I
nte
rna
t
io
na
l J
o
urna
l o
f
Rec
o
nfig
ura
ble a
nd
E
m
be
dd
e
d Sy
s
t
e
m
s
(
I
J
R
E
S)
Vo
l.
1
5
,
No
.
1
,
Ma
r
ch
20
2
6
,
p
p
.
1
94
~
2
0
3
I
SS
N:
2089
-
4864
,
DOI
:
1
0
.
1
1
5
9
1
/i
j
r
es
.
v
1
5
.
i
1
.
pp
1
94
-
2
03
194
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
es.ia
esco
r
e.
co
m
EL
LM
W:
a
n enh
a
nced vis
io
n
–
la
ng
ua
g
e
m
o
del
for re
lia
ble t
ex
t
ex
traction f
ro
m
m
a
nua
lly
co
m
po
sed scrip
ts
Dhiv
y
a
V
enk
a
t
es
h
,
B
rint
ha
Ra
j
a
k
u
m
a
ri
S
iv
a
ra
j
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
,
B
h
a
r
a
t
h
I
n
st
i
t
u
t
e
o
f
H
i
g
h
e
r
E
d
u
c
a
t
i
o
n
a
n
d
R
e
se
a
r
c
h
,
C
h
e
n
n
a
i
,
I
n
d
i
a
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
2
4
,
2
0
2
5
R
ev
i
s
ed
J
an
1
2
,
2
0
2
6
A
cc
ep
ted
Feb
1
7
,
2
0
2
6
W
h
il
e
c
o
n
v
e
n
ti
o
n
a
l
o
p
ti
c
a
l
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
(
OCR
)
sy
ste
m
s
c
a
n
d
ig
it
ize
tex
t,
th
e
y
stru
g
g
l
e
w
it
h
d
iv
e
rse
h
a
n
d
w
rit
in
g
sty
les
,
n
o
isy
in
p
u
ts,
a
n
d
u
n
stru
c
t
u
re
d
lay
o
u
ts,
li
m
it
in
g
th
e
ir
e
ff
e
c
ti
v
e
n
e
ss
.
T
h
is
stu
d
y
p
ro
p
o
se
s
e
n
h
a
n
c
e
d
larg
e
lan
g
u
a
g
e
m
o
d
e
l
w
h
isp
e
re
r
(
E
LL
M
W
),
a
v
isio
n
–
lan
g
u
a
g
e
f
ra
m
e
w
o
rk
f
o
r
a
c
c
u
ra
te
tex
t
e
x
tr
a
c
ti
o
n
(T
E)
f
ro
m
f
u
ll
y
h
a
n
d
w
rit
ten
sc
rip
ts.
T
h
e
m
e
th
o
d
o
lo
g
y
in
teg
ra
tes
a
d
v
a
n
c
e
d
p
re
p
r
o
c
e
ss
in
g
(n
o
ise
re
d
u
c
ti
o
n
,
b
in
a
riza
ti
o
n
,
a
n
d
sk
e
w
c
o
rre
c
ti
o
n
)
,
d
e
e
p
lea
rn
in
g
–
b
a
se
d
h
a
n
d
w
rit
in
g
re
c
o
g
n
it
io
n
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
–
lo
n
g
sh
o
rt
-
term
m
e
m
o
ry
(CNN
–
L
S
T
M
),
a
n
d
L
L
M
-
b
a
se
d
p
o
st
-
c
o
rre
c
ti
o
n
t
o
e
n
su
re
c
o
n
tex
t
-
a
w
a
r
e
a
n
d
stru
c
tu
ra
ll
y
c
o
h
e
re
n
t
o
u
tp
u
ts.
T
h
e
sy
ste
m
c
o
n
v
e
rts
sc
a
n
n
e
d
im
a
g
e
s
,
p
o
rtab
le
d
o
c
u
m
e
n
t
f
o
rm
a
t
s
(
P
DFs
)
,
a
n
d
irreg
u
larly
f
o
r
m
a
tt
e
d
a
n
s
w
e
r
s
h
e
e
ts
in
to
m
a
c
h
in
e
-
re
a
d
a
b
le
tex
t,
w
h
il
e
a
u
to
m
a
ti
c
a
ll
y
c
o
rre
c
ti
n
g
e
rro
rs
in
sp
e
ll
i
n
g
,
g
ra
m
m
a
r,
a
n
d
lay
o
u
t.
Ex
p
e
rime
n
tal
e
v
a
lu
a
ti
o
n
o
n
a
c
u
ra
ted
d
a
tas
e
t
o
f
h
a
n
d
w
rit
ten
e
x
a
m
in
a
ti
o
n
a
n
sw
e
r
sc
rip
ts
(
HEA
S
)
d
e
m
o
n
stra
tes
th
a
t
EL
L
M
W
a
c
h
iev
e
s
9
7
.
8
%
a
c
c
u
ra
c
y
,
1
.
0
4
%
-
c
h
a
ra
c
ter
e
rro
r
ra
te
(
CER)
,
a
n
d
3
.
2
4
%
-
w
o
rd
e
rro
r
ra
t
e
,
o
u
t
p
e
rf
o
rm
in
g
wid
e
ly
u
se
d
OCR
to
o
ls
in
c
l
u
d
i
n
g
T
e
ss
e
ra
c
t,
Eas
y
OCR,
G
o
o
g
le
Clo
u
d
V
isio
n
(G
CV
)
,
P
a
d
d
leO
CR,
A
BBYY
F
i
n
e
Re
a
d
e
r,
a
n
d
T
ra
n
sy
m
OCR.
T
h
e
re
su
lt
s
h
ig
h
li
g
h
t
t
h
e
m
o
d
e
l’s
ro
b
u
stn
e
ss
a
c
ro
ss
v
a
r
y
in
g
h
a
n
d
w
rit
in
g
st
y
les
,
n
o
is
y
b
a
c
k
g
ro
u
n
d
s,
a
n
d
c
o
m
p
lex
d
o
c
u
m
e
n
t
stru
c
tu
re
s.
K
ey
w
o
r
d
s
:
E
x
tr
ac
tin
g
te
x
t
L
L
MW
h
i
s
p
er
er
L
o
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
Ma
ch
i
n
e
lear
n
i
n
g
Op
tical
ch
ar
ac
ter
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
:
Dh
i
v
y
a
V
e
n
k
ates
h
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
B
h
ar
ath
I
n
s
tit
u
te
o
f
H
ig
h
er
E
d
u
ca
tio
n
an
d
R
esear
c
h
C
h
e
n
n
ai,
I
n
d
ia
E
m
ai
l: d
h
iv
y
a
m
asc
@
g
m
ai
l.c
o
m
1.
I
NT
RO
D
UCT
I
O
N
B
ac
k
g
r
o
u
n
d
an
d
p
r
o
b
lem
o
v
er
v
ie
w
:
e
v
al
u
ati
n
g
h
a
n
d
w
r
itt
en
ex
a
m
i
n
atio
n
a
n
s
w
er
s
cr
ip
ts
(
HE
AS)
m
an
u
all
y
r
e
m
ain
s
t
h
e
s
ta
n
d
ar
d
ap
p
r
o
ac
h
in
ed
u
ca
tio
n
al
i
n
s
t
itu
tio
n
s
w
o
r
ld
w
id
e.
W
h
ile
h
u
m
an
g
r
ad
in
g
allo
ws
teac
h
er
s
to
ass
es
s
n
u
a
n
ce
d
s
tu
d
en
t
u
n
d
er
s
ta
n
d
in
g
,
r
ea
s
o
n
i
n
g
,
an
d
cr
ea
tiv
it
y
,
it
is
in
h
er
en
tl
y
ti
m
e
-
co
n
s
u
m
i
n
g
,
in
co
n
s
is
te
n
t,
a
n
d
p
r
o
n
e
to
b
ia
s
,
p
ar
ticu
lar
l
y
d
u
r
i
n
g
lar
g
e
-
s
c
a
le
ex
a
m
i
n
atio
n
s
w
it
h
h
u
n
d
r
e
d
s
o
r
th
o
u
s
an
d
s
o
f
an
s
w
er
s
cr
ip
ts
.
R
esear
c
h
in
d
i
ca
tes
th
at
teac
h
er
s
m
a
y
s
p
e
n
d
2
0
–
2
5
m
i
n
u
te
s
p
er
s
cr
ip
t,
d
ep
en
d
in
g
o
n
le
n
g
t
h
an
d
co
m
p
le
x
it
y
,
w
h
ich
ca
n
l
ea
d
to
s
ig
n
i
f
ica
n
t
d
ela
y
s
i
n
r
etu
r
n
i
n
g
r
es
u
lt
s
to
s
tu
d
e
n
ts
[1
]
-
[
4]
.
Fu
r
th
er
m
o
r
e,
m
an
u
al
e
v
al
u
atio
n
in
tr
o
d
u
ce
s
v
ar
iab
ilit
y
i
n
s
co
r
i
n
g
d
u
e
to
s
u
b
j
ec
tiv
e
in
ter
p
r
etatio
n
s
,
f
ati
g
u
e,
a
n
d
in
co
n
s
is
te
n
cie
s
ac
r
o
s
s
d
if
f
e
r
en
t
ev
al
u
ato
r
s
,
co
m
p
r
o
m
is
i
n
g
t
h
e
f
air
n
e
s
s
an
d
r
eliab
ilit
y
o
f
ac
ad
e
m
ic
ass
es
s
m
en
t
s
[5
]
-
[
8]
.
T
h
e
ad
v
en
t
o
f
d
ig
ital
a
s
s
es
s
m
en
t
to
o
ls
h
a
s
p
ar
tiall
y
alle
v
iat
ed
th
is
b
u
r
d
en
;
h
o
w
e
v
er
,
m
o
s
t
s
y
s
te
m
s
ar
e
d
esig
n
ed
f
o
r
m
u
ltip
le
-
ch
o
ice
o
r
d
ig
itall
y
s
u
b
m
itted
as
s
i
g
n
m
e
n
ts
,
lea
v
i
n
g
h
a
n
d
w
r
itte
n
r
esp
o
n
s
es
lar
g
el
y
u
n
a
u
to
m
ated
.
I
n
ad
d
itio
n
,
m
a
n
u
al
tr
an
s
cr
ip
tio
n
o
r
d
ig
itizat
io
n
o
f
t
h
ese
h
a
n
d
w
r
itte
n
s
cr
i
p
ts
f
o
r
ar
ch
iv
a
l
o
r
an
al
y
tic
p
u
r
p
o
s
es
i
s
co
s
tl
y
a
n
d
er
r
o
r
-
p
r
o
n
e
[9
]
-
[
12]
.
C
o
n
s
eq
u
en
tl
y
,
t
h
er
e
is
a
p
r
ess
in
g
n
e
ed
f
o
r
an
au
to
m
ated
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
E
LL
M
W
:
a
n
en
h
a
n
ce
d
visi
o
n
–
la
n
g
u
a
g
e
mo
d
el
f
o
r
r
elia
b
le
text
ex
tr
a
ctio
n
fr
o
m
…
(
Dh
ivya
V
en
ka
tesh
)
195
ac
cu
r
ate,
an
d
s
ca
lab
le
s
o
l
u
ti
o
n
ca
p
ab
le
o
f
co
n
v
er
ti
n
g
h
an
d
w
r
i
tten
tex
t
i
n
to
m
ac
h
in
e
-
r
ea
d
ab
le
f
o
r
m
ats
,
p
r
eser
v
in
g
s
e
m
a
n
ti
c
i
n
te
g
r
it
y
,
an
d
s
u
p
p
o
r
tin
g
d
o
w
n
s
tr
ea
m
ed
u
ca
tio
n
al
ap
p
licatio
n
s
s
u
c
h
as
a
u
to
m
ated
g
r
ad
in
g
,
p
lag
iar
is
m
d
etec
tio
n
,
an
d
f
ee
d
b
ac
k
g
e
n
er
atio
n
[
1
3
]
-
[
16]
.
L
i
m
itatio
n
s
o
f
ex
i
s
ti
n
g
m
et
h
o
d
s
:
co
n
v
e
n
tio
n
al
o
p
tical
ch
ar
ac
ter
r
ec
o
g
n
itio
n
(
OC
R
)
s
y
s
te
m
s
,
in
cl
u
d
in
g
T
ess
er
a
ct,
E
as
y
O
C
R
,
P
ad
d
leOCR
,
A
B
B
YY
Fin
eRead
er
,
an
d
Go
o
g
le
C
lo
u
d
Vis
io
n
(
G
C
V)
,
h
a
v
e
b
ee
n
w
id
el
y
u
s
ed
to
d
ig
itiz
e
p
r
in
ted
an
d
s
o
m
e
h
a
n
d
w
r
itten
te
x
t.
W
h
ile
t
h
ese
s
y
s
te
m
s
e
x
h
ib
it
r
o
b
u
s
t
p
er
f
o
r
m
a
n
ce
o
n
p
r
in
ted
d
o
cu
m
en
ts
w
ith
s
ta
n
d
ar
d
f
o
n
ts
a
n
d
la
y
o
u
t
s
,
t
h
e
y
s
tr
u
g
g
le
to
h
an
d
le
t
h
e
i
n
h
er
e
n
t
v
ar
iab
ilit
y
o
f
h
u
m
a
n
h
an
d
w
r
itin
g
[
1
7
]
-
[
21]
.
C
h
allen
g
es
a
r
is
e
d
u
e
to
d
iv
er
s
e
h
an
d
w
r
iti
n
g
s
t
y
le
s
,
v
ar
y
i
n
g
s
tr
o
k
e
th
ic
k
n
es
s
es,
in
co
n
s
i
s
te
n
t
s
p
ac
in
g
,
cu
r
s
iv
e
s
cr
ip
ts
,
an
d
th
e
in
cl
u
s
io
n
o
f
s
y
m
b
o
ls
,
d
i
ag
r
a
m
s
,
tab
les,
an
d
m
at
h
e
m
a
tical
n
o
ta
t
io
n
s
.
Mo
r
eo
v
er
,
tr
ad
itio
n
al
OC
R
m
et
h
o
d
s
g
en
er
all
y
f
o
llo
w
a
t
w
o
-
s
ta
g
e
p
ip
elin
e:
te
x
t
r
eg
io
n
d
etec
tio
n
f
o
llo
w
ed
b
y
ch
ar
ac
ter
r
ec
o
g
n
i
tio
n
.
E
r
r
o
r
s
in
eith
er
s
tag
e
p
r
o
p
ag
ate,
r
ed
u
cin
g
o
v
er
all
ac
cu
r
ac
y
.
T
h
ese
s
y
s
te
m
s
also
lack
co
n
te
x
t
u
al
a
n
d
s
e
m
a
n
tic
u
n
d
er
s
ta
n
d
in
g
,
m
ak
in
g
t
h
e
m
i
n
ca
p
ab
le
o
f
co
r
r
ec
tin
g
m
i
s
r
ec
o
g
n
ized
w
o
r
d
s
o
r
in
ter
p
r
etin
g
d
o
m
ain
-
s
p
ec
if
i
c
ter
m
i
n
o
lo
g
y
.
Fo
r
ex
a
m
p
le,
OC
R
e
n
g
i
n
es
m
a
y
m
is
r
ea
d
“O”
as
“
0
”
o
r
f
ail
to
co
r
r
ec
tly
in
ter
p
r
et
m
u
lti
-
li
n
e
eq
u
atio
n
s
a
n
d
an
n
o
tatio
n
s
co
m
m
o
n
in
ex
a
m
i
n
atio
n
s
cr
ip
ts
.
Su
c
h
li
m
itatio
n
s
r
esu
lt
in
h
ig
h
er
r
o
r
r
ates,
r
ed
u
ce
d
r
eliab
ilit
y
,
an
d
p
o
o
r
p
er
f
o
r
m
an
ce
i
n
d
o
w
n
s
tr
ea
m
tas
k
s
,
i
n
clu
d
i
n
g
au
to
m
ated
g
r
ad
in
g
,
k
e
y
w
o
r
d
ex
tr
ac
tio
n
,
an
d
s
e
m
a
n
tic
an
al
y
s
i
s
,
h
i
g
h
li
g
h
t
in
g
th
e
in
ad
eq
u
ac
y
o
f
ex
is
tin
g
OC
R
s
o
l
u
tio
n
s
f
o
r
co
m
p
le
x
h
a
n
d
w
r
itte
n
a
ca
d
e
m
ic
d
o
cu
m
en
t
s
[
2
2
]
-
[
25]
.
R
ec
en
t
ad
v
a
n
ce
s
i
n
d
ee
p
lear
n
in
g
h
av
e
i
n
tr
o
d
u
ce
d
co
n
v
o
lu
ti
o
n
al
n
eu
r
al
n
et
w
o
r
k
(
C
NN
)
,
lo
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
(
L
ST
M
)
,
an
d
tr
an
s
f
o
r
m
er
-
b
ased
m
o
d
els
f
o
r
h
an
d
w
r
iti
n
g
r
ec
o
g
n
itio
n
.
W
h
ile
t
h
ese
m
et
h
o
d
s
i
m
p
r
o
v
e
c
h
ar
ac
ter
-
le
v
el
r
ec
o
g
n
i
ti
o
n
,
t
h
e
y
s
till
f
ail
to
i
n
teg
r
ate
li
n
g
u
i
s
tic
co
n
te
x
t,
p
o
s
t
-
co
r
r
ec
tio
n
,
a
n
d
s
tr
u
ct
u
r
ed
o
u
tp
u
t
g
en
er
atio
n
,
w
h
ic
h
ar
e
ess
e
n
tial
f
o
r
r
ea
l
-
wo
r
ld
ac
ad
em
ic
ap
p
licatio
n
s
.
R
esear
ch
g
ap
:
d
esp
ite
p
r
o
g
r
es
s
i
n
O
C
R
a
n
d
d
ee
p
lear
n
i
n
g
-
b
ased
h
a
n
d
w
r
i
tin
g
r
ec
o
g
n
it
io
n
,
ac
cu
r
at
el
y
ex
tr
ac
ti
n
g
s
tr
u
ct
u
r
ed
an
d
co
n
tex
t
u
all
y
co
h
er
en
t
te
x
t
f
r
o
m
f
u
ll
y
h
an
d
w
r
itte
n
ex
a
m
i
n
atio
n
s
cr
ip
ts
r
e
m
ai
n
s
a
s
ig
n
i
f
ica
n
t c
h
al
len
g
e.
E
x
i
s
ti
n
g
m
o
d
els ar
e
li
m
ited
i
n
th
e
ir
ab
ilit
y
to
:
−
Han
d
le
co
m
p
lex
d
o
cu
m
e
n
t la
y
o
u
ts
,
in
c
lu
d
i
n
g
m
u
lti
-
co
lu
m
n
f
o
r
m
at
s
,
tab
les,
f
i
g
u
r
es,
an
d
m
a
r
g
in
n
o
tes.
−
R
ec
o
g
n
ize
d
iv
er
s
e
h
an
d
w
r
iti
n
g
s
t
y
les,
i
n
cl
u
d
in
g
cu
r
s
iv
e,
m
i
x
ed
,
an
d
m
u
ltil
i
n
g
u
a
l scr
ip
ts
.
−
C
o
r
r
ec
t e
r
r
o
r
s
u
s
i
n
g
co
n
te
x
t
-
a
w
ar
e
o
r
la
n
g
u
a
g
e
-
a
w
ar
e
m
ec
h
an
is
m
s
,
leav
in
g
o
u
tp
u
t
s
p
r
o
n
e
to
s
e
m
a
n
tic
an
d
s
y
n
tactic
i
n
ac
c
u
r
ac
ies.
−
Su
p
p
o
r
t
d
o
w
n
s
tr
ea
m
ed
u
ca
tio
n
al
ap
p
licatio
n
s
th
at
r
eq
u
ir
e
s
t
r
u
ctu
r
ed
,
m
ac
h
i
n
e
-
r
ea
d
ab
le,
an
d
s
e
m
an
t
icall
y
r
eliab
le
d
ata
f
o
r
au
to
m
ated
g
r
a
d
in
g
o
r
an
al
y
tics
.
I
n
p
r
ac
tice,
th
ese
g
ap
s
m
ea
n
th
at
cu
r
r
en
t
s
y
s
te
m
s
eit
h
er
r
eq
u
ir
e
ex
ten
s
iv
e
m
an
u
al
in
ter
v
en
tio
n
o
r
p
r
o
d
u
ce
o
u
tp
u
ts
u
n
s
u
itab
le
f
o
r
in
telli
g
en
t
ev
al
u
atio
n
,
li
m
itin
g
t
h
eir
ap
p
licab
ilit
y
in
ac
ad
em
ic
as
s
ess
m
e
n
t
w
o
r
k
f
lo
w
s
.
T
h
is
s
tu
d
y
id
en
t
if
ies
t
h
e
n
ee
d
f
o
r
a
n
i
n
teg
r
ated
f
r
a
m
e
w
o
r
k
t
h
at
co
m
b
in
e
s
i
m
ag
e
p
r
ep
r
o
ce
s
s
in
g
,
h
an
d
w
r
iti
n
g
r
ec
o
g
n
itio
n
,
a
n
d
l
ar
g
e
lan
g
u
ag
e
m
o
d
el
-
b
ased
(
L
L
Ms
)
p
o
s
t
-
p
r
o
ce
s
s
i
n
g
to
b
r
id
g
e
th
ese
l
i
m
i
tatio
n
s
an
d
p
r
o
v
id
e
ac
cu
r
ate,
s
tr
u
ct
u
r
ed
,
an
d
co
n
tex
t
-
a
w
ar
e
tex
t e
x
t
r
ac
tio
n
(
T
E
)
.
Ob
j
ec
tiv
e
o
f
th
e
s
tu
d
y
:
t
h
e
p
r
i
m
ar
y
o
b
j
ec
tiv
e
o
f
th
is
s
t
u
d
y
i
s
to
d
ev
elo
p
a
r
eliab
le,
s
c
alab
le,
an
d
co
n
tex
t
-
a
w
ar
e
f
r
a
m
e
w
o
r
k
f
o
r
au
to
m
at
i
c
TE
f
r
o
m
f
u
l
l
y
HE
AS
.
Sp
ec
if
icall
y
,
t
h
e
s
y
s
te
m
is
d
esig
n
ed
to
:
−
P
r
ep
r
o
ce
s
s
s
ca
n
n
ed
d
o
cu
m
e
n
t
s
to
r
ed
u
ce
n
o
is
e,
co
r
r
ec
t sk
e
w,
an
d
n
o
r
m
alize
i
m
a
g
e
q
u
alit
y
.
−
R
ec
o
g
n
ize
d
iv
er
s
e
h
an
d
w
r
it
i
n
g
s
t
y
le
s
,
in
clu
d
i
n
g
cu
r
s
i
v
e,
p
r
in
ted
,
an
d
m
i
x
ed
f
o
r
m
s
,
w
h
ile
ac
cu
r
atel
y
d
etec
tin
g
c
h
ar
ac
ter
s
,
w
o
r
d
s
,
an
d
n
o
n
-
te
x
t e
le
m
en
ts
.
−
I
n
teg
r
ate
p
o
s
t
-
p
r
o
ce
s
s
i
n
g
tec
h
n
iq
u
es
u
s
in
g
L
L
M
s
to
co
r
r
ec
t
m
i
s
r
ec
o
g
n
ized
w
o
r
d
s
,
e
n
h
an
ce
co
n
te
x
t
u
al
ac
cu
r
ac
y
,
an
d
s
tr
u
ct
u
r
e
o
u
tp
u
t
tex
t in
m
ea
n
i
n
g
f
u
l f
o
r
m
ats.
−
E
n
ab
le
d
o
w
n
s
tr
ea
m
ap
p
licati
o
n
s
s
u
c
h
as
au
to
m
ated
g
r
ad
in
g
,
d
ig
i
tal
ar
ch
i
v
i
n
g
,
p
lag
iar
i
s
m
d
etec
tio
n
,
an
d
ed
u
ca
tio
n
al
a
n
al
y
tics
,
f
ac
ilit
ati
n
g
m
o
r
e
ef
f
icie
n
t a
n
d
f
air
ev
al
u
atio
n
o
f
s
t
u
d
en
t p
er
f
o
r
m
a
n
ce
.
T
h
r
o
u
g
h
t
h
ese
o
b
j
ec
tiv
es,
t
h
e
s
tu
d
y
a
i
m
s
to
b
r
id
g
e
th
e
g
ap
b
et
w
ee
n
u
n
s
tr
u
ctu
r
ed
h
a
n
d
w
r
itte
n
i
n
p
u
ts
a
n
d
m
ac
h
in
e
-
r
ea
d
ab
le,
co
n
tex
t
u
al
l
y
co
h
er
en
t
o
u
tp
u
ts
,
p
r
o
v
id
i
n
g
a
p
r
ac
tical,
A
I
-
d
r
iv
en
s
o
lu
tio
n
f
o
r
m
o
d
er
n
ed
u
ca
tio
n
al
as
s
es
s
m
e
n
t
s
y
s
te
m
s
.
C
o
n
tr
ib
u
t
io
n
o
f
t
h
e
r
esear
ch
:
−
A
n
o
v
e
l
en
h
a
n
ce
d
L
L
MW
h
i
s
p
er
er
(
E
L
L
MW
)
f
r
a
m
e
w
o
r
k
co
m
b
in
i
n
g
ad
v
a
n
ce
d
p
r
ep
r
o
ce
s
s
i
n
g
,
C
NN
–
L
ST
M
–
b
ased
h
an
d
w
r
iti
n
g
r
ec
o
g
n
itio
n
,
an
d
L
L
M
-
b
ased
p
o
s
t
-
co
r
r
ec
tio
n
.
−
D
e
m
o
n
s
tr
ated
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
o
v
er
co
n
v
e
n
tio
n
a
l
O
C
R
to
o
ls
,
ac
h
ie
v
in
g
9
7
.
8
%
ac
c
u
r
ac
y
a
n
d
1
.
0
4
%
ch
ar
ac
ter
er
r
o
r
r
ate
(
C
E
R
)
o
n
f
u
ll
y
h
a
n
d
w
r
it
ten
a
n
s
w
er
s
cr
ip
ts
.
E
n
ab
les
d
o
w
n
s
tr
ea
m
ap
p
li
ca
t
io
n
s
s
u
c
h
as
a
u
to
m
ated
g
r
ad
in
g
,
p
la
g
iar
is
m
d
etec
tio
n
,
an
d
d
ig
ital
ar
ch
i
v
i
n
g
,
b
r
id
g
in
g
t
h
e
g
ap
b
et
w
ee
n
h
a
n
d
w
r
itte
n
in
p
u
ts
a
n
d
d
ig
ital o
u
t
p
u
ts
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
1
5
,
No
.
1
,
Ma
r
c
h
202
6
:
1
9
4
-
203
196
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
Usi
n
g
s
o
p
h
is
ticated
p
r
ep
r
o
ce
s
s
i
n
g
tec
h
n
iq
u
e
s
li
k
e
b
in
ar
izatio
n
as
w
ell
d
es
k
e
w
i
n
g
,
an
d
n
o
is
e
r
ed
u
ctio
n
,
P
atien
ce
et
a
l
.
[
9
]
d
escr
ib
es
to
tack
les
t
y
p
ical
p
r
o
b
lem
s
i
n
clu
d
i
n
g
lo
w
-
r
eso
lu
tio
n
i
m
ag
e
s
illu
s
tr
atio
n
s
,
n
o
is
y
i
m
a
g
es
,
a
n
d
d
if
f
er
e
n
t
T
R
.
P
er
f
o
r
m
an
c
e
ev
al
u
atio
n
s
s
h
o
w
n
o
tab
le
g
ain
s
in
co
m
p
u
ti
n
g
s
p
ee
d
an
d
ac
cu
r
ac
y
f
o
r
T
R
.
Fu
r
th
er
m
o
r
e,
a
co
m
p
ar
is
o
n
w
it
h
alter
n
ati
v
e
OC
R
s
y
s
te
m
s
i
s
o
f
f
er
ed
to
em
p
h
a
s
ize
th
e
b
en
e
f
it
s
o
f
th
e
e
m
p
lo
y
ed
m
et
h
o
d
.
T
ess
er
ac
t'
s
ab
ilit
ies c
an
b
e
g
r
ea
tl
y
e
n
h
a
n
ce
d
v
ia
s
u
i
tab
le
in
te
g
r
atio
n
a
n
d
p
r
ep
r
o
ce
s
s
in
g
,
ac
co
r
d
in
g
to
th
e
r
esear
ch
,
m
ak
in
g
it
a
p
o
ten
t
to
o
l
f
o
r
a
v
ar
iet
y
o
f
T
R
ap
p
licatio
n
s
.
C
r
o
s
illa
et
a
l
.
[
1
0
]
d
escr
ib
es
t
o
d
eliv
er
an
o
v
er
v
ie
w
o
f
th
e
cu
r
r
en
t
ca
p
ab
ilit
ies
o
f
m
u
lti
m
o
d
al
lar
g
e
lan
g
u
ag
e
m
o
d
el
s
(
M
L
L
Ms)
f
o
r
h
a
n
d
w
r
itten
T
R
(
HT
R
)
,
a
s
s
e
s
s
i
n
g
th
eir
p
o
ten
tial
w
h
en
co
m
p
ar
ed
to
tr
ad
itio
n
al
tas
k
-
s
p
ec
if
ic,
s
u
p
er
v
is
ed
m
o
d
els.
T
h
e
r
esu
lts
s
h
o
w
th
at
L
L
Ms
cu
r
r
en
tl
y
s
h
o
w
a
s
tr
o
n
g
p
er
f
o
r
m
an
ce
o
n
E
n
g
lis
h
tex
ts
,
y
et
t
h
e
y
d
e
m
o
n
s
tr
ate
a
w
ea
k
er
p
er
f
o
r
m
a
n
ce
o
n
la
n
g
u
ag
e
s
o
th
er
t
h
a
n
E
n
g
li
s
h
,
a
n
d
d
o
n
o
t
p
o
s
s
ess
a
s
ig
n
i
f
ica
n
t
ca
p
ab
ilit
y
f
o
r
s
elf
-
co
r
r
ec
tio
n
.
Ka
m
p
elo
p
o
u
lo
s
et
a
l
.
[
1
1
]
attem
p
ts
to
f
i
ll
t
h
is
v
a
cu
u
m
b
y
o
f
f
er
i
n
g
a
th
o
r
o
u
g
h
an
a
l
y
s
is
o
f
th
e
c
u
r
r
en
t
ap
p
licatio
n
s
a
n
d
u
s
e
ex
a
m
p
les
o
f
L
L
Ms
i
n
t
h
e
ar
ch
itect
u
r
e,
en
g
i
n
ee
r
in
g
,
an
d
co
n
s
tr
u
ct
io
n
(
A
E
C
)
s
ec
to
r
th
at
h
av
e
alr
ea
d
y
b
ee
n
estab
lis
h
ed
.
I
n
ad
d
itio
n
,
it
w
a
s
f
ea
s
i
b
le
to
class
if
y
t
h
e
m
,
id
en
ti
f
y
n
e
w
is
s
u
e
s
an
d
p
o
ten
tial
p
ath
s
f
o
r
th
e
f
ield
,
a
n
d
p
r
o
v
id
e
p
r
ac
tical
s
u
g
g
est
io
n
s
f
o
r
i
n
d
u
s
tr
y
s
tak
e
h
o
ld
er
s
b
y
ex
a
m
i
n
i
n
g
th
e
m
ai
n
ad
v
an
tag
e
s
an
d
d
is
ad
v
an
tag
e
s
o
f
th
e
s
e
ap
p
licatio
n
s
as
w
ell
as
b
y
tak
i
n
g
in
to
ac
co
u
n
t a
p
p
r
o
p
r
iate
s
tu
d
ies o
n
th
e
to
p
ic.
L
i
et
a
l
.
[
1
2
]
p
r
o
v
id
e
a
s
i
m
p
le
v
is
io
n
tr
an
s
f
o
r
m
er
(
ViT
-
b
a
s
ed
)
m
o
d
el
s
o
lel
y
u
til
izin
g
th
e
en
co
d
er
p
o
r
tio
n
o
f
th
e
tr
ad
itio
n
al
tr
an
s
d
u
ce
r
f
o
r
HT
R
.
I
n
o
r
d
e
r
to
a
ch
iev
e
g
o
o
d
p
er
f
o
r
m
an
ce
o
n
th
is
tas
k
w
it
h
litt
le
ch
an
g
es
to
th
e
u
s
u
al
ViT
ar
ch
itectu
r
e,
th
i
s
r
esear
ch
s
u
g
g
es
ts
a
n
o
v
el
ViT
lik
e
m
o
d
el.
A
cc
o
r
d
in
g
to
th
e
in
itia
l
r
esu
lt
s
,
ViT
ca
n
p
r
o
d
u
ce
g
o
o
d
r
esu
lts
,
e
s
p
ec
iall
y
o
n
t
h
e
la
r
g
est
d
atase
t,
th
e
L
AM
d
atas
et
[
1
3
]
,
w
h
ic
h
h
a
s
19,
8
3
0
tr
ain
in
g
s
a
m
p
le
s
.
A
o
n
e
-
s
h
o
t
h
an
d
w
r
it
ten
te
x
t
s
y
n
t
h
esi
s
f
r
a
m
e
w
o
r
k
ca
lled
W
r
iteViT
is
p
r
esen
ted
b
y
Na
m
et
a
l
.
[
1
4
]
.
I
t
in
te
g
r
ates
th
e
ViT
f
a
m
il
y
o
f
m
o
d
els,
w
h
ic
h
h
a
s
d
e
m
o
n
s
tr
ated
o
u
ts
tan
d
in
g
r
es
u
lt
s
i
n
a
v
ar
iet
y
o
f
co
m
p
u
ter
v
i
s
io
n
ta
s
k
s
.
A
li
g
h
t
w
ei
g
h
t
V
iT
-
b
ased
r
ec
o
g
n
izer
,
a
m
u
lti
-
s
ca
le
p
r
o
d
u
ce
r
co
n
s
tr
u
cted
u
s
i
n
g
t
r
an
s
f
o
r
m
er
e
n
co
d
er
-
d
ec
o
d
er
m
o
d
u
le
s
a
u
g
m
e
n
ted
b
y
co
n
d
itio
n
al
p
o
s
itio
n
al
e
n
co
d
in
g
(
C
P
E
)
,
an
d
a
ViT
-
b
ased
w
r
iter
id
en
t
if
ier
f
o
r
id
en
tify
i
n
g
s
t
y
le
i
n
co
r
p
o
r
atio
n
ar
e
all
in
clu
d
ed
in
to
W
r
iteVi
T
.
T
h
is
p
ap
er
s
h
o
w
s
th
at
ViT
-
b
ased
s
y
n
t
h
esi
s
ar
ch
itectu
r
e
s
h
a
v
e
g
r
ea
t
p
o
ten
tial
to
en
h
a
n
ce
th
e
p
r
o
d
u
ctio
n
o
f
h
an
d
w
r
itte
n
te
x
t,
esp
ec
iall
y
i
n
m
u
ltid
i
s
cip
lin
ar
y
o
r
lo
w
-
r
es
o
u
r
ce
s
ettin
g
s
.
A
li
g
h
t
w
ei
g
h
t
ar
ch
itectu
r
e
ca
lled
v
is
io
n
an
d
s
p
atial
l
y
-
a
w
ar
e
te
x
t
an
al
y
s
i
s
OC
R
(
VI
ST
A
-
OC
R
)
,
w
h
ic
h
co
m
b
i
n
es
te
x
t
d
etec
t
io
n
an
d
T
R
in
to
a
s
in
g
le
d
y
n
a
m
ic
m
o
d
el,
is
p
r
e
s
en
ted
b
y
Ha
m
d
i
et
a
l
.
[
1
5
]
.
I
n
co
n
tr
ast
to
tr
ad
itio
n
al
ap
p
r
o
ac
h
es
th
at
n
ee
d
d
is
tin
ct
b
r
an
c
h
es
w
i
th
s
p
ec
if
ic
T
R
an
d
d
etec
tio
n
p
ar
am
et
er
s
,
th
is
m
et
h
o
d
u
s
e
s
a
tr
an
s
f
o
r
m
er
d
ec
o
d
er
to
p
r
o
d
u
ce
tex
t tr
an
s
cr
ip
ts
a
n
d
th
eir
s
p
atial
co
o
r
d
in
ates in
a
s
i
n
g
le
b
r
an
ch
w
it
h
s
eq
u
e
n
tial
m
a
n
n
er
.
T
h
e
o
u
tco
m
e
s
d
em
o
n
s
tr
ate
t
h
at
t
h
e
m
o
d
el
s
u
cc
e
s
s
f
u
ll
y
d
ev
e
lo
p
s
g
eo
m
e
tr
ic
d
escr
ip
tio
n
s
o
f
s
p
atial
to
k
en
s
,
allo
w
in
g
f
o
r
ac
cu
r
ate
an
d
f
i
n
e
-
g
r
ain
ed
d
o
cu
m
e
n
t
O
C
R
.
T
h
ese
r
esu
lt
s
in
c
lu
d
e
b
o
th
s
tatis
tical
m
etr
ics
a
n
d
q
u
alitati
v
e
er
r
o
r
ev
alu
a
tio
n
s
.
T
h
e
ef
f
icie
n
t
an
d
ac
cu
r
ate
s
ce
n
e
tex
t
d
etec
to
r
(
E
A
ST
)
tech
n
iq
u
e
is
i
m
p
le
m
e
n
ted
an
d
ev
al
u
ated
b
y
So
n
i
et
a
l
.
[
1
6
]
f
o
r
tex
t
id
en
ti
f
icatio
n
an
d
d
etec
tio
n
in
n
atu
r
al
s
ce
n
e
p
h
o
to
s
.
C
o
m
p
ar
i
n
g
t
h
e
ef
f
ec
t
iv
e
n
es
s
o
f
th
r
ee
w
e
ll
-
k
n
o
w
n
O
C
R
m
e
th
o
d
s
is
th
e
g
o
al
o
f
th
e
s
t
u
d
y
.
B
o
u
n
d
in
g
b
o
x
es
e
m
p
h
a
s
izi
n
g
th
e
id
en
ti
f
ied
tex
t
p
o
r
tio
n
s
w
er
e
u
s
ed
to
v
is
u
all
y
d
is
p
la
y
th
e
o
u
tco
m
es
o
f
ap
p
ly
i
n
g
th
e
E
A
ST
m
o
d
el
to
a
s
et
o
f
test
i
m
ag
e
s
a
m
p
les.
T
h
e
alg
o
r
it
h
m
's
e
f
f
i
cien
c
y
w
a
s
d
e
m
o
n
s
tr
ated
b
y
r
ec
o
r
d
in
g
th
e
r
es
u
lti
n
g
ti
m
i
n
g
s
f
o
r
ea
ch
i
m
a
g
e,
w
h
ic
h
s
h
o
w
ed
a
v
er
ag
e
ti
m
i
n
g
r
an
g
e
b
et
w
ee
n
0
.
4
3
9
to
0
.
4
4
6
s
ec
o
n
d
s
f
o
r
t
h
e
co
r
r
esp
o
n
d
in
g
test
i
m
ag
e
s
.
T
h
ese
f
i
n
d
in
g
s
s
h
o
w
t
h
at
t
h
e
E
AST
m
et
h
o
d
w
o
r
k
s
i
n
r
ea
l
-
t
i
m
e
an
d
is
ac
c
u
r
ate,
w
h
ic
h
m
a
k
e
s
it
ap
p
r
o
p
r
iate
f
o
r
s
itu
a
tio
n
s
t
h
at
n
ee
d
in
s
tan
t T
R
.
Sh
y
lesh
et
a
l
[
1
7
]
c
r
e
ate
d
a
m
o
d
el
s
p
ec
if
i
ca
l
ly
f
o
r
ass
ess
i
n
g
4
0
-
w
o
r
d
r
es
p
o
n
s
es
.
B
y
v
ar
y
in
g
th
e
p
a
r
am
ete
r
s
,
d
e
ep
l
ay
er
s
,
n
u
m
b
e
r
o
f
n
eu
r
o
n
s
,
ac
t
iv
ati
o
n
f
u
n
cti
o
n
,
an
d
b
i
d
ir
ec
t
io
n
a
l
l
o
n
g
s
h
o
r
t
-
t
er
m
m
e
m
o
r
y
(
B
iL
S
T
M
)
lay
er
s
,
th
e
m
o
d
el
is
c
o
n
s
t
r
u
cte
d
u
s
in
g
a
r
an
g
e
o
f
p
o
ten
ti
al
w
ay
s
.
T
o
f
in
d
th
e
lig
h
t
est
an
d
b
est
m
o
d
el,
th
is
p
a
p
e
r
m
ad
e
n
u
m
er
o
u
s
a
d
j
u
s
tm
en
ts
to
ea
ch
p
a
r
a
m
eter
an
d
ch
an
g
e
d
th
e
n
u
m
b
er
o
f
lay
er
s
,
L
S
T
Ms
,
o
r
n
o
d
es
.
T
h
e
m
o
d
el'
s
p
e
r
f
o
r
m
an
ce
u
s
in
g
th
e
t
est s
et
,
ac
h
iev
i
n
g
an
ac
cu
r
ac
y
o
f
a
b
o
u
t
8
0
%.
T
h
e
ac
cu
r
a
cy
m
ig
h
t
b
e
r
a
is
e
d
b
y
im
p
r
o
v
in
g
th
e
t
r
a
in
in
g
d
at
a.
T
h
is
r
esu
lt
s
u
g
g
est
s
th
a
t
a
g
r
ea
te
r
d
eg
r
ee
o
f
an
a
l
y
s
is
an
d
r
ese
a
r
ch
is
n
ee
d
e
d
t
o
c
r
ea
te
a
m
o
d
el
f
o
r
l
en
g
th
ier
tex
ts
(
2
0
0
–
2
5
0
w
o
r
d
s
)
th
at
in
clu
d
e
im
ag
es
an
d
e
q
u
atio
n
s
.
I
n
o
r
d
er
to
in
cr
ea
s
e
th
e
m
o
d
el'
s
ac
cu
r
a
cy
an
d
ci
r
cu
m
v
en
t
th
e
tim
e
-
co
n
s
u
m
in
g
p
r
o
c
ess
o
f
t
r
ain
in
g
it
o
n
h
u
g
e
d
atas
ets
,
Pre
r
an
a
et
a
l
.
[
1
8
]
s
u
g
g
est
u
s
in
g
a
G
o
o
g
le
C
l
o
u
d
P
la
tf
o
r
m
(
G
C
P
)
OC
R
tex
t
ex
tr
ac
t
m
o
d
el
t
h
at
h
as
b
ee
n
t
r
ain
e
d
o
n
en
o
r
m
o
u
s
v
o
lu
m
es
o
f
p
u
b
li
cly
ac
c
ess
i
b
le
d
at
a.
I
n
t
er
m
s
o
f
m
ar
k
in
g
ac
cu
r
a
cy
,
th
e
s
u
g
g
ested
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
m
eth
o
d
f
a
r
e
d
b
et
te
r
th
an
an
y
o
f
th
e
a
lg
o
r
ith
m
s
.
T
h
e
d
is
tin
c
tiv
e
f
ea
tu
r
e
o
f
th
is
t
ec
h
n
i
q
u
e
is
th
at
,
b
ef
o
r
e
t
o
d
ir
ec
t
co
m
p
a
r
is
o
n
,
t
h
e
k
ey
w
o
r
d
s
ar
e
l
em
m
atize
d
,
c
ase
-
f
o
ld
ed
,
s
t
o
p
-
w
o
r
d
s
r
em
o
v
e
d
,
an
d
d
u
p
lic
ate
v
er
if
ie
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
E
LL
M
W
:
a
n
en
h
a
n
ce
d
visi
o
n
–
la
n
g
u
a
g
e
mo
d
el
f
o
r
r
elia
b
le
text
ex
tr
a
ctio
n
fr
o
m
…
(
Dh
ivya
V
en
ka
tesh
)
197
3.
M
E
T
H
O
D
3
.
1
.
L
L
M
Whis
pere
r
f
ra
m
e
w
o
rk
ba
s
ed
o
n ha
nd
w
rit
t
en
a
ns
w
er
s
cr
ipt
t
ex
t
ex
t
ra
ct
io
n
T
h
e
ef
f
ec
ti
v
e
n
es
s
o
f
O
C
R
i
n
HE
A
S
T
E
s
y
s
te
m
s
i
s
li
m
ited
b
y
v
ar
iatio
n
s
i
n
h
an
d
wr
itin
g
an
d
f
o
r
m
atti
n
g
p
r
esen
t
s
i
g
n
i
f
ican
t
ch
allen
g
es
f
o
r
OC
R
s
y
s
te
m
s
,
as
th
e
y
m
u
s
t
ac
co
m
m
o
d
ate
d
iv
er
s
e
s
t
y
les
an
d
in
co
n
s
is
te
n
cie
s
ac
r
o
s
s
d
if
f
er
e
n
t
an
s
w
er
s
h
ee
ts
.
A
d
d
itio
n
all
y
,
th
e
p
r
esen
ce
o
f
n
o
n
-
te
x
t
u
a
l
ar
tif
ac
ts
,
s
u
c
h
as
s
m
u
d
g
e
s
,
s
tai
n
s
,
o
r
s
tr
a
y
m
ar
k
s
,
ca
n
f
u
r
t
h
er
co
m
p
licate
th
e
d
ig
itizat
io
n
p
r
o
ce
s
s
,
p
o
ten
tiall
y
lead
in
g
to
in
ac
cu
r
ac
ie
s
in
T
R
.
R
ec
o
g
n
i
zin
g
co
m
p
le
x
c
h
ar
ac
ter
s
,
esp
ec
iall
y
t
h
o
s
e
f
o
u
n
d
in
d
ia
g
r
a
m
s
,
m
ath
e
m
at
ical
n
o
tatio
n
s
,
o
r
s
p
ec
ialized
s
y
m
b
o
ls
,
ad
d
s
an
o
th
er
la
y
er
o
f
d
if
f
ic
u
lt
y
,
r
eq
u
ir
i
n
g
ad
v
an
ce
d
f
ea
tu
r
e
ex
tr
ac
tio
n
tech
n
iq
u
es
,
a
n
d
r
o
b
u
s
t
alg
o
r
i
th
m
s
to
e
n
s
u
r
e
ac
c
u
r
ate
p
r
o
ce
s
s
i
n
g
[
1
9
]
-
[
2
1
]
.
T
o
o
v
er
co
m
e
th
ese
is
s
u
e
s
t
h
i
s
p
ap
er
p
r
o
p
o
s
es
a
s
o
lu
t
io
n
to
ass
is
t
p
r
o
f
ess
o
r
s
b
y
r
ec
o
g
n
ize
s
h
a
n
d
w
r
i
tten
i
m
a
g
e
te
x
t.
T
h
i
s
r
esear
ch
ad
o
p
ts
a
m
u
lti
-
p
h
ase
m
et
h
o
d
o
lo
g
y
t
h
at
co
m
b
in
es
i
m
a
g
e
p
r
ep
r
o
c
ess
i
n
g
,
s
eg
m
e
n
tatio
n
,
H
R
,
an
d
L
L
M
b
ased
en
h
a
n
ce
m
en
t
tech
n
iq
u
e
s
to
ac
cu
r
atel
y
e
x
tr
ac
t
tex
t
u
al
d
ata
f
r
o
m
s
ca
n
n
ed
d
o
cu
m
e
n
t
o
f
f
u
ll
y
HE
A
S
.
C
o
n
v
en
t
io
n
al
O
C
R
e
n
g
in
e
s
,
lik
e
T
ess
er
ac
t,
h
av
e
tr
o
u
b
le
with
co
m
p
licate
d
o
r
n
o
n
-
s
ta
n
d
ar
d
lay
o
u
ts
d
u
e
to
th
eir
r
el
y
o
n
s
tatic
m
o
d
els
an
d
p
r
eset
p
atter
n
s
.
T
h
e
s
u
g
g
e
s
ted
E
L
L
MW
h
i
s
p
er
er
f
r
a
m
e
w
o
r
k
e
m
p
lo
y
s
s
o
p
h
is
ticated
al
g
o
r
ith
m
s
to
ad
j
u
s
t
d
y
n
a
m
icall
y
,
m
ai
n
tai
n
in
g
d
o
cu
m
e
n
t
co
n
te
x
t
w
h
ile
g
u
ar
a
n
teei
n
g
tex
t
p
ar
s
in
g
ac
c
u
r
ac
y
.
T
h
e
s
u
g
g
e
s
t
ed
p
ar
ad
ig
m
is
a
s
o
p
h
is
t
icate
d
tex
t
p
ar
s
er
t
h
at
g
ets
r
ea
d
y
f
o
r
L
L
Ms
to
an
al
y
ze
co
m
p
lica
ted
d
o
cu
m
e
n
ts
la
ter
o
n
,
s
u
c
h
as
s
ca
n
n
ed
p
o
r
tab
le
d
o
cu
m
en
t
f
o
r
m
ats
(
P
DFs
)
,
p
h
o
to
s
,
an
d
tab
les.
Fig
u
r
e
1
d
is
p
la
y
s
t
h
e
o
v
er
all
b
lo
ck
d
iag
r
a
m
f
o
r
th
e
s
u
g
g
e
s
te
d
m
et
h
o
d
.
Fig
u
r
e
1
.
Ov
er
all
b
lo
ck
d
iag
r
a
m
f
o
r
p
r
o
p
o
s
ed
m
o
d
el
T
h
e
ap
p
licatio
n
o
f
E
L
L
MW
is
h
p
er
er
f
o
r
th
e
HT
R
tas
k
i
n
v
o
lv
es
a
d
if
f
er
en
t
w
o
r
k
f
lo
w
t
h
a
n
t
h
e
o
n
e
b
ased
o
n
p
r
e
-
p
r
o
ce
s
s
i
n
g
,
s
eg
m
en
tatio
n
,
a
n
d
T
R
.
W
h
ile
p
r
e
p
r
o
ce
s
s
in
g
m
a
y
b
e
r
eq
u
ir
ed
i
n
s
o
m
e
ca
s
es
w
h
er
e
m
o
d
e
l
s
r
eq
u
ir
e
i
m
a
g
e
r
esizi
n
g
o
r
co
r
r
ec
tio
n
s
to
o
r
ien
tatio
n
,
s
e
g
m
en
ta
tio
n
i
s
n
o
lo
n
g
er
n
ec
es
s
ar
y
,
n
o
r
ar
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
1
5
,
No
.
1
,
Ma
r
c
h
202
6
:
1
9
4
-
203
198
s
p
ec
if
ic
g
r
o
u
n
d
tr
u
th
(
GT
)
an
n
o
tatio
n
s
r
elate
d
to
b
o
th
tex
t
an
d
la
y
o
u
t.
T
h
is
s
tu
d
y
p
r
i
m
ar
il
y
f
o
cu
s
es
o
n
T
E
f
r
o
m
s
ca
n
n
ed
d
o
cu
m
e
n
t
o
f
H
E
A
S
i
n
v
e
s
ti
g
ates
th
e
p
o
ten
tia
l
o
f
i
n
co
r
p
o
r
atin
g
E
L
L
MW
a
s
a
p
o
s
t
-
co
r
r
ec
tio
n
s
tep
in
th
e
w
o
r
k
f
lo
w
,
to
ass
e
s
s
w
h
et
h
er
L
L
M
s
co
u
ld
d
etec
t
an
d
co
r
r
ec
t
er
r
o
r
s
in
th
eir
in
i
t
ial
p
r
ed
ictio
n
s
an
d
p
o
ten
tiall
y
i
m
p
r
o
v
e
ac
cu
r
ac
y
.
Fin
all
y
,
m
o
d
els
’
p
r
ed
ictio
n
s
ar
e
ev
alu
ated
u
s
i
n
g
tr
ad
itio
n
al
m
etr
ics
s
u
ch
a
s
C
E
R
,
w
o
r
d
er
r
o
r
r
ate
(
W
E
R
)
,
an
d
b
ag
o
f
w
o
r
d
s
-
W
E
R
B
o
W
-
W
E
R
,
g
u
ar
an
teein
g
a
co
m
p
ar
ab
le
ev
alu
atio
n
w
it
h
tr
ad
itio
n
a
l O
C
R
m
o
d
els.
3
.
2
.
Da
t
a
s
et
des
cr
iptio
n
T
h
is
d
ataset
is
cu
r
ated
to
s
u
s
ten
an
ce
th
e
g
r
o
w
t
h
a
n
d
ass
e
s
s
m
e
n
t
o
f
t
h
e
en
h
a
n
ce
d
L
L
M
W
h
is
p
e
r
er
f
r
a
m
e
w
o
r
k
f
o
r
au
to
m
at
ic
T
E
f
r
o
m
f
u
l
l
y
HE
A
S
.
I
t
ai
m
s
to
ad
d
r
ess
th
e
ch
alle
n
g
es
o
f
u
n
s
tr
u
ctu
r
ed
,
d
iv
er
s
e,
an
d
n
atu
r
al
h
an
d
w
r
iti
n
g
s
t
y
les
ac
r
o
s
s
v
ar
io
u
s
s
tu
d
e
n
t
p
o
p
u
latio
n
s
an
d
ac
ad
e
m
ic
s
u
b
j
ec
ts
.
T
h
e
d
ataset
co
m
p
r
is
e
s
s
ca
n
n
ed
i
m
ag
e
s
o
f
r
ea
l
o
r
s
i
m
u
lated
h
a
n
d
w
r
i
tten
ex
a
m
i
n
atio
n
an
s
w
er
s
h
ee
ts
co
llect
ed
f
r
o
m
ac
ad
e
m
ic
in
s
t
itu
tio
n
s
o
r
g
en
er
ated
u
s
i
n
g
s
y
n
t
h
etic
h
a
n
d
w
r
iti
n
g
to
o
ls
to
p
r
o
tect
p
r
iv
ac
y
.
T
h
er
ef
o
r
e,
th
e
i
m
ag
e
s
v
ar
y
i
n
q
u
alit
y
an
d
r
eso
l
u
tio
n
,
o
f
te
n
as
a
r
es
u
lt
o
f
s
tu
d
e
n
t
d
i
f
f
er
en
ce
s
i
n
h
a
n
d
w
r
i
tin
g
s
t
y
le
.
As
s
u
c
h
,
th
e
i
m
ag
e
d
atasets
ar
e
s
u
s
ce
p
tib
le
to
n
o
is
e,
b
lu
r
r
y
r
eg
is
tr
atio
n
,
o
r
d
is
to
r
tio
n
s
.
T
h
e
s
ca
n
n
ed
i
m
ag
e
s
ar
e
p
r
e
-
p
r
o
ce
s
s
ed
f
o
r
q
u
alit
y
i
m
p
r
o
v
e
m
en
ts
,
s
u
ch
a
s
n
o
is
e
r
ed
u
ctio
n
an
d
ali
g
n
m
e
n
t.
I
t
in
cl
u
d
es
s
a
m
p
les
f
r
o
m
s
tu
d
en
t
s
o
f
v
ar
y
i
n
g
ag
es,
g
en
d
er
s
,
a
n
d
w
r
iti
n
g
s
t
y
les
(
c
u
r
s
i
v
e,
p
r
in
t,
a
n
d
m
i
x
ed
)
.
T
h
e
p
ag
e
la
y
o
u
t
s
co
n
t
ain
s
i
n
g
le
-
co
lu
m
n
,
m
u
lti
-
p
ar
ag
r
ap
h
r
esp
o
n
s
es,
f
i
g
u
r
e
s
,
tab
les,
an
d
m
ar
g
i
n
n
o
tes.
T
h
e
n
o
is
e
f
ac
to
r
s
in
cl
u
d
e
s
i
n
g
le
-
co
l
u
m
n
,
m
u
lti
-
p
ar
ag
r
ap
h
r
esp
o
n
s
e
s
,
f
i
g
u
r
es,
tab
les,
an
d
m
ar
g
i
n
n
o
te
s
.
T
h
i
s
d
ataset
is
i
m
p
o
r
tan
t
f
o
r
t
r
ain
in
g
t
h
e
m
o
d
el
f
o
r
au
to
m
at
ic
s
t
u
d
en
t r
esp
o
n
s
e
r
ec
o
g
n
itio
n
a
n
d
ass
e
s
s
m
e
n
t.
3
.
3
.
T
ex
t
det
ec
t
io
n a
nd
s
eg
m
e
nta
t
io
n
T
h
e
in
itial
s
tag
e
in
t
h
e
tex
t
d
etec
tio
n
p
r
o
ce
d
u
r
e
is
to
lo
ca
te
th
e
tex
t
ar
ea
o
n
th
e
p
ag
e.
T
h
is
in
v
o
l
v
e
s
lo
o
k
in
g
f
o
r
g
r
o
u
p
s
o
f
p
ix
el
s
i
n
a
h
an
d
w
r
itte
n
an
s
w
er
s
cr
ip
t
i
m
ag
e
co
m
p
r
is
in
g
ele
m
en
t
s
,
ch
ar
ac
ter
s
;
ea
ch
o
f
th
ese
ele
m
e
n
ts
h
as
a
clas
s
ass
i
g
n
ed
to
it.
I
t
is
n
ec
es
s
ar
y
to
u
t
ilize
an
y
t
h
r
es
h
o
ld
in
g
tec
h
n
iq
u
e
to
en
ab
le
f
u
r
th
e
r
in
v
e
s
ti
g
atio
n
.
A
d
ap
tiv
e
t
h
r
esh
o
ld
i
n
g
o
f
te
n
w
o
r
k
s
b
est
w
h
e
n
th
e
p
r
o
p
er
s
ettin
g
s
ar
e
u
s
e
d
.
T
h
e
f
o
llo
w
in
g
is
h
o
w
th
e
s
e
g
m
e
n
tatio
n
p
r
o
ce
s
s
is
ex
ec
u
ted
in
(
1
)
.
(
)
=
(
)
+
ℎ
ℎ
+
(
)
(
1
)
Fo
r
=
0
,
1
….
N
-
1
.
I
n
(
2
)
an
d
(
3
)
co
m
p
u
te
a
f
ea
tu
r
e
f
o
r
th
e
a
d
ap
tiv
e
p
ix
el
s
et
o
f
a
n
i
m
m
ed
iate
la
y
er
.
=
ƞ
P(a
)
+
μ
(
a,
b
)
f
o
r
a
=
1
,
2
………
n
(
2
)
=
ƞ
P(b
)
–
2
(
x
)
+
μ
(
a,
b
)
f
o
r
a
=
3
,
4
,
5
………
m
(
3
)
I
n
o
r
d
er
to
ar
r
an
g
e
th
e
la
y
e
r
s
in
a
s
eq
u
e
n
tial
f
as
h
io
n
f
o
r
ac
cu
r
ate
an
d
u
n
iq
u
e
T
R
̞
is
th
e
la
y
er
s
ig
n
i
f
ica
n
ce
t
h
at
is
tak
e
n
i
n
to
co
n
s
id
er
atio
n
.
T
h
e
o
u
tco
m
e
is
esti
m
ated
b
y
t
h
e
co
n
s
tr
u
cted
s
in
g
le
la
y
er
as
t
h
e
s
u
m
o
f
all
t
h
e
p
ix
el
s
t
h
at
co
m
p
r
is
e
a
s
et,
w
it
h
a
f
ix
ed
to
tal.
A
n
e
w
la
y
er
is
cr
ea
ted
u
s
i
n
g
(
4
)
.
=
∑
(
)
ℎ
(
)
+
∑
(
)
+μ
(
a,
b
)
(
4
)
3
.
4
.
H
a
nd
w
rit
t
en
t
ex
t
re
co
g
nitio
n
I
n
th
i
s
s
tep
HT
R
is
th
e
p
r
o
ce
s
s
o
f
co
n
v
er
tin
g
h
a
n
d
w
r
itt
en
an
s
w
er
s
cr
ip
t
in
i
m
a
g
es
o
r
s
ca
n
n
ed
d
o
cu
m
en
ts
i
n
to
m
ac
h
in
e
-
r
ea
d
ab
le
tex
t.
I
n
t
h
is
ca
s
e,
m
an
y
h
y
p
er
p
ar
a
m
eter
s
ar
e
ac
ti
v
e.
T
h
ese
eith
er
h
a
v
e
d
ef
au
lt
v
alu
e
s
s
p
ec
if
ied
o
r
h
a
v
e
b
ee
n
g
en
er
ated
u
s
in
g
t
h
e
tr
ain
i
n
g
d
ata.
T
h
e
h
an
d
w
r
it
ten
s
a
m
p
le
h
a
v
e
r
esized
all
o
f
th
e
w
o
r
d
s
eg
m
e
n
ted
i
m
a
g
e
d
ata
to
3
2
×
1
2
8
p
ix
els
in
o
r
d
er
to
f
ee
d
o
u
r
d
ataset
in
to
th
e
C
NN
la
y
er
.
Af
ter
th
at,
th
e
s
e
w
o
r
d
s
ar
e
f
ed
in
to
a
C
NN
l
a
y
er
,
w
h
ic
h
h
as
6
4
n
o
d
es
an
d
a
k
er
n
el
s
ize
o
f
(
3
,
3
)
.
A
p
o
o
lin
g
la
y
er
w
it
h
a
k
er
n
el
s
ize
o
f
(
2
,
2
)
r
e
ce
iv
ed
th
e
o
u
tp
u
t
f
r
o
m
th
e
f
i
r
s
t
C
NN
la
y
er
,
r
ed
u
cin
g
its
f
o
r
m
to
1
6
×6
4
.
T
w
o
C
NN
la
y
er
s
,
ea
ch
h
a
v
i
n
g
2
5
6
n
o
d
es,
r
ec
eiv
e
th
e
o
u
tp
u
t
f
r
o
m
th
e
p
r
ec
ed
in
g
p
o
o
lin
g
la
y
e
r
.
A
p
o
o
lin
g
la
y
e
r
co
m
e
s
a
f
ter
ea
ch
o
f
t
h
ese
C
N
N
la
y
er
s
.
Fo
llo
w
i
n
g
t
h
ese
t
w
o
C
NN
a
n
d
p
o
o
lin
g
la
y
er
s
,
t
h
e
o
u
tp
u
t d
ata
'
s
f
o
r
m
is
s
h
r
u
n
k
to
4
×
3
2
.
On
ce
m
o
r
e,
tw
o
C
N
N
la
y
er
s
w
i
th
5
1
2
n
o
d
es
g
et
t
h
is
d
ata,
an
d
th
e
n
t
w
o
b
atch
n
o
r
m
aliza
tio
n
la
y
er
s
f
o
llo
w
.
A
C
N
N
la
y
er
r
ec
eiv
es
th
e
o
u
tp
u
t
o
f
th
e
las
t
b
atch
n
o
r
m
aliza
tio
n
la
y
er
af
ter
it
h
as
ac
ce
p
ted
th
r
o
u
g
h
o
n
e
p
o
o
lin
g
la
y
er
.
I
n
o
r
d
er
to
id
en
tify
a
w
o
r
d
an
d
u
lti
m
atel
y
co
n
v
er
t
it
in
to
a
tex
t
f
i
le,
th
is
r
esear
c
h
w
o
r
k
e
m
p
lo
y
ed
an
L
ST
M
la
y
e
r
w
it
h
2
5
6
lay
er
s
i
n
t
h
e
f
i
n
a
l l
a
y
er
.
Fo
r
ac
cu
r
ate
T
E
,
th
e
h
id
d
en
lay
er
s
ar
e
tak
en
i
n
to
co
n
s
id
er
ati
o
n
s
in
ce
NN
i
s
u
til
ized
.
In
(
5
)
d
escr
ib
es
th
e
in
p
u
t o
f
ea
c
h
h
id
d
en
la
y
er
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
E
LL
M
W
:
a
n
en
h
a
n
ce
d
visi
o
n
–
la
n
g
u
a
g
e
mo
d
el
f
o
r
r
elia
b
le
text
ex
tr
a
ctio
n
fr
o
m
…
(
Dh
ivya
V
en
ka
tesh
)
199
(
(
,
)
)
=
,
+
∑
∗
(
5
)
w
h
er
e
W
an
d
F
s
tan
d
f
o
r
th
e
in
p
u
t
an
d
h
id
d
en
la
y
er
w
ei
g
h
ts
a
n
d
th
e
b
ias
v
al
u
e
o
f
th
e
h
id
d
en
la
y
er
,
r
esp
ec
tiv
el
y
.
T
h
er
ef
o
r
e,
(
6
)
is
u
s
ed
to
co
m
p
u
te
t
h
e
o
u
tp
u
t o
f
ea
ch
h
id
d
en
la
y
er
.
O
(
I
(
a,
b
)
)
=
∑
(
,
+
,
)
=
1
∑
(
,
+
,
)
=
1
(
6
)
3
.
5
.
P
o
s
t
-
pro
ce
s
s
ing
T
h
en
p
o
s
t
-
p
r
o
ce
s
s
i
n
g
e
n
h
an
ce
s
an
d
co
r
r
ec
ts
t
h
e
o
u
tp
u
t
g
e
n
e
r
ated
b
y
t
h
e
HT
R
s
y
s
te
m
f
r
o
m
O
C
R
.
I
t
in
v
o
l
v
es
s
ev
er
al
s
tep
s
:
er
r
o
r
co
r
r
ec
tio
n
:
u
s
e
co
n
te
x
t
-
b
ased
a
lg
o
r
ith
m
s
to
co
r
r
ec
t
co
m
m
o
n
HT
R
m
is
ta
k
es
(
e.
g
.
,
co
n
f
u
s
i
n
g
'
O
'
w
it
h
'
0
'
o
r
'
I
'
w
it
h
'
1
'
)
.
3.
6
.
T
ex
t
ex
t
ra
ct
io
n us
ing
pro
po
s
ed
E
L
L
M
Wis
hp
er
er
I
n
th
is
s
tep
,
th
is
s
t
u
d
y
e
x
p
lo
r
es
to
in
cr
ea
s
e
th
e
m
o
d
el's
a
b
ilit
y
to
ex
tr
ac
t
r
a
w
te
x
t
f
r
o
m
s
ca
n
n
ed
h
an
d
w
r
itte
n
an
s
w
er
s
cr
ip
t
ac
cu
r
atel
y
,
a
p
r
o
p
o
s
ed
E
L
L
MW
an
al
y
s
i
s
s
tep
is
tak
e
n
af
te
r
p
r
o
ce
s
s
in
g
th
r
o
u
g
h
f
ir
s
t
f
o
u
r
s
t
ep
s
d
escr
ib
ed
ab
o
v
e.
A
co
m
p
r
eh
e
n
s
iv
e
te
x
t
p
ar
s
er
th
at
u
s
e
s
th
e
s
u
g
g
e
s
te
d
E
L
L
MW
h
is
p
er
er
s
tr
u
ct
u
r
e
cr
ea
tes
co
m
p
lex
d
o
cu
m
e
n
t
s
,
s
u
c
h
as
s
ca
n
n
ed
P
DF
s
,
p
ictu
r
es,
a
n
d
tab
les,
f
o
r
L
L
Ms
to
p
r
o
ce
s
s
later
.
I
ts
m
ai
n
g
o
al
i
s
to
tr
an
s
f
o
r
m
s
e
m
i
-
s
tr
u
ct
u
r
ed
o
r
u
n
s
tr
u
ctu
r
ed
d
ata
in
to
s
tr
u
ct
u
r
ed
,
u
s
e
f
u
l
f
o
r
m
ats
li
k
e
J
av
aScr
ip
t
o
b
j
ec
t
n
o
tatio
n
(
J
SON
)
.
E
L
L
MW
ex
ce
ls
at
m
a
n
ag
in
g
h
a
n
d
w
r
i
tten
tex
t,
n
o
is
y
s
ca
n
s
,
an
d
co
m
p
le
x
la
y
o
u
t
s
,
allo
w
in
g
b
u
s
i
n
e
s
s
es
to
o
p
ti
m
ize
d
o
cu
m
e
n
t
p
ar
s
i
n
g
p
r
o
ce
s
s
es.
R
a
th
er
,
it
ap
p
lies
clev
er
p
ar
s
in
g
tech
n
iq
u
es to
i
m
p
r
o
v
e
OC
R
r
esu
lt
s
,
m
a
k
i
n
g
t
h
e
d
ata
ea
s
ier
f
o
r
L
L
M
s
an
d
o
th
er
s
y
s
te
m
s
to
u
n
d
er
s
ta
n
d
.
T
h
e
r
ef
in
e
m
e
n
t
ta
s
k
is
ca
r
r
ied
o
u
t
i
n
t
h
r
ee
s
tep
s
w
h
er
e
t
h
e
m
o
d
el
is
a
s
k
ed
ag
ai
n
to
a
n
al
y
ze
b
o
th
t
h
e
o
r
ig
in
al
i
m
ag
e
an
d
th
e
o
u
tp
u
t
p
r
o
d
u
ce
d
at
ti
m
e
t
–
1
,
w
h
ich
m
ea
n
s
th
a
t
in
th
e
f
ir
s
t
i
ter
atio
n
th
e
m
o
d
el
ex
a
m
in
e
s
t
h
e
o
u
tp
u
t
o
f
t
h
e
z
er
o
-
s
h
o
t
ta
s
k
.
I
n
ea
ch
s
u
b
s
e
q
u
en
t
s
tep
a
d
if
f
er
en
t
u
s
er
p
r
o
m
p
t
i
s
p
r
o
v
id
ed
,
s
p
ec
if
y
in
g
w
h
er
e
th
e
f
o
cu
s
o
f
th
e
m
o
d
el
is
n
ee
d
ed
,
w
h
e
th
e
r
in
th
e
o
r
th
o
g
r
ap
h
y
o
r
s
p
ellin
g
r
ef
i
n
e
m
e
n
t
o
r
in
th
e
l
a
y
o
u
t
f
o
r
m
atti
n
g
o
r
in
b
o
th
as
t
h
e
last
p
r
o
m
p
t
.
Her
e
th
e
ex
tr
ac
ted
r
a
w
tex
t
=T
(
X,
Y)
=(
1
,
2
,
…
,
)
is
p
ass
ed
to
l
a
n
g
u
a
g
e
m
o
d
el
o
f
p
r
o
p
o
s
ed
E
L
L
MW
f
o
r
c
o
r
r
ec
tio
n
,
co
n
tex
t
a
w
ar
e
r
estri
ctin
g
o
r
d
en
o
is
in
g
i
s
m
e
n
tio
n
ed
in
(
7
)
-
(
1
0
)
.
=
(
)
(
7
)
=
(
T
(
X,
Y)
(
8
)
=
((
1
,
2
,
…
,
)
)
(
9
)
Ou
tp
u
t =
–
(
1
∗
,
2
∗
,
…
…
,
∗
)
(1
0
)
On
e
o
f
th
e
k
e
y
ad
v
a
n
ta
g
es
o
f
E
L
L
MW
is
t
h
eir
ab
ilit
y
to
h
an
d
le
b
o
th
s
tr
u
ct
u
r
ed
an
d
u
n
s
tr
u
ctu
r
ed
ev
alu
a
tio
n
tas
k
s
.
B
y
i
n
co
r
p
o
r
atin
g
w
ell
-
d
esi
g
n
ed
p
r
o
m
p
ts
th
at
in
cl
u
d
e
q
u
esti
o
n
s
,
r
u
b
r
ics,
an
d
ex
a
m
p
l
e
an
s
w
er
s
,
E
L
L
MW
ca
n
p
er
f
o
r
m
n
u
a
n
ce
d
ev
al
u
atio
n
s
an
d
p
r
o
v
id
e
j
u
s
tif
icatio
n
s
f
o
r
th
eir
d
ec
is
io
n
s
.
T
h
is
tr
an
s
p
ar
en
c
y
e
n
h
a
n
ce
s
tr
u
s
t i
n
th
e
a
u
to
m
ated
g
r
ad
i
n
g
p
r
o
ce
s
s
.
T
h
is
m
o
d
el
a
ls
o
v
er
s
atile,
c
ap
ab
le
o
f
m
a
n
ag
in
g
o
p
en
en
d
ed
q
u
esti
o
n
s
an
d
s
u
b
j
ec
tiv
e
r
esp
o
n
s
es
ef
f
ec
ti
v
el
y
.
T
h
eir
s
ca
lab
ilit
y
an
d
ad
ap
tab
ilit
y
m
ak
e
t
h
e
m
s
u
itab
le
f
o
r
a
w
id
e
r
an
g
e
o
f
ed
u
ca
tio
n
al
co
n
tex
ts
.
4.
E
XP
E
R
I
M
E
NT
A
L
SE
T
UP
Usi
n
g
a
d
ataset
o
f
f
u
ll
HE
AS
s
a
m
p
les,
t
h
is
s
tu
d
y
ev
al
u
ates
t
h
e
p
r
o
p
o
s
ed
E
L
L
MW
m
o
d
el
o
n
s
tan
d
ar
d
OC
R
tas
k
s
,
s
p
ec
i
f
ic
all
y
tex
t
r
ec
o
g
n
itio
n
(
T
R
)
.
T
h
e
o
b
j
ec
tiv
e
is
to
d
em
o
n
s
tr
at
e
th
at
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
o
u
tp
er
f
o
r
m
s
s
tate
-
of
-
t
h
e
-
ar
t
O
C
R
m
o
d
e
ls
i
n
ter
m
s
o
f
p
er
f
o
r
m
an
ce
w
h
ile
al
s
o
o
f
f
er
i
n
g
g
r
ea
ter
f
le
x
ib
ilit
y
a
n
d
g
e
n
er
aliza
tio
n
ca
p
ab
ilit
y
.
A
l
l
ex
p
er
i
m
e
n
t
s
w
er
e
co
n
d
u
cted
o
n
a
d
esk
to
p
PC
eq
u
ip
p
ed
w
it
h
an
I
n
tel
i
3
p
r
o
ce
s
s
o
r
,
8
GB
o
f
R
A
M,
5
1
2
GB
o
f
s
to
r
ag
e,
an
d
a
h
ar
d
d
is
k
d
r
iv
e
(
HDD)
.
T
o
ev
alu
ate
th
e
ac
c
u
r
ac
y
an
d
q
u
alit
y
o
f
th
e
te
x
t
r
ec
o
g
n
i
ze
d
b
y
th
e
p
r
o
p
o
s
ed
E
L
L
MW
m
o
d
el
f
o
r
HE
AS,
s
ev
er
al
k
e
y
ev
alu
atio
n
m
etr
ics
an
d
cr
iter
ia
w
er
e
e
m
p
lo
y
ed
.
T
h
ese
m
etr
ics
ar
e
u
s
ed
to
ass
es
s
th
e
e
f
f
icac
y
o
f
s
u
g
g
ested
s
o
l
u
tio
n
s
b
y
d
eter
m
in
i
n
g
h
o
w
w
el
l
th
e
y
c
an
co
n
v
er
t
h
a
n
d
w
r
itten
o
r
s
c
an
n
ed
e
x
a
m
a
n
s
w
er
s
cr
ip
t
te
x
t
i
n
to
a
m
ac
h
i
n
e
-
r
ea
d
ab
le
ar
r
an
g
e
m
en
t.
So
m
e
c
o
m
m
o
n
esti
m
atio
n
cr
iter
ia
u
s
e
d
in
th
e
co
n
te
x
t o
f
HE
AS a
r
e
lis
ted
b
elo
w
.
a.
C
E
R
:
it
is
t
h
e
“
i
n
v
er
ted
ac
cu
r
ac
y
”,
m
ea
n
i
n
g
t
h
at
it
r
ep
r
ese
n
ts
t
h
e
er
r
o
r
r
ate
at
ch
ar
ac
ter
le
v
el
b
ased
o
n
th
e
L
e
v
en
s
h
te
in
d
i
s
ta
n
ce
.
T
h
is
s
u
g
g
e
s
ts
th
at
all
t
h
e
m
etr
ics
th
a
t
in
co
r
p
o
r
ate
th
is
a
s
p
ec
t
d
ep
en
d
o
n
th
e
r
ea
d
i
n
g
o
r
d
er
.
T
o
ca
lcu
late
th
e
C
E
R
,
th
e
f
o
llo
w
i
n
g
f
o
r
m
u
la
h
as b
ee
n
u
s
ed
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
1
5
,
No
.
1
,
Ma
r
c
h
202
6
:
1
9
4
-
203
200
=
+
+
=
+
+
+
+
T
h
e
r
ef
er
en
ce
tex
t
m
u
s
t
b
e
tr
an
s
f
o
r
m
ed
in
to
th
e
g
i
v
e
n
GT
b
y
m
a
k
i
n
g
t
h
e
f
o
llo
w
i
n
g
ch
an
g
es
:
I
r
ep
r
esen
t
th
e
n
u
m
b
er
o
f
i
n
s
e
r
tio
n
s
,
S
r
ep
r
esen
t
s
t
h
e
n
u
m
b
er
o
f
s
u
b
s
t
itu
tio
n
s
,
D
r
ep
r
esen
ts
t
h
e
n
u
m
b
er
o
f
d
eletio
n
s
,
a
n
d
N
is
th
e
to
tal
a
m
o
u
n
t
o
f
ch
ar
ac
ter
s
in
th
e
GT
.
T
o
g
iv
e
an
id
ea
o
f
w
h
at
th
ese
p
er
ce
n
tag
e
s
ac
tu
all
y
r
ep
r
esen
t,
a
C
E
R
b
el
o
w
5
%
is
co
n
s
id
er
ed
v
er
y
g
o
o
d
,
if
it
f
alls
i
n
a
r
an
g
e
b
et
w
ee
n
5
to
1
0
%
is
g
o
o
d
,
ex
ce
lle
n
ce
is
ac
h
iev
ed
w
it
h
a
C
E
R
b
elo
w
2
.
5
%.
b.
W
E
R
:
i
t
is
th
e
w
o
r
d
-
le
v
el
eq
u
iv
ale
n
t
o
f
C
E
R
,
ca
lcu
la
ti
n
g
t
h
e
n
u
m
b
er
o
f
ad
d
itio
n
s
,
d
eletio
n
s
,
an
d
s
u
b
s
t
itu
t
io
n
s
i
n
t
h
e
r
ec
o
g
n
i
s
ed
w
o
r
d
s
as a
p
er
ce
n
tag
e
o
f
all
t
h
e
w
o
r
d
s
i
n
GT
.
=
+
+
Si
m
i
lar
l
y
,
t
h
e
n
u
m
b
er
o
f
ter
m
s
u
b
s
ti
tu
t
io
n
s
(
)
,
in
s
er
tio
n
s
(
)
,
an
d
d
eletio
n
s
(
)
r
eq
u
ir
ed
t
o
ch
an
g
e
o
n
e
s
tr
i
n
g
i
n
to
an
o
th
er
is
ad
d
ed
u
p
,
an
d
th
e
to
tal
n
u
m
b
er
o
f
g
r
o
u
n
d
-
tr
u
t
h
ter
m
s
(
)
is
u
s
ed
to
d
eter
m
in
e
t
h
e
.
B
o
W
-
W
E
R
it
is
b
ased
o
n
th
e
r
ea
d
in
g
o
r
d
er
,
m
ea
n
i
n
g
t
h
at,
i
f
a
w
o
r
d
is
d
etec
ted
b
u
t
n
o
t
in
t
h
e
co
r
r
ec
t
p
o
s
itio
n
i
n
g
,
it
w
ill
n
o
t
b
e
co
n
s
id
er
ed
.
T
o
ad
d
r
ess
th
ese
is
s
u
e
s
a
n
d
co
n
s
id
er
th
e
co
r
r
ec
tl
y
d
etec
ted
w
o
r
d
s
r
eg
ar
d
less
o
f
th
eir
o
r
d
er
,
w
h
ich
is
p
ar
tic
u
lar
l
y
u
s
ef
u
l
f
o
r
in
f
o
r
m
a
tio
n
r
etr
iev
al
tas
k
s
,
th
e
W
E
R
-
B
o
W
m
e
tr
ic
h
as b
ee
n
in
tr
o
d
u
ce
d
.
−
=
−
(
∩
)
W
h
er
e,
N
is
th
e
to
tal
n
u
m
b
e
r
o
f
w
o
r
d
s
i
n
th
e
GT
an
d
P
th
e
n
u
m
b
er
o
f
w
o
r
d
s
in
t
h
e
p
r
ed
ictio
n
.
T
h
er
ef
o
r
e,
th
is
v
alu
e
ca
n
b
e
m
u
c
h
lo
w
er
th
a
n
W
E
R
w
h
en
th
e
r
ea
d
in
g
o
r
d
er
b
etw
ee
n
GT
an
d
p
r
ed
ictio
n
is
d
if
f
er
e
n
t.
T
h
e
T
E
r
esu
lts
f
r
o
m
ea
c
h
s
ca
n
n
ed
HE
A
S
s
a
m
p
l
e
d
em
o
n
s
tr
ate
th
a
t
th
e
p
r
o
p
o
s
ed
E
L
L
MW
m
o
d
el
s
ig
n
i
f
ica
n
tl
y
o
u
tp
er
f
o
r
m
s
co
n
v
en
t
io
n
al
O
C
R
A
P
I
s
,
ac
h
ie
v
i
n
g
an
a
v
er
ag
e
ac
c
u
r
ac
y
o
f
9
7
.
8
%
.
Fig
u
r
e
2
illu
s
tr
ates
t
h
e
C
E
R
,
d
em
o
n
s
tr
ati
n
g
th
a
t
th
e
p
r
o
p
o
s
ed
E
L
L
MW
m
o
d
el
co
n
s
is
te
n
tl
y
ac
h
ie
v
e
s
th
e
lo
w
e
s
t
c
h
ar
ac
ter
-
le
v
el
er
r
o
r
s
a
m
o
n
g
al
l
e
v
alu
a
ted
OC
R
f
r
a
m
e
w
o
r
k
s
,
in
d
icati
n
g
it
s
s
u
p
er
io
r
ca
p
ab
ilit
y
i
n
ac
cu
r
atel
y
id
en
ti
f
y
in
g
i
n
d
iv
id
u
al
ch
ar
ac
ter
s
in
h
a
n
d
w
r
i
tten
s
cr
ip
ts
.
Fi
g
u
r
e
3
p
r
esen
t
s
t
h
e
W
E
R
co
m
p
ar
is
o
n
,
co
n
f
ir
m
i
n
g
t
h
at
E
L
L
MW
ef
f
e
ctiv
el
y
r
ec
o
g
n
ize
s
co
m
p
lete
w
o
r
d
s
,
w
i
th
m
in
i
m
a
l
s
u
b
s
t
itu
tio
n
s
,
in
s
er
tio
n
s
,
o
r
d
eletio
n
s
,
t
h
er
eb
y
en
h
a
n
ci
n
g
o
v
er
all
w
o
r
d
-
lev
e
l
ac
cu
r
ac
y
.
Fig
u
r
e
4
s
h
o
w
s
t
h
e
W
E
R
-
B
OW
co
m
p
ar
is
o
n
,
h
ig
h
li
g
h
ti
n
g
t
h
e
m
o
d
el
’
s
s
tr
en
g
th
i
n
p
r
eser
v
i
n
g
co
n
te
x
tu
a
l
i
n
teg
r
i
t
y
a
n
d
ca
p
tu
r
in
g
w
o
r
d
-
le
v
el
s
e
m
a
n
tics
m
o
r
e
p
r
ec
is
el
y
th
a
n
tr
ad
iti
o
n
al
O
C
R
m
eth
o
d
s
.
Fig
u
r
e
5
p
r
o
v
id
es
a
co
m
p
r
eh
en
s
i
v
e
co
m
p
ar
is
o
n
o
f
k
e
y
m
etr
ics
—
p
r
ec
is
io
n
,
r
ec
all,
F
1
-
s
co
r
e,
an
d
ac
cu
r
ac
y
—
ac
r
o
s
s
all
m
et
h
o
d
s
.
T
h
e
p
r
o
p
o
s
ed
E
L
L
MW
m
o
d
el
d
em
o
n
s
tr
ates
t
h
e
h
i
g
h
est
p
er
f
o
r
m
a
n
ce
ac
r
o
s
s
all
m
etr
ics,
w
ith
p
r
ec
is
io
n
o
f
9
7
.
2
%,
r
ec
all
o
f
9
4
.
7
%,
F
1
-
s
c
o
r
e
o
f
9
7
.
4
%,
an
d
o
v
er
all
ac
c
u
r
ac
y
o
f
9
7
.
8
%.
B
y
co
n
tr
ast,
tr
ad
itio
n
al
O
C
R
A
P
I
s
s
u
c
h
as
E
as
y
OC
R
,
T
ess
er
ac
t,
GC
V,
P
ad
d
leOC
R
,
A
B
B
YY
Fi
n
eRead
er
,
an
d
T
r
an
s
y
m
O
C
R
ac
h
ie
v
ed
lo
w
er
p
er
f
o
r
m
a
n
ce
s
r
a
n
g
i
n
g
b
et
w
ee
n
9
2
–
9
6
% a
cr
o
s
s
th
ese
m
etr
ic
s
.
Fig
u
r
e
2
.
C
E
R
ac
r
o
s
s
v
ar
io
u
s
m
et
h
o
d
s
Fig
u
r
e
3
.
W
E
R
ac
r
o
s
s
v
ar
io
u
s
m
et
h
o
d
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
E
LL
M
W
:
a
n
en
h
a
n
ce
d
visi
o
n
–
la
n
g
u
a
g
e
mo
d
el
f
o
r
r
elia
b
le
text
ex
tr
a
ctio
n
fr
o
m
…
(
Dh
ivya
V
en
ka
tesh
)
201
Fig
u
r
e
4
.
W
E
R
B
OW
ac
r
o
s
s
v
ar
io
u
s
m
et
h
o
d
s
Fig
u
r
e
5
.
C
o
m
p
ar
is
o
n
o
f
m
etr
i
cs
ac
r
o
s
s
v
ar
io
u
s
m
e
th
o
d
s
5.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
p
r
ese
n
ted
a
co
m
p
r
e
h
e
n
s
i
v
e
ev
al
u
atio
n
o
f
th
e
p
r
o
p
o
s
ed
E
L
L
MW
f
r
a
m
e
w
o
r
k
f
o
r
au
to
m
ated
TE
f
r
o
m
HE
A
S
.
T
h
e
ex
p
er
i
m
e
n
t
s
d
e
m
o
n
s
tr
ated
th
at
E
L
L
MW
co
n
s
is
te
n
tl
y
o
u
t
p
er
f
o
r
m
s
tr
ad
itio
n
al
OC
R
m
et
h
o
d
s
,
in
cl
u
d
in
g
T
r
an
s
y
m
OC
R
,
GC
V,
P
ad
d
leOCR
,
E
as
y
O
C
R
,
T
ess
er
ac
t,
an
d
A
B
B
YY
Fin
eRead
er
,
ac
r
o
s
s
m
u
ltip
le
ev
al
u
atio
n
m
etr
ics.
Sp
ec
if
icall
y
,
th
e
m
o
d
el
ac
h
iev
ed
a
C
E
R
o
f
1
.
0
4
%
,
W
E
R
o
f
3
.
2
4
%,
W
E
R
-
B
OW
o
f
2
.
45
%,
an
d
a
n
o
v
er
all
ac
c
u
r
ac
y
o
f
9
7
.
8
%,
h
i
g
h
l
ig
h
ti
n
g
its
s
u
p
er
io
r
ca
p
ab
ilit
y
i
n
ac
c
u
r
atel
y
r
ec
o
g
n
izi
n
g
b
o
th
c
h
ar
ac
ter
s
an
d
w
o
r
d
s
.
T
h
e
p
r
o
p
o
s
ed
f
r
am
e
w
o
r
k
al
s
o
s
h
o
w
ed
h
i
g
h
p
r
ec
i
s
io
n
,
r
ec
all
,
an
d
F
1
-
s
co
r
e,
r
ef
lectin
g
its
r
o
b
u
s
t
n
es
s
in
p
r
eser
v
i
n
g
te
x
t
i
n
teg
r
it
y
an
d
co
n
tex
t
u
al
i
n
f
o
r
m
atio
n
.
T
h
e
r
esu
lts
in
d
icate
th
at
E
L
L
MW
is
h
i
g
h
l
y
e
f
f
ec
tiv
e
in
h
a
n
d
lin
g
d
iv
er
s
e
h
a
n
d
w
r
iti
n
g
s
t
y
le
s
,
tex
t
o
r
ien
tati
o
n
s
,
an
d
d
o
cu
m
e
n
t
la
y
o
u
t
s
,
m
ak
in
g
it
a
r
eliab
l
e
s
o
lu
tio
n
f
o
r
lar
g
e
-
s
ca
le
au
to
m
ated
ass
e
s
s
m
e
n
t
an
d
g
r
ad
in
g
s
y
s
te
m
s
.
B
y
s
ig
n
i
f
ica
n
tl
y
r
ed
u
ci
n
g
er
r
o
r
r
ates
an
d
im
p
r
o
v
in
g
ex
tr
a
ctio
n
ac
cu
r
ac
y
,
th
e
f
r
a
m
e
wo
r
k
ca
n
s
tr
ea
m
li
n
e
ev
alu
a
tio
n
w
o
r
k
f
lo
w
s
,
p
r
o
v
id
e
f
aster
f
ee
d
b
ac
k
to
s
t
u
d
en
t
s
,
an
d
r
ed
u
ce
f
ac
u
l
t
y
w
o
r
k
lo
ad
.
T
h
e
li
m
itatio
n
o
f
th
e
s
tu
d
y
i
s
t
h
at
t
h
e
m
o
d
el
m
a
y
e
x
h
ib
it
r
ed
u
ce
d
p
er
f
o
r
m
a
n
ce
o
n
ex
tr
e
m
el
y
n
o
is
y
,
h
e
av
il
y
s
m
u
d
g
ed
,
o
r
m
u
ltil
i
n
g
u
al
h
a
n
d
w
r
itte
n
s
cr
ip
ts
.
F
u
tu
r
e
w
o
r
k
co
u
ld
f
o
cu
s
o
n
f
u
r
t
h
er
en
h
a
n
ci
n
g
th
e
f
r
a
m
e
w
o
r
k
’
s
ad
ap
tab
ilit
y
to
m
u
ltil
i
n
g
u
al
s
cr
ip
ts
,
co
m
p
lex
m
a
th
e
m
atica
l
n
o
tatio
n
s
,
an
d
h
ea
v
il
y
d
eg
r
ad
ed
d
o
cu
m
en
ts
,
as
w
ell
a
s
in
te
g
r
atin
g
t
h
e
m
o
d
el
w
ith
r
ea
l
-
ti
m
e
ed
u
ca
tio
n
al
ass
e
s
s
m
en
t
p
latf
o
r
m
s
.
Ov
er
all,
t
h
e
E
L
L
MW
f
r
a
m
e
w
o
r
k
r
ep
r
esen
ts
a
s
i
g
n
if
ica
n
t
s
tep
to
w
ar
d
ac
cu
r
ate,
ef
f
icie
n
t,
a
n
d
s
ca
lab
le
T
E
f
r
o
m
h
a
n
d
w
r
i
tten
e
x
a
m
in
at
io
n
s
cr
ip
ts
.
F
UNDIN
G
I
NF
O
RM
AT
I
O
N
Au
t
h
o
r
s
s
tate
n
o
f
u
n
d
i
n
g
i
n
v
o
l
v
ed
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
1
5
,
No
.
1
,
Ma
r
c
h
202
6
:
1
9
4
-
203
202
AUTHO
R
CO
NT
RIB
UT
I
O
NS ST
A
T
E
M
E
NT
T
h
is
jo
u
r
n
al
u
s
e
s
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT
)
t
o
r
ec
o
g
n
ize
in
d
i
v
id
u
al
au
t
h
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
t
h
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
lla
b
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
Dh
i
v
y
a
Ve
n
k
ates
h
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
B
r
in
th
a
R
aj
ak
u
m
ar
i
Siv
ar
aj
✓
✓
✓
✓
✓
✓
✓
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
si
s
I
:
I
n
v
e
st
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
si
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
t
h
o
r
s
s
tate
n
o
co
n
f
lic
t o
f
i
n
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
Data
av
ailab
ilit
y
is
n
o
t
ap
p
licab
le
to
th
is
p
ap
er
as
n
o
n
ew
d
ata
w
er
e
cr
ea
ted
o
r
an
aly
ze
d
in
th
i
s
s
tu
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
M
d
.
A
.
R
a
h
a
m
a
n
a
n
d
H
.
M
a
h
m
u
d
,
“
A
u
t
o
mat
e
d
e
v
a
l
u
a
t
i
o
n
o
f
h
a
n
d
w
r
i
t
t
e
n
a
n
sw
e
r
scri
p
t
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
T
ra
n
s
a
c
t
i
o
n
s
o
n
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
a
n
d
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
1
0
,
n
o
.
4
,
p
p
.
1
-
1
6
,
A
u
g
.
2
0
2
2
,
d
o
i
:
1
0
.
1
4
7
3
8
/
t
ml
a
i
.
1
0
4
.
1
2
8
3
1
.
[
2
]
M
.
R
a
v
i
k
u
mar
,
S
.
S
.
S
a
m
p
a
t
h
K
u
ma
r
,
a
n
d
G
.
S
h
i
v
a
k
u
mar
,
“
Ev
a
l
u
a
t
i
o
n
o
f
h
a
n
d
w
r
i
t
t
e
n
a
n
sw
e
r
scri
p
t
s
u
si
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s,”
T
u
r
k
i
s
h
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
a
n
d
M
a
t
h
e
m
a
t
i
c
s
E
d
u
c
a
t
i
o
n
,
v
o
l
.
9
,
n
o
.
1
,
p
p
.
6
2
0
–
6
3
3
,
2
0
1
8
,
d
o
i
:
1
0
.
1
7
7
6
2
/
t
u
r
c
o
mat
.
v
9
i
1
.
1
4
0
4
1
.
[
3
]
R
.
S
m
i
t
h
,
“
A
n
O
v
e
r
v
i
e
w
o
f
t
h
e
Te
sse
r
a
c
t
O
C
R
E
n
g
i
n
e
,
”
i
n
N
i
n
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
D
o
c
u
m
e
n
t
An
a
l
y
si
s
a
n
d
Re
c
o
g
n
i
t
i
o
n
(
I
C
D
A
R
2
0
0
7
)
,
C
u
r
i
t
i
b
a
,
B
r
a
z
i
l
,
2
0
0
7
,
p
p
.
6
2
9
-
6
3
3
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
D
A
R
.
2
0
0
7
.
4
3
7
6
9
9
1
.
[
4
]
Jai
d
e
d
A
I
,
“
Ea
sy
O
C
R
:
R
e
a
d
y
-
to
-
u
se
O
C
R
w
i
t
h
d
e
e
p
l
e
a
r
n
i
n
g
,
”
G
i
t
H
u
b
r
e
p
o
si
t
o
ry
,
2
0
2
0
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
g
i
t
h
u
b
.
c
o
m/
J
a
i
d
e
d
A
I
/
Ea
sy
O
C
R
.
(
A
c
c
e
sse
d
:
N
o
v
.
2
1
,
2
0
2
5
)
.
[
5
]
J.
P
u
i
g
c
e
r
v
e
r
a
n
d
C
.
M
o
c
h
o
l
í
,
“
P
y
La
i
a
2
0
1
8
,
”
G
i
t
H
u
b
r
e
p
o
si
t
o
r
y
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
g
i
t
h
u
b
.
c
o
m
/
j
p
u
i
g
c
e
r
v
e
r
/
P
y
L
a
i
a
.
(
A
c
c
e
ss
e
d
:
N
o
v
.
2
1
,
2
0
2
5
)
.
[
6
]
M
.
K
i
šš,
K
.
B
e
n
e
š,
a
n
d
M
.
H
r
a
d
i
š,
“
A
T
-
S
T
:
S
e
l
f
-
t
r
a
i
n
i
n
g
a
d
a
p
t
a
t
i
o
n
st
r
a
t
e
g
y
f
o
r
O
C
R
i
n
d
o
ma
i
n
s
w
i
t
h
l
i
m
i
t
e
d
t
r
a
n
scr
i
p
t
i
o
n
s,”
i
n
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
D
o
c
u
m
e
n
t
A
n
a
l
y
s
i
s
a
n
d
Re
c
o
g
n
i
t
i
o
n
(
I
C
D
A
R
)
,
2
0
2
1
,
p
p
.
4
6
3
-
4
7
7
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
030
-
8
6
3
3
7
-
1
_
3
1
.
[
7
]
J.
K
o
h
ú
t
a
n
d
M
.
H
r
a
d
i
š,
“
T
S
-
N
e
t
:
O
C
R
t
r
a
i
n
e
d
t
o
sw
i
t
c
h
b
e
t
w
e
e
n
t
e
x
t
t
r
a
n
s
c
r
i
p
t
i
o
n
s
t
y
l
e
s,”
i
n
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
D
o
c
u
m
e
n
t
An
a
l
y
si
s
a
n
d
R
e
c
o
g
n
i
t
i
o
n
(
I
C
D
AR)
,
2
0
2
1
,
p
p
.
4
7
8
-
4
9
3
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
030
-
8
6
3
3
7
-
1
_
3
2
.
[
8
]
P
a
d
d
l
e
P
a
d
d
l
e
C
o
mm
u
n
i
t
y
,
“
P
a
d
d
l
e
O
C
R
:
A
n
o
p
e
n
-
so
u
r
c
e
O
C
R
t
o
o
l
b
a
se
d
o
n
P
a
d
d
l
e
P
a
d
d
l
e
,
”
G
i
t
H
u
b
r
e
p
o
si
t
o
r
y
,
2
0
2
1
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
g
i
t
h
u
b
.
c
o
m/
P
a
d
d
l
e
P
a
d
d
l
e
/
P
a
d
d
l
e
O
C
R
.
(
A
c
c
e
sse
d
:
N
o
v
.
2
1
,
2
0
2
5
)
.
[
9
]
O
.
O
.
P
a
t
i
e
n
c
e
,
E.
M
.
A
mae
c
h
i
,
O
.
G
e
o
r
g
e
,
a
n
d
O
.
N
.
I
saa
c
,
“
En
h
a
n
c
e
d
t
e
x
t
r
e
c
o
g
n
i
t
i
o
n
i
n
i
mag
e
s
u
si
n
g
T
e
sse
r
a
c
t
O
C
R
w
i
t
h
i
n
t
h
e
L
a
r
a
v
e
l
f
r
a
mew
o
r
k
,
”
As
i
a
n
J
o
u
r
n
a
l
o
f
Re
s
e
a
r
c
h
i
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
7
,
n
o
.
9
,
p
p
.
5
8
–
6
9
,
2
0
2
4
,
d
o
i
:
1
0
.
9
7
3
4
/
a
j
r
c
o
s/
2
0
2
4
/
v
1
7
i
9
4
9
9
.
[
1
0
]
G
.
C
r
o
si
l
l
a
,
L
.
K
l
i
c
,
a
n
d
G
.
C
o
l
a
v
i
z
z
a
,
“
B
e
n
c
h
mar
k
i
n
g
l
a
r
g
e
l
a
n
g
u
a
g
e
mo
d
e
l
s
f
o
r
h
a
n
d
w
r
i
t
t
e
n
t
e
x
t
r
e
c
o
g
n
i
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
D
o
c
u
m
e
n
t
a
t
i
o
n
,
v
o
l
.
8
1
,
n
o
.
7
,
p
p
.
3
3
4
-
3
5
4
,
2
0
2
5
,
d
o
i
:
1
0
.
1
1
0
8
/
JD
-
03
-
2025
-
0
0
8
2
.
[
1
1
]
D
.
K
a
mp
e
l
o
p
o
u
l
o
s,
A
.
T
san
o
u
sa
,
S
.
V
r
o
c
h
i
d
i
s,
a
n
d
I
.
K
o
mp
a
t
s
i
a
r
i
s
,
“
A
r
e
v
i
e
w
o
f
L
L
M
s
a
n
d
t
h
e
i
r
a
p
p
l
i
c
a
t
i
o
n
s
i
n
t
h
e
a
r
c
h
i
t
e
c
t
u
r
e
,
e
n
g
i
n
e
e
r
i
n
g
a
n
d
c
o
n
s
t
r
u
c
t
i
o
n
i
n
d
u
st
r
y
,
”
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
Re
v
i
e
w
,
v
o
l
.
5
8
,
p
.
2
5
0
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
4
6
2
-
0
2
5
-
1
1
2
4
1
-
7.
[
1
2
]
Y
.
L
i
,
D
.
C
h
e
n
,
T
.
T
a
n
g
,
a
n
d
X
.
S
h
e
n
,
“
H
T
R
-
V
T:
H
a
n
d
w
r
i
t
t
e
n
t
e
x
t
r
e
c
o
g
n
i
t
i
o
n
w
i
t
h
v
i
si
o
n
t
r
a
n
sf
o
r
me
r
,
”
Pa
t
t
e
rn
R
e
c
o
g
n
i
t
i
o
n
,
v
o
l
.
1
5
8
,
p
.
1
1
0
9
6
7
,
F
e
b
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
a
t
c
o
g
.
2
0
2
4
.
1
1
0
9
6
7
.
[
1
3
]
S
.
C
a
s
c
i
a
n
e
l
l
i
e
t
a
l
.,
“
T
h
e
L
A
M
D
a
t
a
se
t
:
A
N
o
v
e
l
B
e
n
c
h
mar
k
f
o
r
L
i
n
e
-
L
e
v
e
l
H
a
n
d
w
r
i
t
t
e
n
T
e
x
t
R
e
c
o
g
n
i
t
i
o
n
,
”
i
n
2
0
2
2
2
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
o
n
P
a
t
t
e
r
n
R
e
c
o
g
n
i
t
i
o
n
(
I
C
PR)
,
M
o
n
t
r
e
a
l
,
Q
C
,
C
a
n
a
d
a
,
2
0
2
2
,
p
p
.
1
5
0
6
-
1
5
1
3
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
P
R
5
6
3
6
1
.
2
0
2
2
.
9
9
5
6
1
8
9
.
[
1
4
]
D
.
H
.
N
a
m,
H
.
T
.
D
.
K
h
o
a
,
a
n
d
V
.
N
.
L
.
D
u
y
,
“
W
r
i
t
e
V
i
T
:
H
a
n
d
w
r
i
t
t
e
n
t
e
x
t
g
e
n
e
r
a
t
i
o
n
w
i
t
h
v
i
s
i
o
n
t
r
a
n
sf
o
r
me
r
,
”
a
rX
i
v
p
re
p
ri
n
t
,
M
a
y
2
0
2
5
,
d
o
i
:
1
0
.
4
8
5
5
0
/
a
r
X
i
v
.
2
5
0
5
.
1
3
2
3
5
.
[
1
5
]
L
.
H
a
md
i
,
A
.
T
a
mas
n
a
,
P
.
B
o
i
sso
n
,
a
n
d
T
.
P
a
q
u
e
t
,
“
V
I
S
TA
-
O
C
R
:
T
o
w
a
r
d
s
g
e
n
e
r
a
t
i
v
e
a
n
d
i
n
t
e
r
a
c
t
i
v
e
e
n
d
-
to
-
e
n
d
O
C
R
mo
d
e
l
s,”
a
rXi
v
p
re
p
r
i
n
t
,
2
0
2
5
,
d
o
i
:
1
0
.
4
8
5
5
0
/
a
r
X
i
v
.
2
5
0
4
.
0
3
6
2
1
.
[
1
6
]
V
.
K
.
S
o
n
i
,
V
.
S
h
u
k
l
a
,
S
.
R
.
T
a
n
d
a
n
,
A
.
P
i
mp
a
l
k
a
r
,
N
.
K
.
N
e
ma,
a
n
d
M
.
N
a
i
k
,
“
P
e
r
f
o
r
man
c
e
e
v
a
l
u
a
t
i
o
n
o
f
e
f
f
i
c
i
e
n
t
a
n
d
a
c
c
u
r
a
t
e
t
e
x
t
d
e
t
e
c
t
i
o
n
a
n
d
r
e
c
o
g
n
i
t
i
o
n
i
n
n
a
t
u
r
a
l
sce
n
e
i
mag
e
s
u
si
n
g
EA
S
T
a
n
d
O
C
R
f
u
si
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
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
(
I
J
AC
S
A)
,
v
o
l
.
1
6
,
n
o
.
1
,
2
0
2
5
,
d
o
i
:
1
0
.
1
4
5
6
9
/
I
J
A
C
S
A
.
2
0
2
5
.
0
1
6
0
1
4
4
.
[
1
7
]
A
.
S
h
y
l
e
sh
,
A
.
R
a
a
f
e
h
,
S
.
M
a
t
h
i
n
,
V
.
B
.
P
r
a
k
a
s
h
a
n
d
H
.
S
h
a
n
mu
g
a
su
n
d
a
r
a
m,
“
A
u
t
o
mat
e
d
A
n
sw
e
r
S
c
r
i
p
t
Ev
a
l
u
a
t
i
o
n
U
s
i
n
g
D
e
e
p
L
e
a
r
n
i
n
g
,
”
i
n
2
0
2
3
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
C
o
m
m
u
n
i
c
a
t
i
o
n
a
n
d
I
n
f
o
rm
a
t
i
c
s
(
I
C
C
C
I
)
,
C
o
i
mb
a
t
o
r
e
,
I
n
d
i
a
,
2
0
2
3
,
p
p
.
1
-
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
C
I
5
6
7
4
5
.
2
0
2
3
.
1
0
1
2
8
3
1
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
E
LL
M
W
:
a
n
en
h
a
n
ce
d
visi
o
n
–
la
n
g
u
a
g
e
mo
d
el
f
o
r
r
elia
b
le
text
ex
tr
a
ctio
n
fr
o
m
…
(
Dh
ivya
V
en
ka
tesh
)
203
[
1
8
]
M
.
S
.
P
r
e
r
a
n
a
,
S
.
M
.
C
h
a
v
a
n
,
R
.
B
a
t
h
u
l
a
,
S
.
S
a
i
k
u
mar,
a
n
d
G
.
D
a
y
a
l
a
n
,
“
Ev
a
l
:
A
u
t
o
mat
i
c
e
v
a
l
u
a
t
i
o
n
o
f
a
n
sw
e
r
scri
p
t
s u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
I
n
t
e
l
l
i
g
e
n
t
S
y
st
e
m
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
i
n
E
n
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
1
,
n
o
.
1
,
p
p
.
3
1
6
–
3
2
3
,
2
0
2
3
.
[
1
9
]
V
.
A
g
r
a
w
a
l
,
J.
Jag
t
a
p
,
a
n
d
M
.
P
.
K
a
n
t
i
p
u
d
i
,
“
A
n
O
v
e
r
v
i
e
w
o
f
H
a
n
d
-
D
r
a
w
n
D
i
a
g
r
a
m
R
e
c
o
g
n
i
t
i
o
n
M
e
t
h
o
d
s
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
1
9
7
3
9
-
1
9
7
5
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
4
.
3
3
5
7
3
9
8
.
[
2
0
]
Y
.
S
.
C
h
e
r
n
y
sh
o
v
a
,
A
.
V
.
S
h
e
sh
k
u
s
,
a
n
d
V
.
V
.
A
r
l
a
z
a
r
o
v
,
“
Tw
o
-
S
t
e
p
C
N
N
F
r
a
mew
o
r
k
f
o
r
T
e
x
t
L
i
n
e
R
e
c
o
g
n
i
t
i
o
n
i
n
C
a
me
r
a
-
C
a
p
t
u
r
e
d
I
mag
e
s,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
3
2
5
8
7
-
3
2
6
0
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
7
4
0
5
1
.
[
2
1
]
A
.
N
i
k
i
t
h
a
,
J
.
G
e
e
t
h
a
a
n
d
D
.
S
.
J
a
y
a
L
a
k
sh
mi
,
“
H
a
n
d
w
r
i
t
t
e
n
T
e
x
t
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
,
”
2
0
2
0
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
R
e
c
e
n
t
T
r
e
n
d
s
o
n
El
e
c
t
ro
n
i
c
s,
I
n
f
o
rm
a
t
i
o
n
,
C
o
m
m
u
n
i
c
a
t
i
o
n
&
T
e
c
h
n
o
l
o
g
y
(
RT
EI
C
T
)
,
B
a
n
g
a
l
o
r
e
,
I
n
d
i
a
,
2
0
2
0
,
p
p
.
3
8
8
-
3
9
2
,
d
o
i
:
1
0
.
1
1
0
9
/
R
T
EI
C
T
4
9
0
4
4
.
2
0
2
0
.
9
3
1
5
6
7
9
.
[
2
2
]
F
.
J.
A
y
r
i
n
e
t
a
l
.
,
“
En
h
a
n
c
i
n
g
O
C
R
p
o
st
-
p
r
o
c
e
ssi
n
g
t
h
r
o
u
g
h
v
i
si
o
n
-
l
a
n
g
u
a
g
e
mo
d
e
l
,
”
i
n
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
Em
e
rg
i
n
g
T
e
c
h
n
o
l
o
g
i
e
s
a
n
d
C
o
m
p
u
t
i
n
g
I
n
n
o
v
a
t
i
o
n
s
,
2
0
2
5
,
p
p
.
2
4
7
-
2
5
3
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
1
-
9
2
8
5
4
-
3
_
2
9
.
[
2
3
]
K
.
F
e
l
d
h
o
f
f
,
H
.
W
i
e
me
r
,
P
.
T
r
ä
g
e
r
,
R
.
K
ü
h
n
e
,
M
.
Z
i
mm
e
r
man
n
,
a
n
d
S
.
I
h
l
e
n
f
e
l
d
t
,
“
A
u
t
o
mat
i
c
i
n
f
o
r
ma
t
i
o
n
e
x
t
r
a
c
t
i
o
n
f
r
o
m
sci
e
n
t
i
f
i
c
p
u
b
l
i
c
a
t
i
o
n
s
b
a
se
d
o
n
t
h
e
u
se
c
a
se
o
f
a
d
d
i
t
i
v
e
ma
n
u
f
a
c
t
u
r
i
n
g
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
5
,
n
o
.
1
7
,
p
.
9
3
3
1
,
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
5
1
7
9
3
3
1
.
[
2
4
]
X
.
G
a
o
e
t
a
l
.
,
“
S
e
l
e
c
t
i
n
g
p
o
st
-
p
r
o
c
e
ssi
n
g
sc
h
e
me
s
f
o
r
a
c
c
u
r
a
t
e
d
e
t
e
c
t
i
o
n
o
f
smal
l
o
b
j
e
c
t
s
i
n
l
o
w
-
r
e
s
o
l
u
t
i
o
n
w
i
d
e
-
a
r
e
a
a
e
r
i
a
l
i
mag
e
r
y
,
”
Re
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
4
,
n
o
.
2
,
p
.
2
5
5
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
r
s
1
4
0
2
0
2
5
5
.
[
2
5
]
Y
.
Z
h
a
n
g
,
G
.
D
i
n
g
,
D
.
D
i
n
g
,
Z
.
M
a
,
a
n
d
Z
.
L
i
,
“
O
n
c
o
n
t
e
n
t
-
a
w
a
r
e
p
o
st
-
p
r
o
c
e
ssi
n
g
:
A
d
a
p
t
i
n
g
s
t
a
t
i
s
t
i
c
a
l
l
y
l
e
a
r
n
e
d
mo
d
e
l
s
t
o
d
y
n
a
mi
c
c
o
n
t
e
n
t
,
”
AC
M
T
ra
n
s
a
c
t
i
o
n
s
o
n
M
u
l
t
i
m
e
d
i
a
C
o
m
p
u
t
i
n
g
,
C
o
m
m
u
n
i
c
a
t
i
o
n
s
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
0
,
p
p
.
1
–
2
3
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
4
5
/
3
6
1
2
9
2
5
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
M
r
s.
Dh
iv
y
a
Ve
n
k
a
te
s
h
is
p
u
rsu
i
n
g
P
h
.
D.
i
n
th
e
De
p
a
rtm
e
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
Bh
a
ra
th
I
n
stit
u
te
o
f
Hig
h
e
r
Ed
u
c
a
ti
o
n
a
n
d
Re
se
a
rc
h
,
S
e
la
iy
u
r,
Ch
e
n
n
a
i.
S
h
e
h
a
s
g
o
t
7
y
e
a
rs
o
f
e
x
p
e
rien
c
e
in
tea
c
h
in
g
.
S
h
e
is
a
n
Ed
it
o
r
a
t
In
tern
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
f
o
r
th
e
IS
BN
P
r
o
c
e
e
d
in
g
s
.
S
h
e
h
a
s
b
e
e
n
a
wa
rd
e
d
t
h
e
Be
st
F
a
c
u
lt
y
a
w
a
rd
,
Yo
u
n
g
Re
se
a
rc
h
e
r
a
w
a
rd
,
Na
ll
a
siri
y
a
r
a
w
a
rd
,
Ka
lv
is
e
m
m
a
l
a
w
a
rd
a
n
d
S
a
m
u
g
a
A
rp
a
n
ip
a
lar
a
w
a
rd
.
S
h
e
h
a
s
p
u
b
li
s
h
e
d
a
n
d
p
re
se
n
ted
p
a
p
e
rs
a
t
i
n
tern
a
ti
o
n
a
l
a
n
d
n
a
ti
o
n
a
l
c
o
n
f
e
r
e
n
c
e
s.
S
h
e
c
o
m
p
lete
d
NP
T
EL
/S
WA
Y
A
M
c
o
u
rse
in
“
Dig
it
a
l
I
m
a
g
e
P
ro
c
e
ss
in
g
”
.
S
h
e
is
a
n
a
c
ti
v
e
p
a
rti
c
ip
a
n
t
in
v
a
rio
u
s
F
D
P
,
w
e
b
in
a
r,
a
n
d
w
o
rk
sh
o
p
s.
S
h
e
se
rv
e
d
a
s
a
re
so
u
rc
e
P
e
rso
n
f
o
r
v
a
rio
u
s
S
o
u
th
e
rn
Un
iv
e
rsiti
e
s.
S
h
e
h
a
s
o
rg
a
n
ize
d
i
n
tern
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
,
w
o
rk
sh
o
p
s,
a
n
d
se
m
in
a
rs
in
c
o
ll
e
g
e
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
d
h
iv
y
a
m
a
sc
@
g
m
a
il
.
c
o
m
.
Dr
.
B
r
in
th
a
Ra
j
a
k
u
m
a
r
i
S
iv
a
r
a
j
is an
A
ss
o
c
iate
P
ro
f
e
ss
o
r
a
t
B
h
a
ra
th
In
stit
u
te o
f
Hig
h
e
r
Ed
u
c
a
ti
o
n
a
n
d
Re
se
a
rc
h
,
Ch
e
n
n
a
i,
in
t
h
e
De
p
a
rtm
e
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
.
S
h
e
h
a
s
o
v
e
r
2
4
y
e
a
rs
o
f
tea
c
h
in
g
e
x
p
e
rie
n
c
e
a
n
d
a
ro
b
u
st
a
c
a
d
e
m
ic
b
a
c
k
g
ro
u
n
d
i
n
c
o
m
p
u
ter
sc
ien
c
e
,
w
it
h
a
sp
e
c
ializ
a
ti
o
n
in
a
p
p
ly
in
g
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
iq
u
e
s
to
h
e
a
lt
h
c
a
re
a
n
d
d
a
ta
q
u
a
li
ty
re
se
a
rc
h
.
S
h
e
h
a
s
a
u
th
o
re
d
k
e
y
p
u
b
li
c
a
ti
o
n
s
a
d
d
re
ss
in
g
A
lzh
e
i
m
e
r
’s
d
iag
n
o
sis,
o
c
u
lar
d
ise
a
se
d
e
tec
ti
o
n
,
d
a
ta
m
in
in
g
,
a
n
d
c
lo
u
d
-
b
a
se
d
c
o
m
p
u
ti
n
g
se
c
u
rit
y
.
S
h
e
is
a
n
a
c
ti
v
e
m
e
m
b
e
r
o
f
p
ro
f
e
ss
io
n
a
l
b
o
d
ies
su
c
h
a
s
IS
T
E
a
n
d
IA
EN
G
,
a
n
d
h
a
s
p
u
b
li
sh
e
d
m
o
re
th
a
n
4
5
re
se
a
rc
h
p
a
p
e
rs
in
S
c
o
p
u
s
i
n
d
e
x
e
d
,
W
e
b
o
f
S
c
ien
c
e
,
a
n
d
UG
C
C
A
RE
li
ste
d
jo
u
rn
a
ls.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
rin
t
h
a
.
ra
m
e
sh
@
g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.