T
E
L
K
O
M
N
I
K
A
T
elec
o
m
m
un
ica
t
io
n Co
m
pu
t
ing
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
20
,
No
.
1
,
Feb
r
u
ar
y
20
22
,
p
p
.
1
09
~1
17
I
SS
N:
1
6
9
3
-
6
9
3
0
,
DOI
: 1
0
.
1
2
9
2
8
/TE
L
KOM
NI
KA.
v
20
i
1
.
2
2
4
5
7
109
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//telko
mn
ika
.
u
a
d
.
a
c.
i
d
Dev
elo
pment
s
ma
rt
e
y
eg
la
ss
es for
v
isua
lly
i
mpa
ired
p
eo
ple
ba
sed o
n yo
u only
loo
k once
H
a
s
s
a
n Sa
la
m
Abdu
l
-
Am
ee
r
,
H
a
s
s
a
n J
a
leel
H
a
s
s
a
n
,
Sa
lm
a
H
a
m
ee
di Abdu
lla
h
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
E
n
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
si
t
y
o
f
Te
c
h
n
o
l
o
g
y
,
B
a
g
h
d
a
d
,
I
r
a
q
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ap
r
1
0
,
2
0
2
1
R
ev
is
ed
Dec
2
2
,
2
0
2
1
Acc
ep
ted
Dec
3
0
,
2
0
2
1
Visu
a
ll
y
im
p
a
ired
p
e
o
p
le
a
re
fa
c
in
g
m
a
n
y
p
ro
b
lem
s
in
th
e
ir
li
f
e
.
On
e
o
f
th
e
se
p
r
o
b
lem
s
is
h
o
w
t
h
e
y
c
a
n
f
i
n
d
th
e
o
b
jec
ts
i
n
t
h
e
ir
in
d
o
o
r
e
n
v
iro
n
m
e
n
t.
Th
is r
e
se
a
rc
h
wa
s p
re
se
n
ted
to
a
ss
ists v
isu
a
ll
y
imp
a
ired
p
e
o
p
le
i
n
f
in
d
i
n
g
th
e
o
b
jec
ts
i
n
o
ffice
.
Ob
jec
t
d
e
tec
ti
o
n
is
a
m
e
th
o
d
u
se
d
to
d
e
tec
t
t
h
e
o
b
jec
ts
i
n
ima
g
e
s
a
n
d
v
id
e
o
s.
M
a
n
y
a
lg
o
rit
h
m
s
u
se
d
fo
r
o
b
jec
t
d
e
tec
ti
o
n
su
c
h
a
s
c
o
n
v
o
lu
ti
o
n
a
l
n
e
u
ra
l
n
e
two
rk
(C
NN
)
a
n
d
y
o
u
o
n
l
y
lo
o
k
o
n
c
e
(YO
LO).
Th
e
p
ro
p
o
se
d
m
e
th
o
d
wa
s
YO
LO
wh
ich
o
u
t
p
e
rfo
rm
s
t
h
e
o
th
e
r
a
lg
o
rit
h
m
s
su
c
h
a
s
CNN
.
In
CNN
t
h
e
a
lg
o
rit
h
m
s
p
li
ts
t
h
e
ima
g
e
i
n
to
re
g
i
o
n
s.
Th
e
se
re
g
io
n
s
se
q
u
e
n
ti
a
ll
y
e
n
ters
th
e
n
e
u
ra
l
n
e
two
rk
fo
r
o
b
jec
t
d
e
tec
ti
o
n
a
n
d
r
e
c
o
g
n
it
i
o
n
so
CNN
d
o
e
s
n
o
t
d
e
a
l
with
a
ll
th
e
re
g
io
n
s
a
t
th
e
sa
m
e
ti
m
e
b
u
t
Y
OLO
lo
o
k
s
th
e
e
n
t
ire
ima
g
e
th
e
n
it
p
ro
d
u
c
e
s
th
e
b
o
u
n
d
in
g
b
o
x
e
s
w
it
h
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
two
rk
a
n
d
th
e
p
r
o
b
a
b
il
it
ies
o
f
th
e
se
b
o
x
e
s,
t
h
is
m
a
k
e
s
YO
LO
f
a
ste
r
th
a
n
o
th
e
r
a
lg
o
r
it
h
m
s.
O
p
e
n
so
u
rc
e
c
o
m
p
u
ter
v
isi
o
n
(Op
e
n
CV)
u
se
d
to
c
a
p
t
u
re
fra
m
e
s
b
y
u
sin
g
came
ra
.
T
h
e
n
YO
LO
u
se
d
to
d
e
tec
t
a
n
d
re
c
o
g
n
ize
t
h
e
o
b
jec
ts
i
n
e
a
c
h
fra
m
e
.
F
i
n
a
ll
y
,
th
e
so
u
n
d
i
n
Ara
b
ic
lan
g
u
a
g
e
wa
s
g
e
n
e
ra
ted
to
tell
th
e
v
is
u
a
ll
y
imp
a
ired
p
e
o
p
le
a
b
o
u
t
t
h
e
o
b
jec
ts.
Th
e
p
r
o
p
o
s
e
d
sy
ste
m
c
a
n
d
e
tec
t
6
o
b
jec
ts an
d
a
c
h
iev
e
a
n
a
c
c
u
ra
c
y
o
f
9
9
%
.
K
ey
w
o
r
d
s
:
C
o
m
p
u
ter
v
is
io
n
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
Ob
ject
d
ete
ctio
n
Op
en
C
V
Y
OL
O
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
:
Hass
an
Salam
Ab
d
u
l
-
Am
ee
r
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
E
n
g
i
n
ee
r
in
g
,
U
n
iv
er
s
ity
o
f
T
ec
h
n
o
lo
g
y
B
ag
h
d
ad
,
I
r
aq
E
m
ail:
ce
.
1
9
.
0
6
@
g
r
ad
.
u
o
tech
n
o
lo
g
y
.
ed
u
.
i
q
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
n
u
m
b
er
o
f
b
lin
d
a
n
d
v
is
u
ally
im
p
air
ed
p
eo
p
le
is
co
n
s
tan
t
ly
in
cr
ea
s
in
g
.
Acc
o
r
d
in
g
to
o
f
f
icial
s
tatis
t
ics
f
r
o
m
th
e
wo
r
ld
h
ea
lth
o
r
g
an
izatio
n
(
W
HO)
,
g
lo
b
ally
,
u
p
to
th
e
y
ea
r
o
f
2
0
1
1
,
th
er
e
ar
e
ab
o
u
t
2
8
5
m
illi
o
n
v
is
u
ally
im
p
air
ed
p
eo
p
l
e,
3
9
m
illi
o
n
am
o
n
g
th
em
a
r
e
co
m
p
letely
b
lin
d
an
d
2
4
6
m
illi
o
n
h
av
e
wea
k
s
ig
h
t
[
1
]
,
[
2
]
.
T
h
e
s
tatis
tics
o
f
W
HO
in
2
0
1
8
s
h
o
ws
th
at
th
e
r
e
is
n
ea
r
ly
1
b
illi
o
n
b
l
in
d
an
d
v
is
u
ally
im
p
air
ed
p
eo
p
le
[
3
]
.
W
h
ile,
in
2
0
2
0
,
it
b
ec
am
e
2
.
2
b
illi
o
n
,
th
is
in
cr
ea
s
es
th
e
n
ee
d
s
f
o
r
th
e
d
ev
ices
t
h
at
ar
e
u
s
ed
to
h
el
p
th
e
v
is
u
ally
im
p
air
e
d
p
eo
p
le
t
o
p
er
f
o
r
m
d
aily
task
s
.
T
h
e
r
ec
en
t
ad
v
an
ce
s
in
tech
n
o
lo
g
y
lead
to
d
e
v
elo
p
m
an
y
d
ev
ices
th
at
a
r
e
u
s
ed
to
ass
is
t
th
e
v
is
u
ally
im
p
air
e
d
p
eo
p
l
e
s
u
ch
as
s
m
ar
t
ey
e
g
lass
es.
T
h
e
p
r
o
p
o
s
ed
s
ma
r
t
e
y
eg
lass
es
s
y
s
tem
was
b
ased
o
n
c
o
m
p
u
t
e
r
v
is
io
n
.
C
o
m
p
u
te
r
v
is
io
n
is
a
tec
h
n
o
l
o
g
y
wh
ic
h
h
as
th
e
a
b
ilit
y
o
f
p
r
o
ce
s
s
in
g
an
d
u
n
d
er
s
tan
d
in
g
th
e
p
h
o
to
s
an
d
v
i
d
eo
s
by
u
s
in
g
m
ac
h
in
es
[
4
]
,
[
5
]
.
I
t
h
as
m
an
y
task
s
,
o
b
ject
d
etec
tio
n
is
o
n
e
o
f
its
f
u
n
d
am
en
tal
task
s
.
Ob
ject
d
etec
tio
n
is
a
m
e
th
o
d
th
at
d
etec
ts
th
e
o
b
jects
in
im
ag
es
an
d
v
id
eo
s
[6
]
,
[
7]
.
I
t
h
as
v
ar
io
u
s
ap
p
licatio
n
s
s
u
ch
as
s
elf
-
d
r
iv
in
g
ca
r
s
,
th
e
ap
p
licatio
n
s
wh
ich
ar
e
u
s
ed
to
h
el
p
th
e
b
lin
d
p
e
o
p
le
in
r
ec
o
g
n
izin
g
th
e
o
b
jects,
ca
r
p
late
d
etec
ti
o
n
,
au
to
m
ated
p
ar
k
in
g
s
y
s
tem
s
an
d
f
ac
e
d
etec
tio
n
[
8
]
,
[
9
]
.
T
h
er
e
ar
e
m
a
n
y
ty
p
es
o
f
d
ee
p
lear
n
in
g
alg
o
r
ith
m
s
th
at
ar
e
u
s
ed
to
p
er
f
o
r
m
o
b
ject
d
etec
tio
n
f
o
r
in
s
tan
ce
r
eg
io
n
b
ased
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
(
R
-
C
NN
),
an
d
YOL
O
[
1
0
]
.
T
h
e
s
u
g
g
ested
alg
o
r
ith
m
was
y
o
u
o
n
ly
lo
o
k
o
n
ce
v
er
s
io
n
3
(
YOL
O
v3)
.
YOL
O
v
3
is
b
a
s
ed
o
n
C
NN.
C
NN
is
a
d
ee
p
n
eu
r
al
n
etwo
r
k
t
h
at
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
20
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
1
09
-
1
17
110
co
n
s
is
ts
o
f
o
n
e
in
p
u
t
lay
er
,
m
o
r
e
th
an
o
n
e
h
id
d
e
n
lay
er
an
d
o
n
e
o
u
tp
u
t
lay
er
.
E
ac
h
lay
er
h
as
d
if
f
er
en
t
p
r
o
p
er
ties
.
T
h
e
f
ir
s
t
lay
er
in
C
o
n
v
o
lu
tio
n
n
eu
r
al
n
etw
o
r
k
(
C
NN
)
is
in
p
u
t
lay
er
wh
er
e
th
e
im
ag
e
en
ter
s
th
e
n
eu
r
al
n
etwo
r
k
th
r
o
u
g
h
it.
I
n
t
h
is
lay
er
,
th
e
n
u
m
b
er
o
f
n
e
u
r
o
n
s
is
th
e
s
am
e
as
th
e
n
u
m
b
er
o
f
f
ea
tu
r
es.
T
h
e
last
lay
er
in
C
NN
is
th
e
o
u
tp
u
t
lay
er
wh
er
e
th
e
n
u
m
b
er
o
f
n
eu
r
o
n
s
is
th
e
s
am
e
as
th
e
n
u
m
b
er
o
f
cl
ass
es.
Hid
d
en
lay
er
s
ar
e
co
n
v
o
l
u
tio
n
la
y
er
,
a
ctiv
atio
n
lay
er
,
p
o
o
lin
g
lay
er
an
d
f
u
lly
co
n
n
ec
ted
lay
er
.
CN
N
co
n
tain
s
at
least
o
n
e
c
o
n
v
o
lu
tio
n
lay
er
wh
ich
co
m
p
u
tes
a
d
o
t
p
r
o
d
u
ct
b
et
wee
n
th
e
co
n
n
ec
ted
r
eg
io
n
i
n
th
e
in
p
u
t
a
n
d
t
h
e
weig
h
ts
to
p
r
o
d
u
ce
th
e
f
ea
tu
r
e
m
ap
o
r
an
ac
tiv
atio
n
m
ap
.
T
h
e
r
o
le
o
f
ac
tiv
atio
n
lay
e
r
is
to
r
em
o
v
e
th
e
n
eg
ativ
e
v
alu
es
f
o
r
ac
ce
ler
atin
g
th
e
tr
ain
in
g
p
r
o
ce
s
s
.
T
h
e
ac
t
iv
atio
n
lay
e
r
r
esu
lt
is
p
o
o
led
b
y
p
o
o
lin
g
lay
e
r
t
o
s
im
p
lify
th
e
f
ea
tu
r
e
m
ap
.
T
h
e
f
u
lly
co
n
n
ec
ted
lay
er
is
u
s
ed
to
co
n
n
ec
t
th
e
o
u
tp
u
t
s
f
r
o
m
t
h
ese
lay
er
s
.
I
t
is
a
one
-
d
im
e
n
s
io
n
al
lay
er
,
h
a
s
al
l
th
e
lab
els
th
at
ar
e
to
b
e
cla
s
s
if
ied
an
d
it
p
r
o
d
u
ce
s
a
s
co
r
e
f
o
r
ea
ch
lab
el
o
f
class
if
icatio
n
[
1
1
]
-
[
1
3
]
.
T
h
e
v
is
u
ally
im
p
air
ed
p
eo
p
le
ar
e
ex
p
o
s
ed
to
m
an
y
p
r
o
b
lem
s
in
th
eir
d
aily
life
s
u
ch
as
d
is
co
v
er
in
g
th
e
o
b
jects in
th
eir
en
v
ir
o
n
m
e
n
t.
T
h
e
p
r
o
p
o
s
ed
s
m
ar
t e
y
eg
la
s
s
es
s
y
s
tem
s
o
lv
es th
is
p
r
o
b
lem
b
y
co
n
v
er
tin
g
th
e
v
is
u
al
s
ce
n
e
in
to
v
o
ice
m
ess
ag
e.
Ma
n
y
r
esear
ch
es
h
a
v
e
b
e
en
co
n
d
u
cted
to
im
p
lem
e
n
t
s
m
ar
t
ey
eg
lass
es
an
d
th
e
s
y
s
tem
s
th
at
ca
n
b
e
u
s
ed
t
o
h
elp
t
h
e
v
is
u
ally
im
p
air
ed
p
eo
p
le
b
y
u
s
in
g
th
e
d
ee
p
lear
n
i
n
g
alg
o
r
ith
m
s
s
u
ch
as
R
-
C
N
N,
C
NN,
an
d
YOL
O.
W
e
d
is
cu
s
s
s
o
m
e
o
f
r
elativ
e
m
eth
o
d
s
th
at
ca
n
b
e
ap
p
lied
to
d
etec
t
an
d
r
ec
o
g
n
ize
th
e
o
b
jects
[
1
4
].
B
h
ar
ti
et
a
l
.
[1
5
]
,
i
m
p
lem
en
t
s
a
s
y
s
tem
to
ass
is
t
th
e
b
lin
d
p
eo
p
le.
C
NN,
Op
en
-
s
o
u
r
ce
co
m
p
u
ter
v
is
io
n
(
Op
e
n
C
V
)
,
cu
s
t
o
m
d
ata
s
et
an
d
R
asp
b
er
r
y
Pi
ar
e
u
s
ed
.
T
h
e
s
y
s
tem
ca
n
d
etec
t
1
6
class
es.
T
h
e
ac
cu
r
ac
y
o
f
th
is
s
y
s
tem
is
9
0
%.
Ma
s
u
r
ek
ar
et
a
l
.
[1
6
]
,
cr
ea
te
s
an
o
b
ject
d
etec
tio
n
m
o
d
el
to
h
elp
th
e
b
lin
d
a
n
d
v
is
u
ally
im
p
air
e
d
p
eo
p
le.
YOL
O
v
3
a
n
d
th
e
c
u
s
to
m
d
ataset
wh
ich
co
n
tain
th
r
ee
class
es
(
b
u
s
,
m
o
b
ile
an
d
bot
tle)
a
r
e
u
s
ed
.
So
u
n
d
is
g
en
er
ated
u
s
in
g
Go
o
g
le
T
ex
t
T
o
Sp
ee
c
h
.
T
h
ey
f
o
u
n
d
th
at
th
e
a
cc
u
r
ac
y
o
f
th
is
m
o
d
el
is
9
8
%
an
d
th
e
r
eq
u
ir
ed
tim
e
to
d
etec
t
th
e
o
b
j
ec
ts
in
ea
ch
im
ag
e
is
eig
h
t
s
ec
o
n
d
s
.
Vaid
y
a
et
a
l
.
[1
7
]
,
I
m
p
lem
en
t
s
an
an
d
r
o
id
ap
p
licat
io
n
an
d
web
ap
p
licatio
n
f
o
r
o
b
ject
d
etec
tio
n
.
YOL
O
v
3
with
co
m
m
o
n
o
b
jects in
co
n
tex
t
(
C
OC
O)
d
at
aset
,
ar
e
u
s
ed
in
th
is
s
y
s
tem
.
T
h
ey
f
o
u
n
d
th
at
th
e
m
ax
im
u
m
a
cc
u
r
ac
y
in
m
o
b
ile
p
h
o
n
es
is
8
5
.
5
%
an
d
8
9
%
in
web
ap
p
lic
atio
n
s
a
n
d
th
e
re
q
u
ir
ed
tim
e
is
2
s
ec
o
n
d
s
,
th
e
tim
e
will
be
in
cr
ea
s
ed
b
y
in
cr
ea
s
in
g
th
e
n
u
m
b
e
r
o
f
o
b
jects.
Sh
aik
h
et
a
l
.
[1
8
]
,
u
s
es
R
asp
b
er
r
y
Pi,
YOL
O
v
3
an
d
C
OC
O
d
ataset
f
o
r
im
p
lem
en
t
in
g
a
n
o
b
ject
d
etec
t
io
n
s
y
s
tem
.
T
h
e
ac
cu
r
ac
y
is
1
0
0
%
(
f
o
r
clo
ck
,
ch
air
,
ce
llp
h
o
n
e
an
d
p
er
s
o
n
)
an
d
9
5
%
o
n
o
v
er
all
p
er
f
o
r
m
an
ce
.
T
h
e
d
ee
p
lear
n
in
g
al
g
o
r
ith
m
s
ar
e
u
s
ed
to
im
p
lem
en
t
o
th
e
r
k
in
d
s
o
f
s
y
s
tem
s
s
u
ch
as
a
s
ig
n
lan
g
u
ag
e
tr
an
s
latio
n
an
d
th
e
m
o
n
ito
r
in
g
s
y
s
te
m
s
.
Fah
ad
et
a
l
.
[
19
]
,
im
p
lem
en
t
s
a
s
ig
n
lan
g
u
ag
e
tr
an
s
latio
n
s
y
s
tem
.
C
NN
an
d
cu
s
to
m
d
ataset
ar
e
u
s
ed
.
T
h
e
s
y
s
tem
co
n
v
er
ts
th
e
s
ig
n
lan
g
u
ag
e
in
to
a
v
o
ice
m
ess
ag
e.
Fo
u
r
ty
h
a
n
d
g
estu
r
e
s
ar
e
r
ec
o
g
n
ize
d
b
y
th
is
s
y
s
tem
.
T
h
e
ac
h
ie
v
ed
ac
cu
r
ac
y
is
9
8
%.
Ab
d
u
lh
u
s
s
ein
an
d
R
ah
ee
m
[2
0
]
,
im
p
lem
e
n
t
s
a
h
an
d
g
est
u
r
e
r
ec
o
g
n
itio
n
s
y
s
tem
u
s
in
g
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
a
n
d
c
u
s
to
m
d
ataset.
T
wen
ty
-
f
o
u
r
letter
s
ar
e
r
ec
o
g
n
ized
.
T
h
e
ac
cu
r
ac
y
is
9
9
.
3
%.
Ma
h
m
o
o
d
an
d
Sau
d
[2
1
]
,
im
p
lem
en
ts
a
m
o
n
ito
r
in
g
s
y
s
tem
f
o
r
d
etec
tin
g
a
n
d
clas
s
if
y
in
g
th
e
m
o
v
i
n
g
v
eh
icles
in
v
i
d
e
o
s
u
s
in
g
co
n
v
o
l
u
tio
n
al
n
e
u
r
al
n
etwo
r
k
an
d
th
e
c
u
s
to
m
d
ataset.
T
h
e
Acc
u
r
ac
y
is
92%
.
Z
in
et
a
l
.
[2
2
]
,
cr
e
ated
a
h
e
r
b
al
p
lan
t
r
ec
o
g
n
itio
n
s
y
s
tem
b
y
u
s
in
g
co
n
v
o
lu
tio
n
al
n
e
u
r
al
n
e
two
r
k
with
th
e
cu
s
to
m
d
ataset.
T
welv
e
ty
p
es
o
f
p
lan
ts
ca
n
b
e
r
ec
o
g
n
ize
d
b
y
th
is
s
y
s
tem
.
T
h
e
ac
c
u
r
ac
y
is
99%
.
An
an
d
h
alli
et
a
l
.
[2
3
]
,
im
p
lem
en
ts
a
m
o
d
el
f
o
r
d
etec
tin
g
an
d
tr
ac
k
in
g
th
e
v
eh
icle
b
y
u
s
in
g
c
o
n
v
o
lu
tio
n
n
e
u
r
al
n
etwo
r
k
an
d
th
e
cu
s
to
m
d
at
aset.
T
h
e
ac
h
iev
ed
ac
cu
r
ac
y
is
9
0
.
8
8
%
.
T
h
e
p
r
o
p
o
s
ed
s
m
ar
t
ey
eg
lass
es
s
y
s
tem
u
s
es
YOL
O
v
3
with
cu
s
to
m
d
ataset.
T
h
is
s
y
s
tem
p
r
o
d
u
ce
s
a
h
ig
h
ac
cu
r
ac
y
in
d
etec
tin
g
an
d
r
ec
o
g
n
izi
n
g
th
e
o
b
jects wh
ich
is
eq
u
al
to
9
9
%.
2.
P
RO
P
O
SE
D
M
E
T
H
O
D
T
h
e
p
r
o
p
o
s
ed
s
m
ar
t
ey
eg
lass
es
s
y
s
tem
co
n
s
is
t
s
o
f
:
R
asp
b
er
r
y
Pi,
USB
ca
m
er
a
,
p
o
wer
b
an
k
an
d
ea
r
p
ho
n
e.
Fig
u
r
e
1
ex
p
lain
s
t
h
e
b
lo
ck
d
iag
r
a
m
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
.
T
h
e
s
u
g
g
ested
s
y
s
t
em
u
s
e
s
YOL
O
v
3
with
th
e
cu
s
to
m
d
ataset
f
o
r
d
etec
tio
n
an
d
r
ec
o
g
n
itio
n
th
e
s
tatic
o
b
jects
f
o
r
in
d
o
o
r
en
v
ir
o
n
m
en
t
s
u
ch
as
o
f
f
ice
o
r
r
o
o
m
.
Op
en
C
V
lib
r
ar
y
wa
s
u
s
ed
f
o
r
ca
p
t
u
r
in
g
an
d
p
r
o
c
ess
in
g
th
e
im
ag
es.
Play
s
o
u
n
d
lib
r
ar
y
to
p
lay
th
e
s
o
u
n
d
f
r
o
m
s
o
u
n
d
s
d
ataset
was
u
s
ed
with
th
i
s
s
y
s
tem
.
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
was
im
p
lem
en
ted
o
n
R
asp
b
er
r
y
P
i
4
Mo
d
el
B
with
p
y
th
o
n
la
n
g
u
a
g
e
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
f
o
r
t
h
is
s
m
ar
t
ey
e
g
lass
s
y
s
tem
co
n
s
is
ts
o
f
two
p
ar
ts
:
T
h
e
f
ir
s
t
p
a
r
t
was
th
e
tr
ain
in
g
p
r
o
ce
s
s
of
t
h
e
n
e
u
r
al
n
etwo
r
k
wh
ile
th
e
o
th
er
p
ar
t
was
h
o
w
to
u
s
e
th
is
n
eu
r
al
n
etwo
r
k
to
d
etec
t a
n
d
r
ec
o
g
n
ize
th
e
o
b
jects.
Fig
u
r
e
2
ex
p
lain
s
th
e
two
p
a
r
ts
o
f
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
.
T
h
e
f
o
llo
win
g
ar
e
th
e
s
tep
s
f
o
r
tr
ain
in
g
t
h
e
p
r
o
p
o
s
ed
d
ee
p
n
eu
r
al
n
etwo
r
k
:
−
Step
1
: a
s
et
o
f
co
lo
r
an
d
h
ig
h
-
r
eso
lu
tio
n
im
ag
es with
d
if
f
er
en
t sizes is
co
llected
.
−
Step
2
:
l
ab
elin
g
is
u
s
ed
to
lab
el
th
e
o
b
jects
in
ea
ch
im
ag
e.
L
ab
elin
g
is
an
im
ag
e
an
n
o
tatio
n
to
o
l
wh
ich
is
u
s
ed
f
o
r
lab
elin
g
th
e
o
b
jects in
ea
ch
im
ag
e.
−
Step
3
:
an
im
ag
e
an
n
o
tatio
n
f
ile
was
cr
ea
ted
f
o
r
ea
ch
im
ag
e.
T
h
e
d
ataset
n
o
w
is
r
ea
d
y
f
o
r
tr
ain
in
g
th
e
d
ee
p
n
eu
r
al
n
etwo
r
k
.
T
h
e
tr
ain
in
g
p
r
o
ce
s
s
is
ex
ec
u
ted
o
n
g
r
ap
h
ics p
r
o
c
ess
in
g
u
n
it (
GPU)
o
f
Go
o
g
le
C
o
lab
f
o
r
3
0
0
0
iter
a
tio
n
s
an
d
tak
es a
b
o
u
t th
r
ee
h
o
u
r
s
.
−
Step
4
:
at
th
e
en
d
o
f
th
e
tr
ain
in
g
p
r
o
ce
s
s
th
e
weig
h
t
f
ile
was
g
en
er
ated
.
Fig
u
r
e
2
(
a)
,
ex
p
lain
s
th
e
s
tep
s
f
o
r
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
Dev
elo
p
men
t sma
r
t e
ye
g
la
s
s
es
fo
r
visu
a
lly
imp
a
ir
ed
p
eo
p
le
b
a
s
ed
o
n
… (
Ha
s
s
a
n
S
a
la
m
A
b
d
u
l
-
A
mee
r
)
111
th
e
tr
ain
in
g
p
r
o
ce
s
s
.
T
h
e
f
o
llo
win
g
ar
e
th
e
s
tep
s
f
o
r
th
e
s
ec
o
n
d
p
ar
t o
f
p
r
o
p
o
s
ed
m
eth
o
d
:
−
Step
1
:
t
h
e
ca
m
er
a
will c
ap
tu
r
e
th
e
f
r
am
es (
im
ag
es)
b
y
u
s
in
g
Op
en
C
V
lib
r
ar
y
.
−
Step
2
:
e
ac
h
f
r
am
e
(
im
ag
e)
is
r
esized
to
4
1
6
x
4
1
6
u
s
in
g
Op
en
C
V.
−
Step
3
: Y
OL
O
v
3
is
u
s
ed
to
d
etec
t a
n
d
r
ec
o
g
n
ize
th
e
o
b
jects in
ea
ch
im
ag
e
b
ased
o
n
weig
h
t f
ile.
−
Step
4
:
w
h
en
th
e
f
r
am
e
h
as o
b
jects,
th
e
o
b
jects
will
b
e
d
etec
ted
an
d
r
ec
o
g
n
ized
.
I
f
th
er
e
is
n
o
o
b
ject
in
th
e
f
r
am
e,
th
en
th
e
n
ex
t f
r
am
e
will b
e
s
elec
ted
to
d
etec
t a
n
d
r
ec
o
g
n
ize
th
e
o
b
jects in
it.
−
Step
5
:
t
he
s
o
u
n
d
in
Ar
ab
ic
lan
g
u
ag
e
f
r
o
m
th
e
s
o
u
n
d
s
d
ataset
will
b
e
p
lay
ed
b
y
u
s
in
g
p
lay
s
o
u
n
d
lib
r
ar
y
to
tell
th
e
v
is
u
ally
im
p
air
ed
p
eo
p
le
ab
o
u
t
th
e
o
b
jects
in
th
e
f
r
am
e.
Fig
u
r
e
2
(
b
)
,
ex
p
lain
s
th
e
o
b
ject
d
etec
tio
n
an
d
r
ec
o
g
n
itio
n
b
y
th
e
p
r
o
p
o
s
ed
s
y
s
tem
.
Fig
u
r
e
1
.
T
h
e
B
lo
ck
d
iag
r
am
o
f
p
r
o
p
o
s
ed
s
y
s
tem
(
a)
(
b
)
Fig
u
r
e
2
.
T
h
e
two
p
a
r
ts
o
f
p
r
o
p
o
s
ed
m
eth
o
d
(
a)
ex
p
lain
s
th
e
s
tep
s
f
o
r
th
e
tr
ain
in
g
p
r
o
ce
s
s
an
d
(
b
)
ex
p
lai
n
s
th
e
o
b
ject
d
etec
tio
n
a
n
d
r
ec
o
g
n
iti
o
n
b
y
th
e
p
r
o
p
o
s
ed
s
y
s
tem
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
20
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
1
09
-
1
17
112
2
.
1
.
O
penCV
O
p
e
n
-
s
o
u
r
c
e
c
o
m
p
u
t
e
r
v
i
s
i
o
n
i
s
a
l
i
b
r
a
r
y
o
f
p
r
o
g
r
a
m
m
i
n
g
f
u
n
c
t
i
o
n
s
w
h
i
c
h
i
s
u
s
e
d
f
o
r
i
m
a
g
e
p
r
o
c
e
s
s
i
n
g
.
I
m
a
g
e
p
r
o
c
e
s
s
i
n
g
is
a
k
i
n
d
o
f
s
i
g
n
a
l
p
r
o
c
e
s
s
i
n
g
i
n
w
h
i
c
h
t
h
e
i
n
p
u
t
i
s
a
n
i
m
a
g
e
s
u
c
h
a
s
a
v
i
d
e
o
f
r
a
m
e
o
r
p
h
o
t
o
g
r
a
p
h
,
t
h
e
o
u
t
p
u
t
i
s
an
i
m
a
g
e
o
r
s
e
t
o
f
c
h
a
r
a
c
t
e
r
i
s
tic
s
r
e
l
at
e
d
t
o
t
h
e
i
m
a
g
e
.
O
p
e
n
C
V
w
a
s
s
ta
r
t
e
d
as
a
r
e
s
e
a
r
c
h
p
r
o
j
e
ct
b
y
I
n
t
e
l
.
I
t
c
o
n
t
a
i
n
s
v
a
r
i
o
u
s
t
o
o
ls
t
o
s
o
l
v
e
co
m
p
u
t
e
r
v
i
s
i
o
n
p
r
o
b
l
e
m
s
.
I
t
h
a
s
l
o
w
l
e
v
e
l
i
m
a
g
e
p
r
o
c
e
s
s
i
n
g
f
u
n
c
t
i
o
n
s
a
n
d
h
i
g
h
le
v
e
l
a
l
g
o
r
i
t
h
m
s
f
o
r
d
e
t
e
c
ti
n
g
f
a
c
e
s
,
f
e
a
t
u
r
e
m
a
t
c
h
i
n
g
a
n
d
t
r
a
c
k
i
n
g
[
2
4
]
.
2
.
2
.
Ra
s
pb
er
ry
Pi
4
m
o
del
B
R
asp
b
er
r
y
Pi
is
th
e
h
ea
r
t
o
f
th
e
p
r
o
p
o
s
ed
s
m
ar
t
ey
e
g
lass
es
s
y
s
tem
.
E
ar
p
h
o
n
e
is
co
n
n
ec
ted
to
R
asp
b
er
r
y
Pi
to
s
en
d
s
o
u
n
d
to
th
e
v
is
u
ally
im
p
air
ed
p
eo
p
l
e.
USB
C
am
er
a
wa
s
u
s
ed
b
e
ca
u
s
e
th
e
ca
b
le
o
f
R
asp
b
er
r
y
Pi
ca
m
er
a
is
s
tiff
an
d
d
if
f
ic
u
lt
to
m
ain
tain
.
Po
wer
b
an
k
is
th
e
p
o
wer
s
u
p
p
l
y
f
o
r
R
asp
b
er
r
y
Pi
.
R
asp
b
er
r
y
Pi
4
m
o
d
el
B
wit
h
1
2
8
GB
SD
C
ar
d
was
u
s
ed
in
th
e
p
r
o
p
o
s
ed
s
y
s
tem
.
T
h
e
co
d
e
in
p
y
th
o
n
lan
g
u
ag
e
was e
x
ec
u
ted
o
n
R
asp
b
er
r
y
Pi
.
2
.
3
.
YO
L
O
v
3
YOL
O
v
3
is
co
n
s
is
ted
o
f
th
e
C
NN
an
d
an
alg
o
r
ith
m
f
o
r
p
r
o
ce
s
s
in
g
th
e
o
u
tp
u
t
f
r
o
m
n
eu
r
al
n
etwo
r
k
[
2
5
]
.
C
NN
is
a
ty
p
e
o
f
d
ee
p
n
eu
r
al
n
etwo
r
k
s
.
I
t
is
u
s
ed
to
p
r
o
ce
s
s
im
a
g
e
d
ata.
YOL
O
v
3
is
a
f
ast,
m
u
lti
-
o
b
ject
d
etec
tio
n
,
r
ea
l
-
tim
e
d
ee
p
lear
n
i
n
g
al
g
o
r
ith
m
.
T
h
e
ex
ce
llen
t
p
r
o
ce
s
s
in
g
s
p
ee
d
o
f
YOL
O
v
3
is
th
e
f
ea
tu
r
e
th
at
m
a
k
es
it
o
u
tp
er
f
o
r
m
s
th
e
o
th
er
al
g
o
r
ith
m
s
,
as
i
t
ca
n
p
r
o
c
ess
4
5
f
r
am
es
p
e
r
s
ec
o
n
d
.
YOL
O
v
3
ap
p
lies
a
s
in
g
l
e
C
NN
to
an
e
n
ti
r
e
f
r
am
e
,
it
d
iv
id
e
s
f
r
am
e
i
n
to
“
S
”
x
“
S
”
g
r
id
s
th
en
p
r
e
d
icts
b
o
u
n
d
in
g
b
o
x
es
an
d
f
in
d
s
p
r
o
b
a
b
ilit
ies
f
o
r
th
ese
b
o
x
es
t
o
d
etec
t
an
d
r
ec
o
g
n
ize
th
e
o
b
jects
[
2
6
]
.
YOL
O
v
3
c
o
n
s
is
ts
o
f
1
0
6
lay
er
s
an
d
ca
n
d
etec
t
th
e
o
b
jec
ts
at
th
r
ee
d
if
f
er
en
t scale
s
.
As s
h
o
wn
in
F
ig
u
r
e
3
.
Fig
u
r
e
3
.
T
h
e
lay
er
s
o
f
YOL
O
v
3
[
2
5
]
YOL
O
v
3
alg
o
r
ith
m
wo
r
k
s
as
in
th
e
f
o
llo
win
g
:
−
YOL
O
v
3
tak
es
th
e
im
ag
es
(
f
r
am
es
)
f
r
o
m
ca
m
er
a
an
d
an
aly
s
es
ea
ch
im
ag
e
to
d
etec
t
an
d
r
ec
o
g
n
ize
th
e
o
b
jects
in
th
is
im
ag
e
.
I
t
d
iv
id
es
th
e
im
ag
e
in
to
‘
S’
x
‘
S’
g
r
id
s
.
E
ac
h
g
r
id
h
as
a
p
r
o
b
ab
ilit
y
o
f
en
clo
s
in
g
m
u
ltip
le
o
r
s
in
g
le
o
b
ject.
T
h
e
o
b
jects
ar
e
to
b
e
b
o
u
n
d
b
y
a
b
o
u
n
d
in
g
b
o
x
.
As
a
r
esu
lt
,
ea
ch
g
r
id
will
h
av
e
“
B
”
b
o
u
n
d
in
g
b
o
x
es a
n
d
p
r
o
b
ab
ilit
ies o
f
C
class
es
.
−
T
h
e
co
n
f
id
en
ce
s
co
r
e
is
ass
u
m
ed
(
g
en
er
ally
4
0
%
o
r
ab
o
v
e)
ac
co
r
d
in
g
to
wh
ich
b
o
u
n
d
in
g
b
o
x
is
p
r
ed
icted
ag
ain
s
t p
r
o
b
ab
ilit
y
o
f
C
class
es o
f
o
b
jects.
Pre
d
ictio
n
s
with
in
v
alid
co
n
f
id
en
ce
s
co
r
e
will n
o
t b
e
p
r
o
jecte
d
.
−
E
ac
h
b
o
u
n
d
ar
y
b
o
x
p
r
ed
ictio
n
in
clu
d
es
5
v
alu
es:
x
,
y
,
w,
h
,
an
d
co
n
f
id
en
ce
.
T
h
e
(
x
,
y
)
is
th
e
ce
n
ter
o
f
th
e
box
,
h
r
ep
r
esen
ts
th
e
h
eig
h
t
an
d
w
r
ep
r
esen
ts
th
e
wid
th
.
T
h
e
v
alu
es
o
f
“x
”,
“y
”,
“w”
an
d
“h
”
ar
e
b
etwe
en
[
0
,
1
]
.
T
h
er
e
ar
e
6
class
p
r
o
b
ab
ilit
ies
f
o
r
ea
ch
g
r
id
ce
ll
b
u
t
o
n
ly
o
n
e
class
p
r
o
b
ab
ilit
y
ca
n
b
e
p
r
ed
icted
p
er
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
Dev
elo
p
men
t sma
r
t e
ye
g
la
s
s
es
fo
r
visu
a
lly
imp
a
ir
ed
p
eo
p
le
b
a
s
ed
o
n
… (
Ha
s
s
a
n
S
a
la
m
A
b
d
u
l
-
A
mee
r
)
113
ce
ll.
T
h
e
f
in
al
p
r
ed
ictio
n
is
o
f
th
e
f
o
r
m
Sx
S
(
B
x
5
+C).
On
e
o
b
ject
o
n
ly
ca
n
b
e
d
etec
ted
p
er
ce
ll
in
th
e
Sx
S
g
r
id
.
−
E
ac
h
g
r
id
ce
ll
m
ay
d
etec
t
o
n
ly
o
n
e
o
b
ject;
th
er
ef
o
r
e,
an
ch
o
r
b
o
x
will
b
e
u
s
ed
to
en
ab
le
m
u
ltip
le
o
b
ject
d
etec
tio
n
.
C
o
n
s
id
er
th
e
im
ag
e
in
Fig
u
r
e
4
,
th
e
m
id
p
o
in
ts
o
f
th
e
ca
r
an
d
h
u
m
an
ar
e
in
th
e
s
am
e
ce
ll;
th
er
ef
o
r
e,
an
an
ch
o
r
b
o
x
is
u
s
ed
.
T
h
e
g
r
id
ce
lls
in
p
u
r
p
le
co
lo
r
ar
e
th
e
an
ch
o
r
b
o
x
es f
o
r
th
ese
o
b
jects.
An
y
n
u
m
b
er
o
f
th
e
an
c
h
o
r
b
o
x
es
m
ay
b
e
u
s
ed
f
o
r
d
etec
tin
g
m
u
ltip
le
o
b
jects
in
a
s
in
g
le
im
ag
e.
T
wo
an
ch
o
r
b
o
x
es a
r
e
u
s
ed
in
th
is
im
ag
e
[
1
6
]
.
−
W
h
en
two
o
r
m
o
r
e
g
r
id
ce
lls
in
Sx
S g
r
id
h
av
e
th
e
s
am
e
o
b
ject,
th
en
th
e
o
b
ject’
s
ce
n
ter
p
o
in
t is d
eter
m
in
ed
an
d
th
e
ce
ll
w
h
ich
h
as
th
is
ce
n
ter
p
o
in
t
is
s
elec
ted
.
T
h
er
e
ar
e
two
m
eth
o
d
s
f
o
r
d
ea
lin
g
with
m
u
ltip
le
b
o
u
n
d
in
g
b
o
x
es
th
at
ar
e
f
o
u
n
d
ar
o
u
n
d
th
e
o
b
ject
s
.
T
h
e
m
eth
o
d
s
ar
e
non
-
m
ax
s
u
p
p
r
ess
io
n
(
NM
S)
an
d
in
ter
s
ec
tio
n
o
v
er
u
n
io
n
(
I
o
U)
.
I
n
I
o
U
m
eth
o
d
,
I
f
th
e
in
ter
s
ec
tio
n
o
v
er
u
n
io
n
v
alu
e
f
o
r
th
e
b
o
u
n
d
in
g
b
o
x
es
is
eq
u
al
to
o
r
g
r
ea
ter
th
an
th
e
th
r
esh
o
ld
v
alu
e
(
o
u
r
th
r
esh
o
ld
v
alu
e
is
0
.
5
)
th
en
p
r
ed
ictio
n
is
g
o
o
d
.
T
h
e
ac
cu
r
ac
y
will
in
cr
ea
s
e
b
y
in
cr
ea
s
in
g
th
e
th
r
esh
o
ld
v
alu
e
[
1
6
]
.
I
n
th
e
s
ec
o
n
d
m
eth
o
d
(
NM
S
)
,
th
e
b
o
x
es
wh
ich
h
av
e
h
ig
h
p
r
o
b
ab
ilit
y
will
b
e
tak
en
an
d
th
e
b
o
x
es
with
h
ig
h
I
o
U
will
b
e
s
u
p
p
r
ess
ed
.
T
h
is
p
r
o
ce
s
s
is
r
ep
ea
ted
u
n
til a
b
o
x
is
s
elec
ted
an
d
co
n
s
id
er
ed
as th
e
b
o
u
n
d
in
g
b
o
x
f
o
r
th
e
o
b
ject
[
1
0
]
.
−
As
p
r
ev
io
u
s
ly
m
en
tio
n
ed
,
YOL
O
v
3
co
n
s
is
ts
o
f
1
0
6
lay
er
s
.
Ou
r
p
r
o
p
o
s
ed
m
eth
o
d
co
n
s
is
ts
o
f
9
4
lay
er
s
in
s
tead
o
f
1
0
6
lay
er
s
,
b
y
r
em
o
v
in
g
th
e
last
1
2
lay
er
s
f
r
o
m
YOL
O
v
3
alg
o
r
ith
m
to
d
ec
r
ea
s
e
th
e
r
eq
u
ir
ed
tim
e
f
o
r
o
b
ject
d
etec
tio
n
wh
ile
m
ain
tain
th
e
ac
cu
r
ac
y
as
o
u
r
s
y
s
tem
d
o
es
n
o
t
d
ea
l
with
v
er
y
s
m
all
o
b
ject
s
,
b
u
t w
ith
lar
g
e
an
d
m
ed
iu
m
o
b
jects to
en
ab
le
th
e
b
lin
d
p
eo
p
le
to
d
is
co
v
er
th
e
o
b
jects in
f
r
o
n
t o
f
th
em
.
Fig
u
r
e
4
.
T
h
e
an
c
h
o
r
b
o
x
es
2
.
4
.
Cus
t
o
m
da
t
a
s
et
(
im
a
g
e
da
t
a
s
et
)
I
n
th
e
tr
ain
in
g
p
r
o
ce
s
s
f
o
r
n
eu
r
al
n
etwo
r
k
s
,
m
an
y
im
ag
e
s
ar
e
r
eq
u
ir
ed
to
tr
ain
th
e
d
ee
p
lear
n
in
g
m
o
d
el.
T
h
e
p
r
ep
ar
ed
d
ataset
f
o
r
th
e
s
u
g
g
ested
s
y
s
tem
co
n
s
is
ts
o
f
1
5
6
0
lab
eled
im
ag
es
f
o
r
6
o
b
jects
(
T
V,
b
o
ttle,
p
er
s
o
n
,
ch
air
,
lap
to
p
an
d
tab
le
)
.
T
h
e
n
u
m
b
er
o
f
im
ag
es
th
at
b
elo
n
g
s
to
T
V
was
1
8
0
im
ag
es,
f
o
r
p
er
s
o
n
was
6
0
0
,
1
8
0
f
o
r
b
o
ttle,
1
8
0
f
o
r
ch
air
,
2
2
0
f
o
r
tab
le
an
d
2
0
0
f
o
r
lap
to
p
.
T
h
e
im
ag
es
ar
e
in
d
if
f
er
en
t
s
izes.
T
h
e
se
im
ag
es a
r
e
in
.
J
PG f
o
r
m
at.
Fig
u
r
e
5
ex
p
lain
s
s
o
m
e
o
f
d
ataset
im
ag
es.
Fig
u
r
e
5
.
I
m
ag
e
d
ataset
2
.
5
.
So
un
d da
t
a
s
et
A
s
et
o
f
v
o
ice
m
ess
ag
e
s
in
A
r
ab
ic
lan
g
u
ag
e
is
cr
ea
ted
an
d
s
to
r
ed
in
th
e
R
asp
b
er
r
y
Pi
.
W
h
en
th
e
o
b
ject
is
d
etec
ted
an
d
r
ec
o
g
n
i
ze
d
th
en
th
e
s
o
u
n
d
will
b
e
p
l
ay
ed
b
y
u
s
in
g
t
h
e
p
lay
s
o
u
n
d
lib
r
ar
y
,
to
tell
th
e
v
is
u
ally
im
p
air
ed
p
eo
p
le
ab
o
u
t
th
e
o
b
jects
in
ea
ch
f
r
am
e
(
in
f
r
o
n
t
o
f
h
im
/h
e
r
)
.
T
h
e
v
o
i
ce
m
ess
ag
es
ar
e
in
.
MP3
f
o
r
m
at.
T
h
is
m
eth
o
d
w
ill
co
n
v
er
t
th
e
tex
t
in
to
v
o
ice
m
ess
ag
e
with
o
u
t
u
s
in
g
th
e
I
n
ter
n
et
an
d
at
h
ig
h
s
p
ee
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
20
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
1
09
-
1
17
114
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
YOL
O
v
3
wh
ich
is
an
ac
cu
r
ate
an
d
f
ast
alg
o
r
ith
m
was
u
s
ed
in
th
e
p
r
o
p
o
s
ed
s
y
s
tem
.
T
h
e
d
ataset
wh
ic
h
co
n
s
is
ted
o
f
1
5
6
0
co
lo
u
r
im
ag
e
was
s
p
lit
ted
in
to
tw
o
g
r
o
u
p
s
.
T
h
e
f
ir
s
t
co
n
tain
s
ei
g
h
ty
-
f
i
v
e
p
er
ce
n
t
o
f
th
e
to
tal
im
a
g
es
as
tr
ain
in
g
i
m
ag
es
wh
ile
th
e
s
ec
o
n
d
co
n
ta
in
s
th
e
r
e
m
ain
d
er
as
test
in
g
im
ag
es.
T
h
e
tr
ain
in
g
p
r
o
ce
s
s
is
ex
ec
u
ted
o
n
GPU
o
f
Go
o
g
le
C
o
lab
an
d
ta
k
es
ab
o
u
t
th
r
ee
h
o
u
r
s
.
T
h
e
d
ee
p
n
e
u
r
a
l
n
etwo
r
k
is
tr
ain
ed
f
o
r
(
3
0
0
0
)
iter
atio
n
s
.
T
h
e
p
r
o
p
o
s
ed
s
m
ar
t
ey
eg
lass
es
s
y
s
tem
ca
n
b
e
u
s
ed
f
o
r
in
d
o
o
r
en
v
ir
o
n
m
e
n
t
s
s
u
ch
as
r
o
o
m
o
r
o
f
f
ice
an
d
it
ca
n
d
etec
t
m
u
ltip
le
o
b
jects.
Als
o
,
th
e
s
y
s
tem
ca
n
d
etec
t
th
e
o
b
jects
ev
en
if
th
e
d
is
tan
ce
b
etwe
en
th
e
USB
C
am
er
a
an
d
o
b
jects
is
g
r
ea
ter
th
an
3
m
eter
s
.
T
h
e
m
ea
n
av
er
ag
e
p
r
ec
is
io
n
(
m
AP)
o
f
t
h
e
s
m
ar
t
ey
eg
lass
es
s
y
s
tem
was
1
0
0
%.
T
h
e
m
AP
is
u
s
ed
f
o
r
ev
alu
atin
g
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
s
u
g
g
ested
s
y
s
tem
.
Fig
u
r
e
6
s
h
o
ws
th
e
m
AP
an
d
L
o
s
s
f
o
r
th
e
s
u
g
g
ested
m
eth
o
d
.
T
h
er
e
a
r
e
o
th
e
r
v
al
u
es
th
at
ar
e
u
s
ed
f
o
r
ev
alu
atin
g
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
s
u
g
g
ested
s
y
s
tem
s
u
ch
as
p
r
ec
is
io
n
,
I
o
U
,
R
ec
all,
F1
-
s
co
r
e,
tr
u
e
p
o
s
itiv
e
(
TP
),
f
alse
p
o
s
itiv
e
(
FP
)
an
d
f
alse
n
eg
ativ
e
(
FN
)
.
Fig
u
r
e
7
ex
p
lain
s
th
e
r
esu
lt
s
o
f
tr
ain
in
g
p
r
o
ce
s
s
(
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
)
.
Fig
u
r
e
8
s
h
o
ws th
e
o
b
ject
d
etec
tio
n
b
y
u
s
in
g
s
m
ar
t
ey
eg
lass
es
.
T
ab
le
1
ex
p
lain
s
th
e
co
m
p
ar
is
o
n
b
etw
ee
n
s
u
g
g
ested
ap
p
r
o
ac
h
an
d
o
t
h
er
r
elate
d
a
p
p
r
o
ac
h
.
Fig
u
r
e
6
.
L
o
s
s
an
d
m
AP
F
ig
u
r
e
7
.
T
h
e
r
esu
lt o
f
tr
ain
in
g
p
r
o
ce
s
s
3
.
1
.
Co
nfusi
o
n m
a
t
rix
T
h
e
co
n
f
u
s
io
n
m
atr
i
x
,
wh
ich
is
also
ca
lled
an
er
r
o
r
m
atr
ix
,
is
a
s
u
m
m
ar
y
th
at
g
i
v
e
th
e
r
e
s
u
lt
o
f
th
e
p
r
ed
ictio
n
.
T
h
e
n
u
m
b
er
o
f
in
co
r
r
ec
t
an
d
c
o
r
r
ec
t
p
r
ed
ictio
n
s
is
s
u
m
m
ar
ized
with
co
u
n
ted
v
alu
es
an
d
b
r
o
k
en
d
o
wn
class
b
y
class
.
T
h
e
er
r
o
r
m
atr
ix
(
co
n
f
u
s
io
n
m
atr
ix
)
ex
p
lain
s
h
o
w
th
e
m
o
d
el
is
co
n
f
u
s
ed
wh
en
it
m
ak
es
p
r
ed
ictio
n
s
.
T
h
e
co
n
f
u
s
io
n
m
atr
ix
g
iv
es
th
e
in
s
ig
h
t
n
o
t
o
n
l
y
in
to
th
e
er
r
o
r
s
b
ein
g
m
ad
e
b
y
th
e
class
if
ier
b
u
t
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
Dev
elo
p
men
t sma
r
t e
ye
g
la
s
s
es
fo
r
visu
a
lly
imp
a
ir
ed
p
eo
p
le
b
a
s
ed
o
n
… (
Ha
s
s
a
n
S
a
la
m
A
b
d
u
l
-
A
mee
r
)
115
also
g
iv
e
th
e
t
y
p
e
o
f
er
r
o
r
[1
6
].
FP
r
ef
er
s
to
th
e
n
u
m
b
e
r
o
f
in
co
r
r
ec
t
d
etec
tio
n
s
.
T
h
e
n
u
m
b
er
o
f
c
o
r
r
ec
tly
d
etec
ted
o
b
jects
is
r
ep
r
esen
ted
b
y
T
P.
T
h
e
FN r
e
f
er
s
to
n
u
m
b
er
o
f
m
is
s
ed
d
etec
tio
n
.
Fo
r
T
V
T
P =
2
7
an
d
FP
=
0
Fo
r
ch
air
T
P =
2
7
an
d
FP
=
1
T
o
tal
T
P =
234
Fo
r
p
er
s
o
n
T
P =
9
0
an
d
FP
=
0
Fo
r
tab
le
T
P =
3
0
an
d
FP
=
0
T
o
tal
FP
=
2
Fo
r
b
o
ttle
T
P =
2
7
an
d
FP
=
0
Fo
r
lap
to
p
T
P =
3
3
an
d
FP
=
1
Fig
u
r
e
8
.
T
h
e
r
esu
lts
o
f
o
b
ject
d
etec
tio
n
u
s
in
g
s
m
ar
t e
y
eg
las
s
es sy
s
tem
T
ab
le
1.
T
h
e
co
m
p
ar
is
o
n
b
etw
ee
n
th
e
s
u
g
g
ested
a
p
p
r
o
ac
h
an
d
o
th
er
r
elate
d
ap
p
r
o
ac
h
A
u
t
h
o
r
M
e
t
h
o
d
A
c
c
u
r
a
c
y
n
o
.
o
f
o
b
j
e
c
t
s
P
r
o
c
e
ss
i
n
g
t
i
m
e
M
a
s
u
r
e
k
a
r
e
t
a
l
.
[
1
7
]
Y
O
LO
v
3
,
C
u
s
t
o
m
d
a
t
a
se
t
a
n
d
G
o
o
g
l
e
Te
x
t
to
S
p
e
e
c
h
(
G
TTS
)
9
8
%
3
8
se
c
Th
e
p
r
o
p
o
s
e
d
met
h
o
d
Y
O
LO
v
3
,
C
u
s
t
o
m D
a
t
a
s
e
t
,
O
p
e
n
C
V
,
P
l
a
y
s
o
u
n
d
a
n
d
R
a
s
p
b
e
r
r
y
P
i
.
9
9
%
6
N
e
a
r
l
y
2
0
sec
3
.
2
.
P
re
cisi
o
n
T
h
e
p
r
ec
is
io
n
is
u
s
ed
to
m
ea
s
u
r
e
h
o
w
ac
cu
r
ate
th
e
p
r
ed
icti
o
n
s
ar
e.
Pre
cisi
o
n
ca
lcu
latio
n
will
b
e
as:
th
e
d
iv
is
io
n
o
f
TP
o
v
e
r
th
e
s
u
m
o
f
FP
an
d
TP
.
In
(
1
)
e
x
p
lain
s
th
e
p
r
ec
is
io
n
[
2
7
]
.
T
h
e
o
b
tain
ed
v
alu
e
is
0
.
9
9
.
Pr
e
c
ision
=
TP
TP
+
FP
(
1
)
3
.
3
.
Rec
a
ll
T
h
e
R
ec
all
is
u
s
ed
f
o
r
ca
lcu
latin
g
th
e
tr
u
e
p
r
ed
ictio
n
f
r
o
m
t
h
e
all
co
r
r
ec
tly
p
r
ed
icted
d
ata
.
R
ec
all
ca
lcu
latio
n
will
b
e
as:
th
e
d
iv
is
io
n
o
f
T
P
o
v
er
t
h
e
s
u
m
o
f
FN
T
P
.
I
n
(
2
)
ex
p
lain
s
th
e
r
ec
all
m
etr
ic
[
2
7
]
.
T
h
e
o
b
tain
ed
R
ec
all
v
alu
e
is
1
.
0
0
.
R
e
c
a
l
l
=
TP
TP
+
FN
(
2
)
3
.
4
.
F1
-
s
co
re
T
h
e
h
a
r
m
o
n
ic
m
ea
n
(
HM
)
o
f
th
e
Pre
cisi
o
n
an
d
th
e
R
ec
all.
I
t
is
o
n
e
o
f
m
etr
ics
th
at
u
s
ed
f
o
r
p
e
r
f
o
r
m
an
ce
ev
alu
atio
n
.
Ob
tain
e
d
v
alu
e
is
1
.
0
0
.
3
.
5
.
I
o
U
T
h
e
d
iv
is
io
n
o
f
in
ter
s
ec
tio
n
a
r
ea
o
v
er
a
r
ea
o
f
u
n
io
n
b
etwe
en
g
r
o
u
n
d
tr
u
t
h
b
o
u
n
d
i
n
g
b
o
x
a
n
d
d
etec
tio
n
b
o
u
n
d
in
g
b
o
x
f
o
r
a
s
p
ec
if
ic
th
r
esh
o
ld
.
In
(
3
)
s
h
o
ws th
e
I
o
U
[
2
7
]
.
T
h
e
o
b
tain
ed
a
v
er
ag
e
I
o
U
is
8
6
.
4
8
%.
IoU
=
A
r
ea
of
In
t
er
s
ect
i
o
n
A
r
ea
of
Un
i
o
n
×
100%
(
3
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
20
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
1
09
-
1
17
116
3
.
6
.
m
AP
T
h
e
m
ea
n
a
v
er
ag
e
p
r
ec
is
io
n
is
th
e
av
er
ag
e
o
f
a
v
er
ag
e
p
r
ec
is
io
n
(
AP)
th
at
is
ca
lcu
latin
g
f
o
r
all
class
e
s
.
I
n
(
4
)
ex
p
lain
s
th
e
m
ea
n
a
v
er
a
g
e
p
r
ec
is
io
n
.
T
h
e
o
b
tain
ed
m
AP is 1
0
0
%.
mAP
=
s
um
of
AP
f
o
r
t
he
t
o
t
al
cl
as
s
es
no
.
of
t
o
t
al
cl
as
s
es
×
100%
(
4
)
4.
CO
NCLU
SI
O
N
T
h
er
e
ar
e
m
an
y
ty
p
es
o
f
alg
o
r
ith
m
s
th
at
ar
e
u
s
ed
f
o
r
o
b
ject
d
etec
tio
n
an
d
r
ec
o
g
n
itio
n
f
o
r
in
s
tan
ce
R
-
C
NN,
f
ast
R
-
C
NN
an
d
YOL
O.
T
h
e
s
u
g
g
ested
s
m
ar
t
ey
eg
lass
es
s
y
s
tem
u
s
es
YOL
O
v3
b
ec
au
s
e
it
is
a
f
ast
an
d
ac
cu
r
ate
m
eth
o
d
.
YOL
O
v
3
ca
n
d
etec
t
m
u
ltip
le
o
b
jects
in
an
im
ag
e
.
T
h
e
p
lay
s
o
u
n
d
is
a
lib
r
ar
y
th
at
is
u
s
ed
to
p
lay
th
e
s
o
u
n
d
s
o
th
e
s
u
g
g
ested
s
y
s
tem
n
o
t
n
ee
d
s
co
n
n
ec
tio
n
to
th
e
in
ter
n
et
f
o
r
co
n
v
er
tin
g
th
e
tex
t
in
to
a
v
o
ice
m
ess
ag
e.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
ac
h
iev
es
an
ac
cu
r
ac
y
o
f
99
%.
T
h
e
r
eq
u
ir
ed
tim
e
f
o
r
d
etec
tio
n
is
n
ea
r
ly
t
wen
ty
s
ec
o
n
d
s
o
n
R
asp
b
er
r
y
Pi
wh
ile
o
n
th
e
p
er
s
o
n
al
co
m
p
u
ter
is
n
ea
r
ly
two
s
ec
o
n
d
s
.
I
n
f
u
tu
r
e,
th
e
s
m
a
r
t
ey
eg
lass
es
ca
n
b
e
im
p
lem
en
tin
g
o
n
NVI
DI
A
J
et
s
o
n
Nan
o
in
s
tead
o
f
R
asp
b
er
r
y
Pi
to
d
ec
r
ea
s
e
th
e
r
eq
u
ir
ed
tim
e
to
d
etec
t
th
e
o
b
jects
f
o
r
ea
ch
f
r
am
e.
Gen
d
er
p
r
ed
ictio
n
tech
n
iq
u
e
an
d
ag
e
p
r
ed
ictio
n
tech
n
iq
u
e
ca
n
also
b
e
ad
d
ed
to
p
r
ed
ict
th
e
g
en
d
er
an
d
ag
e
if
th
e
d
etec
ted
o
b
ject
is
p
er
s
o
n
.
I
n
ad
d
itio
n
,
th
e
t
ec
h
n
iq
u
e
o
f
f
ac
e
r
ec
o
g
n
itio
n
ca
n
b
e
ad
d
ed
to
ass
is
t th
e
b
lin
d
p
eo
p
le
in
r
ec
o
g
n
izin
g
th
e
p
eo
p
le
in
f
r
o
n
t o
f
th
em
.
RE
F
E
R
E
NC
E
S
[
1
]
J.
B
a
i
,
S
.
L
i
a
n
,
Z.
Li
u
,
K
.
W
a
n
g
,
a
n
d
D
.
L
i
u
,
“
S
mar
t
g
u
i
d
i
n
g
g
l
a
ss
e
s
f
o
r
v
i
su
a
l
l
y
i
mp
a
i
r
e
d
p
e
o
p
l
e
i
n
i
n
d
o
o
r
e
n
v
i
r
o
n
me
n
t
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
o
n
su
m
e
r E
l
e
c
t
r
o
n
i
c
s
,
v
o
l
.
6
3
,
n
o
.
3
,
p
p
.
2
5
8
–
2
6
6
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
E.
2
0
1
7
.
0
1
4
9
8
0
.
[
2
]
H
.
Ja
b
n
o
u
n
,
F
.
B
e
n
z
a
r
t
i
,
a
n
d
H
.
A
mi
r
i
,
“
O
b
j
e
c
t
d
e
t
e
c
t
i
o
n
a
n
d
i
d
e
n
t
i
f
i
c
a
t
i
o
n
f
o
r
b
l
i
n
d
p
e
o
p
l
e
i
n
v
i
d
e
o
sc
e
n
e
,
”
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
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
D
e
si
g
n
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
(
I
S
D
A
)
,
v
o
l
.
2
0
1
6
-
J
u
n
e
,
p
p
.
3
6
3
–
3
6
7
,
2
0
1
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
S
D
A
.
2
0
1
5
.
7
4
8
9
2
5
6
.
[
3
]
N
.
S
a
t
a
n
i
,
S
.
P
a
t
e
l
,
a
n
d
S
.
P
a
t
e
l
,
“
A
I
P
o
w
e
r
e
d
G
l
a
sses
f
o
r
V
i
s
u
a
l
l
y
I
mp
a
i
r
e
d
P
e
r
s
o
n
,
”
In
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
R
e
c
e
n
t
T
e
c
h
n
o
l
o
g
y
a
n
d
En
g
i
n
e
e
ri
n
g
(
I
J
RTE)
,
v
o
l
.
9
,
n
o
.
2
,
p
p
.
4
1
6
–
4
2
1
,
2
0
2
0
,
d
o
i
:
1
0
.
3
5
9
4
0
/
i
j
r
t
e
.
b
3
5
6
5
.
0
7
9
2
2
0
.
[
4
]
H
.
B
h
o
r
sh
e
t
t
i
,
S
.
G
h
u
g
e
,
A
.
K
u
l
k
a
r
n
i
,
P
.
S
.
B
h
i
n
g
a
r
k
a
r
,
a
n
d
P
.
N
.
L
o
k
h
a
n
d
e
,
“
L
o
w
B
u
d
g
e
t
S
m
a
r
t
G
l
a
sses
f
o
r
V
i
su
a
l
l
y
I
mp
a
i
r
e
d
P
e
o
p
l
e
,
”
1
0
t
h
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
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
N
e
t
w
o
r
k
s (I
C
-
I
S
C
N
2
0
1
9
)
,
2
0
1
9
,
p
p
.
4
8
–
5
2
.
[
5
]
M
.
V
a
i
d
h
e
h
i
,
V
.
S
e
t
h
,
a
n
d
B
.
S
i
n
g
h
a
l
,
“
R
e
a
l
-
T
i
me
O
b
j
e
c
t
D
e
t
e
c
t
i
o
n
f
o
r
A
i
d
i
n
g
V
i
s
u
a
l
l
y
I
mp
a
i
r
e
d
u
s
i
n
g
D
e
e
p
L
e
a
r
n
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
E
n
g
i
n
e
e
ri
n
g
a
n
d
Ad
v
a
n
c
e
T
e
c
h
n
o
l
o
g
y
(
I
J
EAT)
,
v
o
l
.
9
,
n
o
.
4
,
p
p
.
1
6
0
0
–
1
6
0
5
,
2
0
2
0
,
d
o
i
:
1
0
.
3
5
9
4
0
/
i
j
e
a
t
.
d
8
3
7
4
.
0
4
9
4
2
0
.
[
6
]
V
.
K
h
a
r
c
h
e
n
k
o
a
n
d
I
.
C
h
y
r
k
a
,
“
D
e
t
e
c
t
i
o
n
o
f
A
i
r
p
l
a
n
e
s
o
n
t
h
e
G
r
o
u
n
d
U
si
n
g
Y
O
LO
N
e
u
r
a
l
N
e
t
w
o
r
k
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
Ma
t
h
e
m
a
t
i
c
a
l
M
e
t
h
o
d
s
El
e
c
t
r
o
m
a
g
n
e
rt
i
c
T
h
e
o
r
y
,
MM
ET
,
2
0
1
8
,
p
p
.
2
9
4
–
2
9
7
,
d
o
i
:
1
0
.
1
1
0
9
/
M
M
E
T.
2
0
1
8
.
8
4
6
0
3
9
2
.
[
7
]
Z.
C
h
e
n
g
,
J
.
L
v
,
A
.
W
u
,
a
n
d
N
.
Q
u
,
“
Y
O
LO
v
3
O
b
j
e
c
t
D
e
t
e
c
t
i
o
n
A
l
g
o
r
i
t
h
m
w
i
t
h
F
e
a
t
u
r
e
P
y
r
a
m
i
d
A
t
t
e
n
t
i
o
n
f
o
r
R
e
m
o
t
e
S
e
n
si
n
g
I
mag
e
s,”
S
e
n
s
o
rs
M
a
t
e
r.
,
v
o
l
.
3
2
,
n
o
.
1
2
,
p
p
.
4
5
3
7
-
4
5
5
8
,
2
0
2
0
,
d
o
i
:
1
0
.
1
8
4
9
4
/
sam
.
2
0
2
0
.
3
1
3
0
.
[
8
]
S
.
N
.
S
r
i
v
a
t
s
a
,
G
.
S
r
e
e
v
a
t
h
sa
,
G
.
V
i
n
a
y
,
a
n
d
P
.
E
l
a
i
y
a
r
a
j
a
,
“
O
b
j
e
c
t
D
e
t
e
c
t
i
o
n
u
s
i
n
g
D
e
e
p
L
e
a
r
n
i
n
g
w
i
t
h
O
p
e
n
C
V
a
n
d
P
y
t
h
o
n
,
”
,
I
n
t
e
r
n
a
t
i
o
n
a
l
Re
s
e
a
r
c
h
J
o
u
r
n
a
l
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
(
I
RJ
ET)
,
v
o
l
.
8,
n
o
.
1,
p
p
.
2
2
7
–
2
3
0
,
2
0
2
1
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
i
r
j
e
t
.
n
e
t
/
a
r
c
h
i
v
e
s/
V
8
/
i
1
/
I
R
JET
-
V
8
I
1
4
5
.
p
d
f
[
9
]
M
.
S
h
a
h
a
n
d
R
.
K
a
p
d
i
,
“
O
b
j
e
c
t
d
e
t
e
c
t
i
o
n
u
s
i
n
g
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s,”
i
n
Pr
o
c
.
2
0
1
7
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
I
n
t
e
l
l
i
g
e
n
t
C
o
m
p
u
t
i
n
g
a
n
d
C
o
n
t
ro
l
S
y
s
t
e
m
s (I
C
I
C
C
S
)
2
0
1
7
,
2
0
1
7
,
p
p
.
7
8
7
–
7
9
0
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
O
N
S
.
2
0
1
7
.
8
2
5
0
5
7
0
.
[
1
0
]
S
.
G
e
e
t
h
a
p
r
i
y
a
,
N
.
D
u
r
a
i
m
u
r
u
g
a
n
,
a
n
d
S
.
P
.
C
h
o
k
k
a
l
i
n
g
a
m,
“
R
e
a
l
t
i
me
o
b
j
e
c
t
d
e
t
e
c
t
i
o
n
w
i
t
h
y
o
l
o
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
En
g
i
n
e
e
ri
n
g
a
n
d
A
d
v
a
n
c
e
d
T
e
c
h
n
o
l
o
g
y
(
I
J
EAT)
,
v
o
l
.
8
,
n
o
.
3
,
p
p
.
5
7
8
–
5
8
1
,
2
0
1
9
.
[
1
1
]
K
.
P
o
t
d
a
r
,
C
.
D
.
P
a
i
,
a
n
d
S
.
A
k
o
l
k
a
r
,
“
A
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
b
a
se
d
Li
v
e
O
b
j
e
c
t
R
e
c
o
g
n
i
t
i
o
n
S
y
s
t
e
m
a
s
B
l
i
n
d
A
i
d
,
”
a
rXi
v
,
2
0
1
8
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
a
r
x
i
v
.
o
r
g
/
p
d
f
/
1
8
1
1
.
1
0
3
9
9
.
p
d
f
[
1
2
]
P
.
S
.
B
h
a
i
r
n
a
l
l
y
k
a
r
,
A
.
P
r
a
j
a
p
a
t
i
,
A
.
R
a
j
b
h
a
r
,
a
n
d
S
.
M
u
j
a
w
a
r
,
“
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
(
C
N
N
)
f
o
r
I
mag
e
D
e
t
e
c
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
R
e
se
a
rc
h
J
o
u
rn
a
l
o
f
En
g
i
n
e
e
ri
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
(
I
RJ
E
T
)
,
v
o
l
.
7
,
n
o
.
1
1
,
p
p
.
1
2
3
9
–
1
2
4
3
,
2
0
2
0
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
w
w
w
.
i
r
j
e
t
.
n
e
t
/
a
r
c
h
i
v
e
s
/
V
7
/
i
1
1
/
I
R
JET
-
V
7
I
1
1
2
0
4
.
p
d
f
.
[
1
3
]
T.
S
.
G
u
n
a
w
a
n
e
t
a
l
.
,
“
D
e
v
e
l
o
p
me
n
t
o
f
v
i
d
e
o
-
b
a
s
e
d
e
m
o
t
i
o
n
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
w
i
t
h
G
o
o
g
l
e
C
o
l
a
b
,
”
T
ELKO
MN
I
K
A
T
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
,
C
o
m
p
u
t
i
n
g
,
El
e
c
t
r
o
n
i
c
s
a
n
d
C
o
n
t
r
o
l
,
v
o
l
.
1
8
,
n
o
.
5
,
p
p
.
2
4
63
–
2
4
7
1
,
2
0
2
0
,
d
o
i
:
1
0
.
1
2
9
2
8
/
TE
LK
O
M
N
I
K
A
.
v
1
8
i
5
.
1
6
7
1
7
.
[
1
4
]
A
.
A
b
d
u
r
r
a
s
y
i
d
,
I
.
I
n
d
r
i
a
n
t
o
,
a
n
d
R
.
A
r
i
a
n
t
o
,
“
D
e
t
e
c
t
i
o
n
o
f
i
mm
o
v
a
b
l
e
o
b
j
e
c
t
s
o
n
v
i
s
u
a
l
l
y
i
m
p
a
i
r
e
d
p
e
o
p
l
e
w
a
l
k
i
n
g
a
i
d
s,
”
T
ELKO
MN
I
K
A
T
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
,
C
o
m
p
u
t
i
n
g
,
El
e
c
t
r
o
n
i
c
s
a
n
d
C
o
n
t
r
o
l
,
v
o
l
.
1
7
,
n
o
.
2
,
p
p
.
5
8
0
–
5
8
5
,
2
0
1
9
,
d
o
i
:
1
0
.
1
2
9
2
8
/
TEL
K
O
M
N
I
K
A
.
V
1
7
I
2
.
9
9
3
3
.
[
1
5
]
R
.
B
h
a
r
t
i
,
K
.
B
h
a
d
a
n
e
,
P
.
B
h
a
d
a
n
e
,
a
n
d
A
.
G
a
d
h
e
,
“
O
b
j
e
c
t
D
e
t
e
c
t
i
o
n
a
n
d
R
e
c
o
g
n
i
t
i
o
n
f
o
r
B
l
i
n
d
A
ssi
s
t
a
n
c
e
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
Re
se
a
rc
h
J
o
u
r
n
a
l
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
(
I
R
J
ET)
,
v
o
l
.
6
,
n
o
.
5
,
p
p
.
7
0
8
5
–
7
0
8
7
,
2
0
1
9
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
i
r
j
e
t
.
n
e
t
/
a
r
c
h
i
v
e
s/
V
6
/
i
5
/
I
R
JET
-
V
6
I
5
1
0
0
2
.
p
d
f
.
[
1
6
]
O
.
M
a
s
u
r
e
k
a
r
,
O
.
Ja
d
h
a
v
,
P
.
K
u
l
k
a
r
n
i
,
a
n
d
S
.
P
a
t
i
l
,
“
R
e
a
l
Ti
m
e
O
b
j
e
c
t
D
e
t
e
c
t
i
o
n
U
s
i
n
g
Y
O
LO
v
3
,
”
I
n
t
.
Re
s
.
J
.
E
n
g
.
T
e
c
h
n
o
l
.
,
v
o
l
.
7
,
n
o
.
3
,
p
p
.
3
7
6
4
–
3
7
6
8
,
2
0
2
0
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
w
w
w
.
i
r
j
e
t
.
n
e
t
/
a
r
c
h
i
v
e
s/
V
7
/
i
3
/
I
R
JET
-
V
7
I
3
7
5
6
.
p
d
f
[
1
7
]
S
.
V
a
i
d
y
a
,
N
.
S
h
a
h
,
N
.
S
h
a
h
,
a
n
d
R
.
S
h
a
n
k
a
r
m
a
n
i
,
“
R
e
a
l
-
T
i
me
O
b
j
e
c
t
D
e
t
e
c
t
i
o
n
f
o
r
V
i
su
a
l
l
y
C
h
a
l
l
e
n
g
e
d
P
e
o
p
l
e
,
”
i
n
Pr
o
c
.
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
t
e
l
l
i
g
e
n
t
C
o
m
p
u
t
i
n
g
a
n
d
C
o
n
t
r
o
l
S
y
st
e
m
s
(
I
C
I
C
C
S
2
0
2
0
)
,
2
0
2
0
,
p
p
.
3
1
1
–
3
1
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
C
C
S
4
8
2
6
5
.
2
0
2
0
.
9
1
2
1
0
8
5
.
[
1
8
]
S
.
S
h
a
i
k
h
,
V
.
K
a
r
a
l
e
,
a
n
d
G
.
T
a
w
d
e
,
“
A
ssi
s
t
i
v
e
O
b
j
e
c
t
R
e
c
o
g
n
i
t
i
o
n
S
y
st
e
m
f
o
r
V
i
s
u
a
l
l
y
I
mp
a
i
r
e
d
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
En
g
i
n
e
e
ri
n
g
Re
s
e
a
rc
h
& T
e
c
h
n
o
l
o
g
y
(
I
J
ERT)
,
v
o
l
.
9
,
n
o
.
9
,
p
p
.
7
3
6
–
7
4
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
7
5
7
7
/
i
j
e
r
t
v
9
i
s
0
9
0
3
8
2
.
[
1
9
]
A
.
A
.
F
a
h
a
d
,
H
.
J.
H
a
ssa
n
,
a
n
d
S
.
H
.
A
b
d
u
l
l
a
h
,
“
D
e
e
p
L
e
a
r
n
i
n
g
-
b
a
s
e
d
D
e
a
f
&
M
u
t
e
G
e
st
u
r
e
Tr
a
n
s
l
a
t
i
o
n
S
y
s
t
e
m,”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
S
c
i
e
n
c
e
a
n
d
Re
s
e
a
r
c
h
(
I
J
S
R)
,
v
o
l
.
9
,
n
o
.
5
,
p
p
.
2
8
8
–
2
9
2
,
2
0
2
0
,
d
o
i
:
1
0
.
2
1
2
7
5
/
S
R
2
0
5
0
3
0
3
1
8
0
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
Dev
elo
p
men
t sma
r
t e
ye
g
la
s
s
es
fo
r
visu
a
lly
imp
a
ir
ed
p
eo
p
le
b
a
s
ed
o
n
… (
Ha
s
s
a
n
S
a
la
m
A
b
d
u
l
-
A
mee
r
)
117
[
2
0
]
A
.
A
.
A
b
d
u
l
h
u
s
sei
n
a
n
d
F
.
A
.
R
a
h
e
e
m,
“
H
a
n
d
G
e
s
t
u
r
e
R
e
c
o
g
n
i
t
i
o
n
o
f
S
t
a
t
i
c
Le
t
t
e
r
s
A
meri
c
a
n
S
i
g
n
L
a
n
g
u
a
g
e
(
A
S
L)
U
si
n
g
D
e
e
p
Le
a
r
n
i
n
g
,
”
En
g
i
n
e
e
r
i
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
J
o
u
r
n
a
l
,
v
o
l
.
3
8
,
n
o
.
6
,
p
p
.
9
2
6
–
9
3
7
,
2
0
2
0
,
d
o
i
:
1
0
.
3
0
6
8
4
/
e
t
j
.
v
3
8
i
6
a
.
5
3
3
.
[
2
1
]
S
.
S
.
M
a
h
m
o
o
d
a
n
d
L
.
J
.
S
a
u
d
,
“
A
n
Ef
f
i
c
i
e
n
t
A
p
p
r
o
a
c
h
f
o
r
D
e
t
e
c
t
i
n
g
a
n
d
C
l
a
ssi
f
y
i
n
g
M
o
v
i
n
g
V
e
h
i
c
l
e
s
i
n
a
V
i
d
e
o
B
a
se
d
M
o
n
i
t
o
r
i
n
g
S
y
s
t
e
m
,
”
En
g
i
n
e
e
ri
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
J
o
u
r
n
a
l
,
v
o
l
.
3
8
,
n
o
.
6
A
,
p
p
.
8
3
2
–
8
4
5
,
2
0
2
0
,
d
o
i
:
1
0
.
3
0
6
8
4
/
e
t
j
.
v
3
8
i
6
a
.
4
3
8
.
[
2
2
]
I
.
A
.
M
.
Zi
n
,
Z
.
I
b
r
a
h
i
m,
D
.
I
sa,
S
.
A
l
i
ma
n
,
N
.
S
a
b
r
i
,
a
n
d
N
.
N
.
A
.
M
a
n
g
sh
o
r
,
“
H
e
r
b
a
l
p
l
a
n
t
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
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
,
”
B
u
l
l
e
t
i
n
o
f
El
e
c
t
r
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
a
n
d
I
n
f
o
rm
a
t
i
c
s
(
BEEI)
,
v
o
l
.
9
,
n
o
.
5
,
p
p
.
2
1
9
8
–
2
2
0
5
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
5
9
1
/
e
e
i
.
v
9
i
5
.
2
2
5
.
[
2
3
]
M
.
A
n
a
n
d
h
a
l
l
i
,
V
.
P
.
B
a
l
i
g
a
r
,
P
.
B
a
l
i
g
a
r
,
P
.
D
e
e
p
si
r
,
a
n
d
M
.
I
t
i
,
“
V
e
h
i
c
l
e
d
e
t
e
c
t
i
o
n
a
n
d
t
r
a
c
k
i
n
g
f
o
r
t
r
a
f
f
i
c
m
a
n
a
g
e
me
n
t
,
”
I
A
ES
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
(
I
J
-
AI
)
,
v
o
l
.
1
0
,
n
o
.
1
,
p
p
.
6
6
-
73
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
a
i
.
v
1
0
.
i
1
.
p
p
6
6
-
7
3
.
[
2
4
]
M
.
N
a
v
e
e
n
k
u
mar
a
n
d
V
.
A
y
y
a
s
a
my
,
“
O
p
e
n
C
V
f
o
r
C
o
m
p
u
t
e
r
V
i
s
i
o
n
A
p
p
l
i
c
a
t
i
o
n
s,
”
i
n
Pro
c
o
f
N
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
B
i
g
D
a
t
a
a
n
d
C
l
o
u
d
C
o
m
p
u
t
i
n
g
(
N
C
BD
C
’
1
5
)
,
2
0
1
5
,
p
p
.
5
2
–
5
6
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
r
e
se
a
r
c
h
g
a
t
e
.
n
e
t
/
p
u
b
l
i
c
a
t
i
o
n
/
3
0
1
5
9
0
5
7
1
_
O
p
e
n
C
V
_
f
o
r
_
C
o
mp
u
t
e
r
_
V
i
si
o
n
_
A
p
p
l
i
c
a
t
i
o
n
s
.
[
2
5
]
A
.
C
o
r
o
v
i
c
,
V
.
I
l
i
c
,
S
.
D
u
r
i
c
,
M
.
M
a
r
i
j
a
n
,
a
n
d
B
.
P
a
v
k
o
v
i
c
,
“
Th
e
R
e
a
l
-
Ti
me
D
e
t
e
c
t
i
o
n
o
f
Tr
a
f
f
i
c
P
a
r
t
i
c
i
p
a
n
t
s
U
s
i
n
g
Y
O
LO
A
l
g
o
r
i
t
h
m
,
”
i
n
Pr
o
c
.
2
0
1
8
2
6
t
h
T
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
F
o
r
u
m
,
T
ELFO
R
2
0
1
8
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
TELFO
R
.
2
0
1
8
.
8
6
1
1
9
8
6
.
[
2
6
]
M
.
B
u
r
i
ć
,
M
.
P
o
b
a
r
,
a
n
d
M
.
I
v
a
ši
ć
-
K
o
s,
“
A
d
a
p
t
i
n
g
y
o
l
o
n
e
t
w
o
r
k
f
o
r
b
a
l
l
a
n
d
p
l
a
y
e
r
d
e
t
e
c
t
i
o
n
,
”
i
n
Pr
o
c
.
o
f
t
h
e
8
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
P
a
t
t
e
r
n
R
e
c
o
g
n
i
t
i
o
n
Ap
p
l
i
c
a
t
i
o
n
s
a
n
d
Me
t
h
o
d
s
(
I
C
PRAM
2
0
1
9
)
,
2
0
1
9
p
p
.
8
4
5
–
8
5
1
,
d
o
i
:
1
0
.
5
2
2
0
/
0
0
0
7
5
8
2
0
0
8
4
5
0
8
5
1
.
[
2
7
]
Y
.
H
u
a
n
g
,
J.
Zh
e
n
g
,
S
.
S
u
n
,
C
.
Y
a
n
g
,
a
n
d
J.
Li
u
,
“
A
p
p
l
i
e
d
sc
i
e
n
c
e
s
O
p
t
i
m
i
z
e
d
Y
O
LO
v
3
A
l
g
o
r
i
t
h
m
a
n
d
I
t
s
A
p
p
l
i
c
a
t
i
o
n
i
n
Tr
a
f
f
i
c
F
l
o
w
D
e
t
e
c
t
i
o
n
s,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
2
0
2
0
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
H
a
ss
a
n
S
a
la
m
Abd
u
l
-
Am
e
e
r
re
c
e
iv
e
d
h
is
Ba
c
h
e
lo
r
o
f
C
o
m
p
u
ter
E
n
g
i
n
e
e
rin
g
fro
m
Un
iv
e
rsit
y
o
f
Tec
h
n
o
l
o
g
y
(
UO
T),
Ira
q
,
in
2
0
1
5
.
He
is
c
u
rre
n
tl
y
w
o
rk
i
n
g
o
n
h
is
M
a
ste
r
d
e
g
re
e
a
t
Un
i
v
e
rsity
o
f
Tec
h
n
o
l
o
g
y
.
His
a
re
a
o
f
i
n
tere
st
a
re
c
o
m
p
u
ter
v
isio
n
,
d
e
e
p
lea
r
n
i
n
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
c
e
.
1
9
.
0
6
@
g
ra
d
.
u
o
tec
h
n
o
l
o
g
y
.
e
d
u
.
i
q
As
st.
p
r
o
fe
ss
o
r
H
a
ss
a
n
J
.
H
a
ss
a
n
is
c
u
rre
n
tl
y
a
h
e
a
d
o
f
d
e
p
a
rtme
n
t
o
f
c
o
m
p
u
ter
e
n
g
in
e
e
rin
g
,
u
n
iv
e
rsit
y
o
f
tec
h
n
o
lo
g
y
(UO
T),
Ira
q
,
si
n
c
e
2
0
1
9
.
He
o
b
tain
e
d
h
is
P
h
D
i
n
c
o
m
p
u
ter
n
e
two
r
k
fr
o
m
c
o
n
tr
o
l
a
n
d
sy
ste
m
s
e
n
g
in
e
e
rin
g
d
e
p
a
rtme
n
t,
u
n
iv
e
rsit
y
o
f
tec
h
n
o
lo
g
y
in
2
0
1
3
.
His
re
se
a
rc
h
in
tere
st
in
c
lu
d
e
s,
wire
les
s
c
o
m
m
u
n
ica
ti
o
n
n
e
two
rk
s,
I
n
tern
e
t
o
f
th
in
g
s
(Io
T),
Ne
two
r
k
c
o
n
g
e
stio
n
c
o
n
tro
l,
Ad
v
a
n
c
e
d
ima
g
e
p
ro
c
e
ss
in
g
.
H
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
6
0
0
1
2
@
u
o
tec
h
n
o
lo
g
y
.
e
d
u
.
i
q
Dr
.
S
a
l
m
a
H
a
m
e
e
d
i
Ab
d
u
ll
a
h
is
c
u
rre
n
tl
y
a
lec
tu
re
r
a
t
t
h
e
fa
c
u
lt
y
o
f
C
o
m
p
u
ter
En
g
i
n
e
e
rin
g
De
p
a
rtme
n
t,
U
n
iv
e
rs
it
y
o
f
Tec
h
n
o
l
o
g
y
,
Ira
q
.
S
h
e
o
b
ta
in
e
d
h
e
r
P
h
.
D.
i
n
c
o
m
p
u
ter
v
isio
n
fro
m
th
e
Un
iv
e
rsit
y
o
f
T
e
c
h
n
o
l
o
g
y
in
2
0
1
8
.
He
r
re
se
a
rc
h
in
tere
st
in
c
lu
d
e
s
c
o
m
p
u
ter
v
isio
n
,
d
e
e
p
lea
rn
in
g
,
i
n
tern
e
t
o
f
t
h
in
g
s
(I
o
T),
a
n
d
p
a
tt
e
rn
re
c
o
g
n
it
i
o
n
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
1
2
0
0
1
5
@u
o
tec
h
n
o
lo
g
y
.
e
d
u
.
iq
Evaluation Warning : The document was created with Spire.PDF for Python.