
p
p
.
1
5
–
3
2
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
sp
e
c
o
m.
2
0
1
9
.
0
1
.
0
0
4
.
[
9
]
F
.
Ta
o
a
n
d
C
.
B
u
ss
o
,
“
E
n
d
-
to
-
e
n
d
a
u
d
i
o
v
i
s
u
a
l
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
s
y
st
e
m
w
i
t
h
mu
l
t
i
t
a
sk
l
e
a
r
n
i
n
g
,
”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
Mu
l
t
i
m
e
d
i
a
,
v
o
l
.
2
3
,
p
p
.
1
–
1
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
TM
M
.
2
0
2
0
.
2
9
7
5
9
2
2
.
[
1
0
]
A
.
H
a
b
b
a
l
,
H
.
H
a
m
o
u
d
a
,
A
.
M
.
A
l
n
a
j
i
m,
S
.
K
h
a
n
,
a
n
d
M
.
F
.
A
l
r
i
f
a
i
e
,
“
P
r
i
v
a
c
y
a
s
a
Li
f
e
s
t
y
l
e
:
e
m
p
o
w
e
r
i
n
g
a
ss
i
s
t
i
v
e
t
e
c
h
n
o
l
o
g
i
e
s
f
o
r
p
e
o
p
l
e
w
i
t
h
d
i
sa
b
i
l
i
t
i
e
s
,
c
h
a
l
l
e
n
g
e
s
a
n
d
f
u
t
u
r
e
d
i
r
e
c
t
i
o
n
s
,
”
J
o
u
rn
a
l
o
f
K
i
n
g
S
a
u
d
U
n
i
v
e
rs
i
t
y
-
C
o
m
p
u
t
e
r
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
3
6
,
n
o
.
4
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
k
s
u
c
i
.
2
0
2
4
.
1
0
2
0
3
9
.
[
1
1
]
J.
G
u
g
l
a
n
i
a
n
d
A
.
N
.
M
i
s
h
r
a
,
“
A
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
s
y
s
t
e
m
w
i
t
h
p
i
t
c
h
d
e
p
e
n
d
e
n
t
f
e
a
t
u
r
e
s
f
o
r
P
u
n
j
a
b
i
l
a
n
g
u
a
g
e
o
n
K
A
LD
I
t
o
o
l
k
i
t
,
”
A
p
p
l
i
e
d
A
c
o
u
st
i
c
s
,
v
o
l
.
1
6
7
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
p
a
c
o
u
s
t
.
2
0
2
0
.
1
0
7
3
8
6
.
[
1
2
]
A
.
W
i
n
u
r
si
t
o
,
R
.
H
i
d
a
y
a
t
,
a
n
d
A
.
B
e
j
o
,
“
I
mp
r
o
v
e
m
e
n
t
o
f
M
F
C
C
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
a
c
c
u
r
a
c
y
u
s
i
n
g
P
C
A
i
n
I
n
d
o
n
e
s
i
a
n
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
2
0
1
8
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
f
o
rm
a
t
i
o
n
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
s
T
e
c
h
n
o
l
o
g
y
,
I
C
O
I
AC
T
2
0
1
8
,
2
0
1
8
,
p
p
.
3
7
9
–
3
8
3
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
O
I
A
C
T.
2
0
1
8
.
8
3
5
0
7
4
8
.
[
1
3
]
A
.
T.
Tr
a
n
,
T
.
D
.
L
u
o
n
g
,
a
n
d
V
.
N
.
H
u
y
n
h
,
“
A
c
o
m
p
r
e
h
e
n
si
v
e
su
r
v
e
y
a
n
d
t
a
x
o
n
o
my
o
n
p
r
i
v
a
c
y
-
p
r
e
ser
v
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
5
7
6
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m.
2
0
2
4
.
1
2
7
3
4
5
.
[
1
4
]
T.
A
.
M
.
C
e
l
i
n
,
T.
N
a
g
a
r
a
j
a
n
,
a
n
d
P
.
V
i
j
a
y
a
l
a
k
s
h
mi
,
“
D
a
t
a
a
u
g
m
e
n
t
a
t
i
o
n
u
si
n
g
v
i
r
t
u
a
l
m
i
c
r
o
p
h
o
n
e
a
r
r
a
y
s
y
n
t
h
e
s
i
s
a
n
d
m
u
l
t
i
-
r
e
so
l
u
t
i
o
n
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
f
o
r
i
so
l
a
t
e
d
w
o
r
d
d
y
s
a
r
t
h
r
i
c
sp
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
,
”
I
EEE
J
o
u
r
n
a
l
o
n
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
S
i
g
n
a
l
Pro
c
e
ssi
n
g
,
v
o
l
.
1
4
,
n
o
.
2
,
p
p
.
3
4
6
–
3
5
4
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
JS
TSP
.
2
0
2
0
.
2
9
7
2
1
6
1
.
[
1
5
]
K
.
A
i
d
a
-
Z
a
d
e
,
A
.
X
o
c
a
y
e
v
,
a
n
d
S
.
R
u
st
a
m
o
v
,
“
S
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
s
u
p
p
o
r
t
v
e
c
t
o
r
ma
c
h
i
n
e
s,
”
i
n
Ap
p
l
i
c
a
t
i
o
n
o
f
I
n
f
o
rm
a
t
i
o
n
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
T
e
c
h
n
o
l
o
g
i
e
s
,
AI
C
T
2
0
1
6
-
C
o
n
f
e
re
n
c
e
Pr
o
c
e
e
d
i
n
g
s
,
2
0
1
7
,
p
p
.
1
–
4
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
I
C
T.
2
0
1
6
.
7
9
9
1
6
6
4
.
[
1
6
]
V
.
P
a
ssr
i
c
h
a
a
n
d
R
.
K
.
A
g
g
a
r
w
a
l
,
“
A
h
y
b
r
i
d
o
f
d
e
e
p
C
N
N
a
n
d
b
i
d
i
r
e
c
t
i
o
n
a
l
LS
TM
f
o
r
a
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
,
v
o
l
.
2
9
,
n
o
.
1
,
p
p
.
1
2
6
1
–
1
2
7
4
,
2
0
2
0
,
d
o
i
:
1
0
.
1
5
1
5
/
j
i
s
y
s
-
2
0
1
8
-
0
3
7
2
.
[
1
7
]
U
.
S
h
r
a
w
a
n
k
a
r
a
n
d
V
.
T
h
a
k
a
r
e
,
“
N
o
i
se
e
s
t
i
m
a
t
i
o
n
a
n
d
n
o
i
se
r
e
m
o
v
a
l
t
e
c
h
n
i
q
u
e
s
f
o
r
sp
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
i
n
a
d
v
e
r
se
e
n
v
i
r
o
n
me
n
t
,
”
I
FI
P
A
d
v
a
n
c
e
s
i
n
I
n
f
o
rm
a
t
i
o
n
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
3
4
0
A
I
C
T,
p
p
.
3
3
6
–
3
4
2
,
2
0
1
0
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
642
-
1
6
3
2
7
-
2
_
4
0
.
[
1
8
]
Y
.
H
.
T
a
g
u
c
h
i
a
n
d
Y
.
M
u
r
a
k
a
m
i
,
“
P
r
i
n
c
i
p
a
l
c
o
m
p
o
n
e
n
t
a
n
a
l
y
si
s
-
b
a
s
e
d
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
a
p
p
r
o
a
c
h
t
o
i
d
e
n
t
i
f
y
c
i
r
c
u
l
a
t
i
n
g
mi
c
r
o
R
N
A
b
i
o
mar
k
e
r
s,
”
PL
o
S
O
N
E
,
v
o
l
.
8
,
n
o
.
6
,
2
0
1
3
,
d
o
i
:
1
0
.
1
3
7
1
/
j
o
u
r
n
a
l
.
p
o
n
e
.
0
0
6
6
7
1
4
.
[
1
9
]
J.
W
.
H
u
n
g
,
J.
S
.
Li
n
,
a
n
d
P
.
J.
W
u
,
“
Emp
l
o
y
i
n
g
r
o
b
u
st
p
r
i
n
c
i
p
a
l
c
o
m
p
o
n
e
n
t
a
n
a
l
y
s
i
s
f
o
r
n
o
i
s
e
-
r
o
b
u
s
t
s
p
e
e
c
h
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
i
n
a
u
t
o
m
a
t
i
c
sp
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
w
i
t
h
t
h
e
s
t
r
u
c
t
u
r
e
o
f
a
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
A
p
p
l
i
e
d
S
y
s
t
e
m
I
n
n
o
v
a
t
i
o
n
,
v
o
l
.
1
,
n
o
.
3
,
p
p
.
1
–
1
4
,
2
0
1
8
,
d
o
i
:
1
0
.
3
3
9
0
/
a
s
i
1
0
3
0
0
2
8
.
[
2
0
]
Y
.
Zh
a
n
g
e
t
a
l
.
,
“
To
w
a
r
d
s e
n
d
-
to
-
e
n
d
sp
e
e
c
h
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,”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
A
n
n
u
a
l
C
o
n
f
e
re
n
c
e
o
f
t
h
e
I
n
t
e
r
n
a
t
i
o
n
a
l
S
p
e
e
c
h
C
o
m
m
u
n
i
c
a
t
i
o
n
Asso
c
i
a
t
i
o
n
,
I
N
T
ERS
P
EE
C
H
,
2
0
1
6
,
v
o
l
.
0
8
-
12
-
S
e
p
t
,
p
p
.
4
1
0
–
4
1
4
,
d
o
i
:
1
0
.
2
1
4
3
7
/
I
n
t
e
r
sp
e
e
c
h
.
2
0
1
6
-
1
4
4
6
.
[
2
1
]
W
.
H
a
n
e
t
a
l
.
,
“
C
o
n
t
e
x
t
n
e
t
:
i
mp
r
o
v
i
n
g
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
o
r
a
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
w
i
t
h
g
l
o
b
a
l
c
o
n
t
e
x
t
,
”
a
rX
i
v
p
re
p
ri
n
t
a
rXi
v
:
2
0
0
5
.
0
3
1
9
1
,
2
0
2
0
.
[
2
2
]
K
.
C
h
o
u
h
a
n
,
A
.
S
i
n
g
h
,
A
.
S
h
r
i
v
a
s
t
a
v
a
,
S
.
A
g
r
a
w
a
l
,
B
.
D
.
S
h
u
k
l
a
,
a
n
d
P
.
S
.
To
mar,
“
S
t
r
u
c
t
u
r
a
l
su
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
f
o
r
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
c
l
a
ssi
f
i
c
a
t
i
o
n
w
i
t
h
C
N
N
a
p
p
r
o
a
c
h
,
”
i
n
2
0
2
1
9
th
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
C
y
b
e
r
a
n
d
I
T
S
e
rv
i
c
e
M
a
n
a
g
e
m
e
n
t
,
C
I
T
S
M
2
0
2
1
,
2
0
2
1
,
p
p
.
1
–
7
,
d
o
i
:
1
0
.
1
1
0
9
/
C
I
TSM
5
2
8
9
2
.
2
0
2
1
.
9
5
8
8
9
1
8
.
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
Ho
mo
mo
r
p
h
ic
en
cryp
tio
n
,
p
r
iva
cy
-
p
r
eser
vin
g
fea
tu
r
e
ex
tr
a
ctio
n
,
a
n
d
…
(
K
a
th
ir
esh
Mu
r
u
g
esa
n
)
2159
[
2
3
]
A
.
A
l
so
b
h
a
n
i
,
H
.
M
.
A
.
A
l
a
b
b
o
o
d
i
,
a
n
d
H
.
M
a
h
d
i
,
“
S
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
c
o
n
v
o
l
u
t
i
o
n
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s,”
J
o
u
r
n
a
l
o
f
Ph
y
s
i
c
s:
C
o
n
f
e
r
e
n
c
e
S
e
ri
e
s
,
v
o
l
.
1
9
7
3
,
n
o
.
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
4
2
-
6
5
9
6
/
1
9
7
3
/
1
/
0
1
2
1
6
6
.
[
2
4
]
N
.
K
h
a
l
i
d
,
A
.
Q
a
y
y
u
m
,
M
.
B
i
l
a
l
,
A
.
A
l
-
F
u
q
a
h
a
,
a
n
d
J.
Q
a
d
i
r
,
“
P
r
i
v
a
c
y
-
p
r
e
serv
i
n
g
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
i
n
h
e
a
l
t
h
c
a
r
e
:
Te
c
h
n
i
q
u
e
s
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s,
”
C
o
m
p
u
t
e
rs
i
n
B
i
o
l
o
g
y
a
n
d
M
e
d
i
c
i
n
e
,
v
o
l
.
1
5
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
b
i
o
m
e
d
.
2
0
2
3
.
1
0
6
8
4
8
.
[
2
5
]
A
.
I
f
t
i
k
h
a
r
,
K
.
N
.
Q
u
r
e
s
h
i
,
M
.
S
h
i
r
a
z
,
a
n
d
S
.
A
l
b
a
h
l
i
,
“
S
e
c
u
r
i
t
y
,
t
r
u
s
t
a
n
d
p
r
i
v
a
c
y
r
i
s
k
s,
r
e
s
p
o
n
ses
,
a
n
d
so
l
u
t
i
o
n
s
f
o
r
h
i
g
h
-
sp
e
e
d
smar
t
c
i
t
i
e
s
n
e
t
w
o
r
k
s:
a
sy
s
t
e
m
a
t
i
c
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
,
”
J
o
u
r
n
a
l
o
f
K
i
n
g
S
a
u
d
U
n
i
v
e
rsi
t
y
-
C
o
m
p
u
t
e
r
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
3
5
,
n
o
.
9
,
p
.
1
0
1
7
8
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
k
s
u
c
i
.
2
0
2
3
.
1
0
1
7
8
8
.
[
2
6
]
S
.
S
i
n
g
h
,
S
.
R
a
t
h
o
r
e
,
O
.
A
l
f
a
r
r
a
j
,
A
.
T
o
l
b
a
,
a
n
d
B
.
Y
o
o
n
,
“
A
f
r
a
mew
o
r
k
f
o
r
p
r
i
v
a
c
y
-
p
r
e
s
e
r
v
a
t
i
o
n
o
f
I
o
T
h
e
a
l
t
h
c
a
r
e
d
a
t
a
u
s
i
n
g
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
a
n
d
b
l
o
c
k
c
h
a
i
n
t
e
c
h
n
o
l
o
g
y
,
”
F
u
t
u
r
e
G
e
n
e
r
a
t
i
o
n
C
o
m
p
u
t
e
r
S
y
s
t
e
m
s
,
v
o
l
.
1
2
9
,
p
p
.
3
8
0
–
3
8
8
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
f
u
t
u
r
e
.
2
0
2
1
.
1
1
.
0
2
8
.
[
2
7
]
C
.
S
a
n
d
e
e
p
a
,
B
.
S
i
n
i
a
r
s
k
i
,
N
.
K
o
u
r
t
e
l
l
i
s,
S
.
W
a
n
g
,
a
n
d
M
.
Li
y
a
n
a
g
e
,
“
A
s
u
r
v
e
y
o
n
p
r
i
v
a
c
y
f
o
r
B
5
G
/
6
G
:
n
e
w
p
r
i
v
a
c
y
c
h
a
l
l
e
n
g
e
s
,
a
n
d
r
e
s
e
a
r
c
h
d
i
r
e
c
t
i
o
n
s
,
”
J
o
u
r
n
a
l
o
f
I
n
d
u
s
t
ri
a
l
I
n
f
o
rm
a
t
i
o
n
I
n
t
e
g
r
a
t
i
o
n
,
v
o
l
.
3
0
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
i
i
.
2
0
2
2
.
1
0
0
4
0
5
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
K
a
th
ire
sh
Mu
r
u
g
e
sa
n
is
a
n
a
ss
istan
t
p
r
o
fe
ss
o
r
i
n
th
e
De
p
a
rt
m
e
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
Ra
th
i
n
a
m
Co
ll
e
g
e
o
f
Ar
ts
a
n
d
S
c
ien
c
e
C
o
imb
a
to
re
.
He
r
e
c
e
iv
e
d
h
is
B.
S
c
.
d
e
g
re
e
i
n
m
a
th
e
m
a
ti
c
s
fro
m
Bh
a
ra
th
iar
Un
iv
e
rsity
.
M
CA
d
e
g
re
e
a
t
B
h
a
ra
th
iar
Un
i
v
e
rsity
a
n
d
M
.
P
h
i
l
.
d
e
g
re
e
a
t
Ka
rp
a
g
a
m
Un
iv
e
rsity
.
P
h
.
D
.
d
e
g
re
e
fro
m
Bh
a
ra
th
iar
Un
iv
e
rsit
y
.
He
h
a
s
m
o
re
th
a
n
1
3
y
e
a
rs
o
f
tea
c
h
i
n
g
e
x
p
e
rie
n
c
e
a
n
d
h
e
p
u
b
l
ish
e
d
8
re
se
a
rc
h
a
rti
c
les
.
His
re
se
a
rc
h
i
n
tere
st
in
c
lu
d
e
s
n
e
two
rk
se
c
u
rit
y
,
c
o
m
p
il
e
r
d
e
sig
n
,
a
n
d
so
ftwa
re
e
n
g
i
n
e
e
rin
g
.
He
h
a
s
fin
ish
e
d
o
n
e
UG
C
sp
o
n
so
re
d
m
in
o
r
re
se
a
rc
h
p
ro
jec
t
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t:
k
a
t
h
ires
h
c
s8
3
@
g
m
a
il
.
c
o
m
.
La
v
a
n
y
a
S
u
b
b
a
r
a
y
a
lu
Ra
m
a
m
u
r
th
y
is
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
t
KPR
C
o
ll
e
g
e
o
f
Arts
S
c
ie
n
c
e
a
n
d
Re
se
a
rc
h
,
Co
imb
a
t
o
re
.
S
h
e
e
a
rn
e
d
h
e
r
P
h
.
D
.
in
c
o
m
p
u
ter
sc
ien
c
e
fro
m
Bh
a
ra
th
iar
Un
i
v
e
rsity
.
Wi
t
h
o
v
e
r
1
9
y
e
a
rs
o
f
a
c
a
d
e
m
ic
e
x
p
e
rien
c
e
,
sh
e
is
a
lea
d
in
g
e
x
p
e
rt
in
d
a
ta
m
in
in
g
a
n
d
m
a
c
h
in
e
lea
rn
in
g
.
Dr.
La
v
a
n
y
a
h
a
s
p
u
b
li
sh
e
d
1
5
p
a
p
e
rs i
n
p
re
stig
io
u
s S
c
o
p
u
s
-
i
n
d
e
x
e
d
jo
u
rn
a
ls,
p
e
e
r
-
r
e
v
iew
e
d
p
u
b
li
c
a
ti
o
n
s,
o
n
li
n
e
b
o
o
k
s,
a
n
d
h
a
s
p
re
se
n
ted
m
o
re
t
h
a
n
1
5
p
a
p
e
rs
a
t
n
a
ti
o
n
a
l
a
n
d
in
t
e
rn
a
ti
o
n
a
l
c
o
n
fe
re
n
c
e
s.
He
r
wo
rk
a
lso
in
c
l
u
d
e
s
p
a
ten
ts,
u
n
d
e
rsc
o
rin
g
h
e
r
sig
n
ifi
c
a
n
t
c
o
n
tri
b
u
ti
o
n
s
to
t
h
e
field
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
d
rsrla
v
a
n
y
a
@
g
m
a
il
.
c
o
m
,
lav
a
n
y
a
.
s.r@k
p
rc
a
s.a
c
.
in
.
Bo
o
p
a
th
i
P
a
l
a
n
isa
m
y
is
a
n
a
ss
istan
t
p
ro
fe
ss
o
r
in
t
h
e
De
p
a
rt
m
e
n
t
o
f
Co
m
p
u
ter
Ap
p
li
c
a
ti
o
n
s
a
t
Ne
h
ru
Arts
a
n
d
S
c
ien
c
e
Co
ll
e
g
e
,
Co
imb
a
to
re
.
He
is
p
u
rsu
i
n
g
a
P
h
.
D.
d
e
g
re
e
in
c
o
m
p
u
ter
sc
ien
c
e
a
t
Dr.
S
NS
R
a
jala
k
sh
m
i
Co
ll
e
g
e
o
f
Arts
a
n
d
S
c
ien
c
e
,
Co
imb
a
to
re
wit
h
a
sp
e
c
ializa
ti
o
n
in
d
a
ta
m
in
i
n
g
.
He
h
a
s
m
o
re
th
a
n
1
5
+
y
e
a
rs
o
f
e
x
p
e
rien
c
e
in
c
o
re
c
o
m
p
u
te
r
sc
ien
c
e
su
b
jec
ts.
He
h
a
s
p
u
b
li
s
h
e
d
m
o
re
a
rti
c
le
p
a
p
e
rs
i
n
re
p
u
te
d
n
a
ti
o
n
a
l
a
n
d
i
n
tern
a
ti
o
n
a
l
jo
u
r
n
a
ls
a
n
d
a
ls
o
fi
ll
e
d
m
o
re
t
h
a
n
p
a
tt
e
rn
p
a
p
e
rs
i
n
th
e
sa
m
e
.
He
h
a
s
b
e
e
n
a
c
ti
n
g
a
s
a
re
so
u
rc
e
p
e
rso
n
i
n
t
h
e
v
a
ri
o
u
s i
n
stit
u
ti
o
n
s.
He
c
a
n
b
e
c
o
n
tac
ted
a
t:
b
o
o
p
a
th
i
m
c
a
1
9
8
5
@g
m
a
il
.
c
o
m
.
Ya
m
in
i
Cha
n
d
r
a
se
k
a
r
is
wo
rk
i
n
g
a
s
a
ss
o
c
iate
p
r
o
fe
ss
o
r
a
n
d
h
e
a
d
–
BCA
i
n
Co
m
p
u
ter
S
c
ien
c
e
De
p
a
rtme
n
t.
S
h
e
re
c
e
iv
e
d
h
e
r
P
h
.
D.
in
c
o
m
p
u
ter
a
p
p
li
c
a
ti
o
n
s
fr
o
m
An
n
a
Un
iv
e
rsity
d
u
rin
g
2
0
1
4
.
He
r
a
re
a
o
f
re
se
a
rc
h
in
c
lu
d
e
s
d
a
ta
m
in
in
g
,
a
rti
ficia
l
in
tell
ig
e
n
c
e
,
d
e
e
p
lea
rn
in
g
a
n
d
m
a
c
h
i
n
e
lea
rn
in
g
.
S
h
e
h
a
s
se
rv
e
d
a
s
a
re
so
u
rc
e
p
e
rso
n
in
F
DP,
se
m
in
a
r
a
n
d
se
rv
e
d
a
s
a
c
h
a
ir
fo
r
n
a
ti
o
n
a
l
a
n
d
in
ter
n
a
ti
o
n
a
l
c
o
n
fe
re
n
c
e
.
S
h
e
h
a
s
o
rg
a
n
ize
d
v
a
rio
u
s
in
tern
a
ti
o
n
a
l
c
o
n
fe
re
n
c
e
,
F
DP,
w
o
rk
sh
o
p
s
a
n
d
se
m
in
a
rs.
S
h
e
h
a
s
p
u
b
li
s
h
e
d
a
r
o
u
n
d
3
0
+
p
a
p
e
rs
in
n
a
ti
o
n
a
l
a
n
d
i
n
tern
a
ti
o
n
a
l
Jo
u
rn
a
ls
wh
ich
i
n
c
lu
d
e
s
S
COPUS
,
Th
o
m
so
n
Re
u
ters
,
UG
C,
IEE
E,
e
tc.
a
n
d
a
lso
p
re
se
n
ted
2
5
+
p
a
p
e
rs i
n
v
a
rio
u
s n
a
ti
o
n
a
l
a
n
d
i
n
t
e
rn
a
ti
o
n
a
l
c
o
n
fe
re
n
c
e
s.
S
h
e
h
a
s
p
u
b
l
ish
e
d
5
b
o
o
k
s.
S
h
e
is
th
e
b
o
a
r
d
o
f
stu
d
ies
m
e
m
b
e
r
f
o
r
b
o
th
P
G
-
CS
a
n
d
UG
-
IT
i
n
Bh
a
ra
th
iar
Un
i
v
e
rsity
Co
imb
a
to
r
e
.
Re
c
e
iv
e
d
sp
o
n
s
o
r
to
c
o
n
d
u
c
te
d
F
DP
u
n
d
e
r
AICTE
ATAL
F
DP
S
c
h
e
m
e
.
S
h
e
c
a
n
b
e
c
o
n
tac
t
e
d
a
t
e
m
a
il
:
y
a
m
in
ics
@s
rc
w.ac
.
in
.
Evaluation Warning : The document was created with Spire.PDF for Python.