Co
m
pu
t
er
Science
a
nd
I
nfo
r
m
a
t
io
n T
ec
hn
o
lo
g
ies
Vo
l.
7
,
No
.
2
,
J
u
ly
20
26
,
p
p
.
2
2
4
~
2
3
0
I
SS
N:
2722
-
3
2
2
1
,
DOI
:
1
0
.
1
1
5
9
1
/csi
t
.
v
7
i
2
.
p
2
2
4
-
2
3
0
224
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
p
r
ime.
co
m/in
d
ex
.
p
h
p
/csi
t
A nov
el appro
a
ch
f
o
r re
a
l
-
time
traf
fic
sig
n
r
ec
o
g
n
i
ti
o
n
f
r
a
m
ew
o
r
k
K
s
ha
t
ra
pa
l Sing
h
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
a
n
d
En
g
i
n
e
e
r
i
n
g
,
K
C
C
I
n
st
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
a
n
d
M
a
n
a
g
e
me
n
t
,
G
r
e
a
t
e
r
N
o
i
d
a
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Dec
1
5
,
2
0
2
5
R
ev
is
ed
Ma
r
2
0
,
2
0
2
6
Acc
ep
ted
Ma
y
1
6
,
2
0
2
6
Traffic
sig
n
re
c
o
g
n
it
io
n
p
lay
s
a
c
rit
ica
l
ro
le
in
e
n
h
a
n
c
in
g
ro
a
d
s
a
fe
ty
a
n
d
e
n
a
b
li
n
g
a
u
to
n
o
m
o
u
s
d
ri
v
i
n
g
sy
s
tem
s.
Th
is
p
a
p
e
r
p
re
se
n
ts
a
c
o
m
p
re
h
e
n
siv
e
a
p
p
ro
a
c
h
to
re
a
l
-
ti
m
e
traffic
si
g
n
re
c
o
g
n
i
ti
o
n
u
sin
g
a
d
v
a
n
c
e
d
c
o
m
p
u
ter
v
isio
n
tec
h
n
i
q
u
e
s
a
n
d
m
a
c
h
i
n
e
lea
rn
i
n
g
m
o
d
e
ls.
Th
e
p
ro
p
o
se
d
sy
ste
m
e
m
p
lo
y
s
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
two
rk
s
(CNN
s)
f
o
r
a
c
c
u
ra
te
d
e
te
c
ti
o
n
a
n
d
c
las
sifica
ti
o
n
o
f
traffic
sig
n
s
u
n
d
e
r
d
i
v
e
rse
e
n
v
ir
o
n
m
e
n
tal
c
o
n
d
it
io
n
s,
in
c
lu
d
in
g
v
a
ry
i
n
g
l
ig
h
ti
n
g
,
we
a
th
e
r,
a
n
d
o
c
c
lu
sio
n
s.
Re
a
l
-
ti
m
e
p
ro
c
e
ss
in
g
is
a
c
h
i
e
v
e
d
th
r
o
u
g
h
t
h
e
in
teg
ra
ti
o
n
o
f
o
p
ti
m
ize
d
a
lg
o
rit
h
m
s
a
n
d
h
a
rd
wa
re
a
c
c
e
ler
a
ti
o
n
tec
h
n
iq
u
e
s,
e
n
su
ri
n
g
m
in
ima
l
late
n
c
y
a
n
d
h
ig
h
t
h
ro
u
g
h
p
u
t.
Ex
p
e
rime
n
tal
re
su
lt
s
d
e
m
o
n
stra
t
e
th
a
t
th
e
sy
ste
m
a
c
h
iev
e
s
sta
te
-
of
-
th
e
-
a
rt
p
e
rfo
rm
a
n
c
e
o
n
b
e
n
c
h
m
a
rk
d
a
ta
se
ts,
with
a
n
a
c
c
u
ra
c
y
o
f
o
v
e
r
9
5
%
a
n
d
a
re
c
o
g
n
it
io
n
sp
e
e
d
su
it
a
b
le
fo
r
re
a
l
-
wo
rld
a
p
p
li
c
a
ti
o
n
s.
Th
e
fin
d
i
n
g
s
u
n
d
e
rsc
o
re
th
e
p
o
te
n
ti
a
l
o
f
t
h
e
s
y
ste
m
to
imp
ro
v
e
d
r
iv
e
r
a
ss
istan
c
e
sy
ste
m
s
a
n
d
p
a
v
e
th
e
wa
y
fo
r
sa
fe
r
a
u
t
o
n
o
m
o
u
s v
e
h
icle
s.
K
ey
w
o
r
d
s
:
Ar
tific
ial
in
tellig
en
ce
A
u
to
n
o
m
o
u
s
v
eh
icles a
n
d
ad
v
an
ce
d
d
r
iv
e
r
-
ass
is
tan
ce
s
y
s
tem
s
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
Occ
lu
s
io
n
s
T
r
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Ksh
atr
ap
al
Sin
g
h
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
KC
C
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
an
d
Ma
n
ag
em
en
t
Gr
ea
ter
No
id
a,
I
n
d
ia
E
m
ail:
m
ek
p
s
in
g
h
1
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
r
af
f
ic
s
ig
n
s
p
la
y
a
v
ital
r
o
le
i
n
r
eg
u
latin
g
,
war
n
in
g
,
an
d
g
u
i
d
in
g
d
r
iv
er
s
to
en
s
u
r
e
s
af
e
an
d
ef
f
icien
t
tr
an
s
p
o
r
tatio
n
.
R
ec
o
g
n
izin
g
a
n
d
in
ter
p
r
etin
g
th
ese
s
ig
n
s
ac
cu
r
ately
is
ess
en
tial
f
o
r
b
o
th
h
u
m
an
d
r
iv
e
r
s
an
d
au
to
n
o
m
o
u
s
v
eh
icles.
W
ith
th
e
r
ap
id
ad
v
an
ce
m
e
n
ts
in
a
r
tif
icial
in
tellig
en
ce
an
d
c
o
m
p
u
t
er
v
is
io
n
,
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
h
as
em
er
g
ed
as
a
k
ey
ar
ea
of
r
esear
ch
in
in
tellig
en
t
tr
an
s
p
o
r
tatio
n
s
y
s
tem
s
[
1
]
.
T
h
is
tech
n
o
lo
g
y
h
as
s
ig
n
if
ican
t
im
p
licatio
n
s
f
o
r
en
h
a
n
cin
g
r
o
ad
s
af
ety
,
r
ed
u
cin
g
ac
cid
en
ts
,
an
d
en
ab
lin
g
f
u
lly
au
to
n
o
m
o
u
s
d
r
iv
i
n
g
s
y
s
tem
s
.
Desp
ite
co
n
s
id
er
ab
le
p
r
o
g
r
ess
in
th
e
f
ield
,
d
ev
elo
p
in
g
a
r
o
b
u
s
t
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
s
y
s
tem
r
em
ain
s
a
ch
allen
g
in
g
task
.
Facto
r
s
s
u
ch
as
v
ar
y
in
g
li
g
h
tin
g
co
n
d
itio
n
s
,
wea
th
er
ch
an
g
es,
p
ar
tial
o
cc
lu
s
io
n
s
,
m
o
tio
n
b
lu
r
,
a
n
d
th
e
d
iv
er
s
ity
o
f
tr
af
f
ic
s
ig
n
d
esig
n
s
p
o
s
e
s
ig
n
if
ican
t
h
u
r
d
les
[
2
]
,
[
3
]
.
Fu
r
th
er
m
o
r
e
,
ac
h
iev
in
g
h
ig
h
ac
cu
r
ac
y
a
n
d
p
r
o
ce
s
s
in
g
s
p
ee
d
s
s
im
u
ltan
eo
u
s
ly
is
cr
itical
f
o
r
p
r
ac
tical
d
ep
lo
y
m
e
n
t
in
r
ea
l
-
wo
r
l
d
s
ce
n
ar
io
s
,
esp
ec
ially
in
tim
e
-
s
en
s
itiv
e
en
v
ir
o
n
m
en
ts
lik
e
au
to
n
o
m
o
u
s
v
eh
icles
an
d
ad
v
an
ce
d
d
r
iv
e
r
-
ass
is
tan
ce
s
y
s
tem
s
(
ADAS)
.
B
y
p
u
ttin
g
o
u
t
a
th
o
r
o
u
g
h
f
r
am
ewo
r
k
f
o
r
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
th
at
i
n
c
o
r
p
o
r
ates
cu
ttin
g
-
ed
g
e
d
ee
p
lear
n
in
g
m
o
d
els
an
d
o
p
tim
al
p
r
o
ce
s
s
in
g
m
eth
o
d
s
,
th
is
r
esear
ch
s
tu
d
y
s
ee
k
s
to
o
v
er
c
o
m
e
th
ese
is
s
u
es.
T
o
g
u
ar
an
tee
r
ea
l
-
tim
e
p
er
f
o
r
m
an
ce
,
t
h
e
s
u
g
g
e
s
ted
s
y
s
tem
u
s
es
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NNs)
f
o
r
ef
f
icien
t
f
ea
tu
r
e
e
x
tr
ac
tio
n
ac
ce
ler
atio
n
tec
h
n
iq
u
es.
Nu
m
er
o
u
s
test
s
o
n
r
ea
l
-
wo
r
ld
s
ce
n
ar
io
s
an
d
b
en
ch
m
ar
k
d
atasets
s
h
o
w
th
at
th
e
s
y
s
tem
ca
n
ac
h
iev
e
l
o
w
laten
cy
a
n
d
h
ig
h
ac
c
u
r
a
cy
,
wh
ich
m
ak
es
it
ap
p
r
o
p
r
iate
f
o
r
u
s
e
in
co
n
tem
p
o
r
ar
y
tr
a
n
s
p
o
r
tatio
n
s
y
s
tem
s
[
4
]
,
[
5
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
C
o
m
p
u
t Sci
I
n
f
T
ec
h
n
o
l
I
SS
N:
2722
-
3
2
2
1
A
n
o
ve
l a
p
p
r
o
a
c
h
fo
r
r
ea
l
-
time
tr
a
ffic sig
n
r
ec
o
g
n
itio
n
fr
a
mewo
r
k
(
K
s
h
a
tr
a
p
a
l S
in
g
h
)
225
T
h
e
r
est
o
f
th
e
p
ap
er
is
o
r
g
an
ized
as
f
o
llo
ws:
se
ctio
n
2
r
ev
iews
r
elate
d
wo
r
k
i
n
t
r
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
a
n
d
its
ap
p
licatio
n
s
.
Sectio
n
3
d
etails
th
e
m
eth
o
d
o
lo
g
y
,
in
cl
u
d
in
g
d
ataset
p
r
ep
ar
atio
n
,
m
o
d
el
ar
ch
itectu
r
e,
an
d
o
p
tim
izatio
n
s
tr
ateg
ies.
Sectio
n
4
p
r
esen
ts
th
e
ex
p
er
im
e
n
tal
r
esu
lts
an
d
s
ec
tio
n
5
co
n
t
ain
s
d
is
cu
s
s
io
n
s
.
Fin
ally
,
s
ec
tio
n
6
co
n
clu
d
es
th
e
p
ap
er
a
n
d
o
u
tlin
es
p
o
ten
tial
d
ir
ec
tio
n
s
f
o
r
f
u
t
u
r
e
r
esear
ch
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
e
f
i
e
l
d
of
t
r
a
f
f
i
c
s
i
g
n
r
e
c
o
g
n
i
t
i
o
n
h
a
s
g
a
r
n
e
r
e
d
s
ig
n
if
ican
t
atten
tio
n
in
r
ec
en
t
y
ea
r
s
d
u
e
to
it
i
s
v
ital
r
o
le
in
in
tellig
en
t
tr
an
s
p
o
r
tat
io
n
s
y
s
tem
s
an
d
au
to
n
o
m
o
u
s
v
eh
icles.
R
esear
ch
er
s
h
av
e
ex
p
lo
r
ed
v
ar
io
u
s
ap
p
r
o
ac
h
es,
r
a
n
g
in
g
f
r
o
m
t
r
ad
itio
n
al
im
a
g
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
to
m
o
d
er
n
d
e
ep
lear
n
i
n
g
-
b
ased
f
r
am
ewo
r
k
s
,
ea
c
h
co
n
tr
ib
u
tin
g
u
n
iq
u
e
s
o
lu
tio
n
s
to
th
e
c
h
allen
g
es in
th
is
d
o
m
ai
n
.
2
.
1
.
O
v
er
v
iew
o
f
ex
is
t
ing
t
ra
f
f
ic
s
ig
n r
ec
o
g
nitio
n sy
s
t
em
s
a
nd
t
heir
lim
it
a
t
io
ns
T
r
a
f
f
i
c
s
i
g
n
r
e
c
o
g
n
i
t
i
o
n
(
T
SR
)
s
y
s
t
e
m
s
h
a
v
e
a
d
v
a
n
c
e
d
s
i
g
n
i
f
i
c
a
n
t
l
y
,
e
v
o
l
v
i
n
g
f
r
o
m
r
u
l
e
-
b
a
s
e
d
a
p
p
r
o
a
c
h
e
s
u
s
i
n
g
h
a
n
d
c
r
a
f
t
e
d
f
e
a
t
u
r
e
s
t
o
m
o
d
e
r
n
m
a
c
h
i
n
e
l
ea
r
n
i
n
g
-
b
a
s
e
d
m
e
t
h
o
d
s
.
E
a
r
l
y
s
y
s
t
e
m
s
,
r
el
y
i
n
g
o
n
s
h
a
p
e
d
e
te
c
t
i
o
n
a
n
d
c
o
l
o
r
s
e
g
m
e
n
t
a
t
i
o
n
,
s
t
r
u
g
g
l
e
d
w
i
t
h
e
n
v
i
r
o
n
m
e
n
t
a
l
c
h
al
l
e
n
g
es
li
k
e
li
g
h
t
i
n
g
v
a
r
i
a
ti
o
n
s
a
n
d
o
c
c
l
u
s
i
o
n
s
[
6
]
.
W
h
i
l
e
m
o
d
e
r
n
s
y
s
t
e
m
s
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
h
a
v
e
i
m
p
r
o
v
e
d
a
c
c
u
r
a
c
y
,
t
h
e
y
f
a
c
e
li
m
i
t
at
i
o
n
s
i
n
r
e
a
l
-
t
i
m
e
p
e
r
f
o
r
m
a
n
c
e
,
d
a
ta
s
et
g
e
n
e
r
a
l
i
z
a
ti
o
n
,
a
n
d
r
o
b
u
s
t
n
e
s
s
u
n
d
e
r
a
d
v
e
r
s
e
c
o
n
d
i
t
i
o
n
s
.
2
.
2
.
Co
m
pa
riso
n o
f
t
ra
ditio
na
l im
a
g
e
pro
ce
s
s
ing
t
ec
hn
iqu
es
v
s
.
m
o
dern
deep
lea
rni
ng
a
pp
ro
a
ches
T
r
ad
itio
n
al
m
eth
o
d
s
,
in
clu
d
in
g
ed
g
e
d
etec
tio
n
an
d
h
o
u
g
h
tr
an
s
f
o
r
m
,
o
f
f
er
e
d
co
m
p
u
tatio
n
a
l
s
im
p
licity
b
u
t
lack
ed
r
o
b
u
s
t
n
e
s
s
ag
ain
s
t
r
ea
l
-
wo
r
ld
v
ar
iatio
n
s
.
Dee
p
lear
n
i
n
g
m
o
d
els,
p
ar
ticu
lar
ly
C
NNs,
h
av
e
r
ev
o
lu
tio
n
ized
T
SR
b
y
en
ab
lin
g
a
u
to
m
ated
f
ea
tu
r
e
ex
tr
ac
tio
n
an
d
i
m
p
r
o
v
e
d
ac
cu
r
ac
y
[
7
]
,
[
8
]
.
Ho
wev
er
,
th
ese
m
o
d
els
d
em
an
d
s
ig
n
if
ican
t
co
m
p
u
tatio
n
al
re
s
o
u
r
ce
s
,
m
ak
i
n
g
r
ea
l
-
ti
m
e
im
p
lem
en
tatio
n
ch
allen
g
in
g
.
2
.
3
.
Rec
ent
a
dv
a
ncem
ent
s
i
n r
ea
l
-
t
im
e
o
bje
c
t
det
ec
t
io
n
a
nd
cla
s
s
if
ica
t
io
n
R
ec
en
t
r
esear
ch
h
as
f
o
cu
s
ed
o
n
r
ea
l
-
tim
e
T
SR
s
y
s
tem
s
ca
p
ab
le
o
f
p
r
o
ce
s
s
in
g
liv
e
v
id
e
o
s
tr
ea
m
s
with
m
in
im
al
laten
cy
.
Ob
jec
t
d
etec
tio
n
f
r
am
ewo
r
k
s
s
u
ch
as
y
o
u
o
n
ly
lo
o
k
o
n
ce
(
Y
OL
O)
,
Sin
g
le
Sh
o
t
Mu
ltiB
o
x
d
etec
to
r
(
SS
D)
,
an
d
f
aster
R
-
C
NN
h
av
e
b
ee
n
wid
ely
ad
o
p
ted
f
o
r
tr
af
f
ic
s
ig
n
d
etec
tio
n
an
d
class
if
icatio
n
.
YOL
O,
in
p
ar
ticu
lar
,
is
f
av
o
u
r
e
d
f
o
r
its
h
ig
h
-
s
p
ee
d
p
er
f
o
r
m
an
ce
,
m
ak
i
n
g
it
s
u
itab
le
f
o
r
r
ea
l
-
tim
e
ap
p
licatio
n
s
[
6
]
.
T
o
o
p
tim
ize
d
ee
p
lear
n
in
g
m
o
d
els
f
o
r
r
ea
l
-
tim
e
d
ep
l
o
y
m
e
n
t,
r
esear
ch
e
r
s
h
av
e
ex
p
lo
r
ed
lig
h
tweig
h
t
ar
ch
itectu
r
es
s
u
ch
as
tin
y
Y
OL
O
an
d
Mo
b
ile
-
Net.
T
h
ese
m
o
d
els
ac
h
iev
e
a
b
a
lan
ce
b
etwe
en
s
p
ee
d
an
d
ac
cu
r
ac
y
,
en
a
b
lin
g
d
e
p
lo
y
m
e
n
t
o
n
r
eso
u
r
ce
-
co
n
s
tr
ain
ed
d
e
v
ices
lik
e
ed
g
e
p
r
o
ce
s
s
o
r
s
an
d
em
b
ed
d
e
d
s
y
s
tem
s
[9
]
−
[
1
1
]
.
Mo
r
eo
v
e
r
,
h
ar
d
war
e
ac
ce
ler
atio
n
u
s
in
g
GPUs
o
r
s
p
ec
ialized
h
ar
d
war
e
s
u
ch
as
ten
s
o
r
p
r
o
ce
s
s
in
g
u
n
its
(
T
PUs
)
h
as
f
u
r
th
er
e
n
h
a
n
ce
d
th
e
f
ea
s
ib
ilit
y
o
f
r
ea
l
-
tim
e
T
SR
s
y
s
tem
.
2
.
4
.
I
dentif
ied
re
s
ea
rc
h g
a
ps
Desp
ite
th
ese
ad
v
an
ce
m
en
ts
,
s
ev
er
al
r
esear
ch
g
ap
s
r
em
ain
i
n
th
e
f
ield
o
f
r
ea
l
-
tim
e
T
SR
.
−
Data
s
et
d
iv
er
s
ity
:
m
o
s
t
m
o
d
el
s
ar
e
tr
ain
ed
o
n
b
en
ch
m
a
r
k
d
a
tasets
th
at
m
ay
n
o
t
ad
e
q
u
ately
r
ep
r
esen
t
r
ea
l
-
wo
r
ld
co
n
d
itio
n
s
,
in
clu
d
in
g
r
a
r
e
tr
af
f
ic
s
ig
n
s
an
d
r
eg
io
n
al
v
a
r
i
a
t
i
o
n
s
.
−
E
n
v
ir
o
n
m
en
tal
r
o
b
u
s
tn
ess
:
s
y
s
tem
s
s
till
s
tr
u
g
g
le
with
ad
v
er
s
e
co
n
d
itio
n
s
s
u
ch
as
lo
w
v
is
ib
ilit
y
,
g
lar
e
,
a
n
d
o
cc
lu
s
io
n
,
lim
itin
g
th
eir
r
ea
l
-
w
o
r
ld
ap
p
licab
ilit
y
.
−
E
f
f
icien
cy
v
s
.
ac
c
u
r
ac
y
tr
a
d
e
-
o
f
f
:
ac
h
iev
in
g
r
ea
l
-
tim
e
p
er
f
o
r
m
an
ce
with
o
u
t
s
ac
r
if
icin
g
ac
cu
r
ac
y
r
em
ai
n
s
a
cr
itical
ch
allen
g
e.
Ma
n
y
m
o
d
els co
m
p
r
o
m
is
e
ac
cu
r
ac
y
to
m
ee
t sp
ee
d
r
eq
u
ir
em
en
ts
.
−
Scalab
ilit
y
:
i
n
teg
r
atin
g
T
SR
s
y
s
tem
s
in
to
lar
g
e
-
s
ca
le
in
tellig
en
t
tr
an
s
p
o
r
tatio
n
n
etwo
r
k
s
r
eq
u
ir
es
f
u
r
th
er
in
n
o
v
atio
n
in
ter
m
s
o
f
s
ca
lab
ilit
y
an
d
in
ter
o
p
er
a
b
ilit
y
.
3.
P
RO
P
O
SE
D
M
E
T
H
O
DO
L
O
G
Y
T
h
is
s
ec
tio
n
d
escr
ib
es
th
e
s
y
s
tem
atic
ap
p
r
o
ac
h
ad
o
p
ted
f
o
r
d
ev
elo
p
in
g
an
d
ev
alu
atin
g
th
e
p
r
o
p
o
s
ed
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
iti
o
n
s
y
s
tem
.
T
h
e
m
et
h
o
d
o
lo
g
y
in
clu
d
es
d
ataset
p
r
ep
ar
atio
n
,
m
o
d
el
ar
ch
itectu
r
e
d
esig
n
,
tr
ain
in
g
s
tr
ateg
ies,
an
d
r
ea
l
-
tim
e
im
p
lem
en
tatio
n
.
3
.
1
.
Sy
s
t
e
m
o
v
er
v
iew
T
h
e
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
s
y
s
tem
c
o
n
s
is
ts
o
f
two
m
ain
c
o
m
p
o
n
en
ts
:
i)
d
etec
tio
n
:
lo
ca
tin
g
tr
af
f
ic
s
ig
n
s
in
an
in
p
u
t
im
ag
e
o
r
v
id
eo
f
r
a
m
e
;
an
d
ii)
c
lass
if
icatio
n
:
id
en
tify
in
g
th
e
ty
p
e
o
f
tr
af
f
ic
s
ig
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
3
2
2
1
C
o
m
p
u
t Sci
I
n
f
T
ec
h
n
o
l
,
Vo
l.
7
,
No
.
2
,
J
u
ly
20
26
:
224
-
2
3
0
226
d
etec
ted
.
T
h
e
f
r
am
ewo
r
k
is
b
u
ilt
u
s
in
g
a
d
e
ep
lear
n
in
g
-
b
ased
o
b
ject
d
etec
tio
n
m
o
d
e
l,
wi
th
a
f
o
cu
s
on
b
alan
cin
g
ac
c
u
r
ac
y
a
n
d
s
p
ee
d
as sh
o
wn
in
Fig
u
r
e
1
.
Fig
u
r
e
1
.
R
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
s
y
s
tem
3
.
2
.
Da
t
a
s
et
prepa
ra
t
io
n
T
h
e
s
y
s
tem
was
tr
ain
ed
an
d
te
s
ted
on
p
u
b
licly
av
ailab
le
d
atasets
s
u
ch
as:
−
Ger
m
an
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
b
en
c
h
m
ar
k
(
GT
SR
B
)
:
p
r
o
v
id
es
lab
eled
tr
af
f
ic
s
ig
n
im
ag
es
f
o
r
class
if
icatio
n
.
−
T
s
in
g
h
u
a
-
ten
ce
n
t d
ataset: c
o
n
t
ain
s
r
ea
l
-
wo
r
ld
tr
a
f
f
ic
s
ig
n
im
ag
es f
o
r
d
etec
tio
n
task
s
.
Step
s
:
−
Data
au
g
m
en
tatio
n
:
tech
n
i
q
u
e
s
lik
e
r
o
tatio
n
,
s
ca
lin
g
,
b
r
ig
h
t
n
ess
ad
ju
s
tm
en
t,
an
d
f
lip
p
in
g
wer
e
ap
p
lied
to
in
cr
ea
s
e
d
ataset
d
iv
er
s
ity
an
d
r
o
b
u
s
tn
ess
.
−
Pre
p
r
o
ce
s
s
in
g
:
im
ag
es
wer
e
r
e
s
ized
to
2
2
4
×
2
2
4
,
2
2
4
tim
es
2
2
4
,
2
2
4
×2
2
4
p
ix
els
an
d
n
o
r
m
al
ized
to
en
h
a
n
ce
m
o
d
el
p
er
f
o
r
m
a
n
ce
.
3
.
3
.
M
o
del
a
rc
hite
ct
ure
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
em
p
lo
y
s
a
m
o
d
if
ied
v
e
r
s
io
n
o
f
t
h
e
Y
OL
Ov
5
o
b
ject
d
etec
tio
n
m
o
d
el
f
o
r
r
ea
l
-
tim
e
d
etec
tio
n
an
d
class
if
icati
o
n
o
f
tr
af
f
ic
s
ig
n
s
,
as
s
ee
n
in
F
ig
u
r
e
s
2
an
d
3
[
1
2
]
−
[
1
4
].
Fig
u
r
e
2
.
Mo
d
el
d
em
o
Evaluation Warning : The document was created with Spire.PDF for Python.
C
o
m
p
u
t Sci
I
n
f
T
ec
h
n
o
l
I
SS
N:
2722
-
3
2
2
1
A
n
o
ve
l a
p
p
r
o
a
c
h
fo
r
r
ea
l
-
time
tr
a
ffic sig
n
r
ec
o
g
n
itio
n
fr
a
mewo
r
k
(
K
s
h
a
tr
a
p
a
l S
in
g
h
)
227
Fig
u
r
e
3
.
Mo
d
el
f
lo
w
d
iag
r
am
−
B
ac
k
b
o
n
e:
C
SP
Dar
k
n
et
f
o
r
f
e
atu
r
e
ex
tr
ac
tio
n
.
−
Nec
k
: Path
ag
g
r
eg
atio
n
n
etwo
r
k
(
PANet)
to
en
h
an
ce
f
ea
t
u
r
e
p
r
o
p
a
g
atio
n
.
−
Hea
d
: Bo
u
n
d
in
g
b
o
x
r
eg
r
ess
io
n
an
d
class
if
icatio
n
lay
er
s
.
3
.
4
.
T
ra
ini
ng
a
nd
o
ptim
iza
t
i
o
n
−
T
r
ain
in
g
p
ar
am
eter
s
:
a.
L
ea
r
n
in
g
r
ate:
0
.
0
0
1
0
.
0
0
1
0
.
0
0
1
b.
B
atch
s
ize:
3
2
c.
E
p
o
ch
s
: 5
0
−
L
o
s
s
f
u
n
ctio
n
: a
c
o
m
b
in
atio
n
o
f
o
b
jectn
ess
,
class
if
icatio
n
,
a
n
d
lo
ca
lizatio
n
lo
s
s
was u
s
ed
.
−
Op
tim
izatio
n
tech
n
iq
u
es
:
a.
T
r
an
s
f
er
lear
n
i
n
g
u
s
in
g
a
p
r
e
-
tr
ain
ed
YOL
Ov
5
m
o
d
el.
b.
R
eg
u
lar
izatio
n
tech
n
iq
u
es su
ch
as d
r
o
p
o
u
t a
n
d
weig
h
t
d
ec
ay
to
r
ed
u
ce
o
v
e
r
-
f
itti
n
g
.
3
.
5
.
Rea
l
-
t
im
e
i
m
plem
ent
a
t
i
o
n
T
h
e
tr
ain
ed
m
o
d
el
was
d
ep
lo
y
ed
on
a
h
ar
d
war
e
p
latf
o
r
m
o
p
t
im
ized
f
o
r
r
ea
l
-
tim
e
p
er
f
o
r
m
a
n
ce
,
s
u
ch
as NV
I
DI
A
J
etso
n
Nan
o
o
r
a
GPU
-
en
ab
led
s
y
s
tem
.
T
h
e
im
p
lem
en
tatio
n
p
i
p
elin
e
in
clu
d
es
[
1
5
]
,
[
1
4
]:
−
I
n
p
u
t
h
an
d
lin
g
: c
a
p
tu
r
i
n
g
liv
e
v
id
eo
f
r
am
es u
s
in
g
Op
en
C
V.
−
I
n
f
er
en
ce
:
p
r
o
ce
s
s
in
g
f
r
am
es th
r
o
u
g
h
th
e
t
r
ain
ed
m
o
d
el.
−
Ou
tp
u
t
v
is
u
aliza
tio
n
: a
n
n
o
tate
r
ec
o
g
n
ize
d
tr
af
f
ic
s
ig
n
s
with
b
o
u
n
d
in
g
b
o
x
es a
n
d
lab
els in
r
ea
l
-
tim
e.
3
.
6
.
P
er
f
o
r
m
a
nce
ev
a
lua
t
io
n
T
h
e
s
y
s
tem
was
ev
alu
ated
b
as
ed
on
[
12
]
,
[
1
6
]
:
−
Acc
u
r
ac
y
:
m
ea
s
u
r
ed
o
n
b
en
ch
m
ar
k
d
atasets
.
−
Sp
ee
d
:
f
r
am
es
p
er
s
ec
o
n
d
(
FP
S)
d
u
r
in
g
r
ea
l
-
tim
e
in
f
e
r
en
ce
.
−
R
o
b
u
s
tn
ess
:
test
ed
u
n
d
er
v
ar
y
i
n
g
co
n
d
itio
n
s
,
in
clu
d
in
g
p
o
o
r
lig
h
tin
g
,
o
cc
l
u
s
io
n
,
an
d
m
o
tio
n
b
lu
r
.
4.
E
XP
E
R
I
M
E
N
T
A
L
RE
SUL
T
S
T
h
is
s
ec
tio
n
p
r
esen
ts
th
e
ev
alu
atio
n
of
th
e
p
r
o
p
o
s
ed
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
s
y
s
tem
,
in
clu
d
in
g
p
er
f
o
r
m
a
n
ce
m
etr
ic
s
,
r
esu
lts
o
n
b
en
ch
m
ar
k
d
atasets
,
r
ea
l
-
wo
r
ld
test
in
g
,
an
d
c
o
m
p
ar
ativ
e
a
n
aly
s
is
with
ex
is
tin
g
ap
p
r
o
ac
h
es.
4
.
1
.
P
er
f
o
r
m
a
nce
m
e
t
rics
T
h
e
s
y
s
tem
’
s
p
er
f
o
r
m
an
ce
wa
s
ev
alu
ated
u
s
in
g
th
e
f
o
llo
win
g
m
etr
ics [
1
7
]:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
3
2
2
1
C
o
m
p
u
t Sci
I
n
f
T
ec
h
n
o
l
,
Vo
l.
7
,
No
.
2
,
J
u
ly
20
26
:
224
-
2
3
0
228
−
Acc
u
r
ac
y
:
th
e
p
e
r
ce
n
tag
e
o
f
c
o
r
r
ec
tly
class
if
ied
tr
af
f
ic
s
ig
n
s
.
−
Pre
cisi
o
n
:
th
e
r
atio
o
f
c
o
r
r
ec
tl
y
p
r
ed
icted
p
o
s
itiv
e
o
b
s
er
v
atio
n
s
to
to
tal
p
r
e
d
icted
p
o
s
itiv
es.
−
R
ec
all:
th
e
r
atio
o
f
co
r
r
ec
tly
p
r
ed
icted
p
o
s
itiv
e
o
b
s
er
v
atio
n
s
to
all
ac
tu
al
p
o
s
itiv
es.
−
F1
-
Sco
r
e:
th
e
h
ar
m
o
n
ic
m
ea
n
o
f
p
r
ec
is
io
n
a
n
d
r
ec
all.
−
L
aten
cy
:
av
er
a
g
e
tim
e
tak
en
t
o
p
r
o
ce
s
s
ea
ch
f
r
am
e
(
m
ea
s
u
r
ed
in
m
illi
s
ec
o
n
d
s
)
.
−
FPS:
a
m
ea
s
u
r
e
o
f
r
ea
l
-
tim
e
s
y
s
tem
p
er
f
o
r
m
an
ce
.
4
.
2
.
Resul
t
s
o
n
benchm
a
rk
da
t
a
s
et
s
T
h
e
s
y
s
tem
was
tes
ted
on
th
e
GT
SR
B
an
d
T
s
in
g
h
u
a
-
T
en
ce
n
t T
r
af
f
ic
Sig
n
Data
s
et.
Key
r
e
s
u
l
t
s
:
−
G
T
SR
B
:
a.
Acc
u
r
ac
y
: 9
7
.
3
%
b.
Pre
cisi
o
n
: 9
6
.
8
%
c.
R
ec
all: 9
7
.
5
%
d.
F1
-
Sco
r
e:
9
7
.
1
%
−
T
s
i
n
g
h
u
a
-
T
e
n
c
e
n
t
:
a.
Acc
u
r
ac
y
: 9
5
.
6
%
b.
Pre
cisi
o
n
: 9
4
.
9
%
c.
R
ec
all: 9
5
.
8
%
d.
F1
-
Sco
r
e:
9
5
.
3
%
4
.
3
.
Rea
l
-
wo
rld t
estin
g
T
h
e
s
y
s
tem
was
d
ep
lo
y
ed
in
a
r
ea
l
wo
r
ld
s
ettin
g
u
s
in
g
li
v
e
v
id
e
o
f
ee
d
s
ca
p
tu
r
e
d
u
n
d
er
v
ar
io
u
s
co
n
d
itio
n
s
:
−
Day
lig
h
t:
ac
h
iev
ed
c
o
n
s
is
ten
t d
etec
tio
n
with
an
ac
c
u
r
ac
y
o
f
9
6
.
5
%.
−
Nig
h
ttime
:
ac
cu
r
ac
y
d
r
o
p
p
ed
s
lig
h
tly
to
9
2
.
4
% d
u
e
to
r
ed
u
c
ed
v
is
ib
ilit
y
.
−
Ad
v
er
s
e
wea
th
er
:
p
er
f
o
r
m
a
n
c
e
u
n
d
er
r
ain
an
d
f
o
g
s
h
o
wed
an
ac
cu
r
ac
y
o
f
8
9
.
8
%,
h
ig
h
li
g
h
tin
g
ar
ea
s
f
o
r
im
p
r
o
v
em
e
n
t.
−
Occ
lu
s
io
n
an
d
m
o
tio
n
b
lu
r
:
th
e
s
y
s
tem
s
u
cc
ess
f
u
lly
id
en
tifie
d
p
ar
tiall
y
v
is
ib
le
s
ig
n
s
with
8
5
.
3
%
ac
cu
r
ac
y
,
b
u
t stru
g
g
le
d
in
ca
s
es o
f
s
ev
er
e
m
o
tio
n
b
lu
r
.
4
.
4
.
Co
m
pa
ra
t
iv
e
a
na
ly
s
is
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
was
co
m
p
ar
ed
with
s
tate
o
f
-
th
e
-
ar
t
m
eth
o
d
s
,
in
clu
d
in
g
YOL
Ov
4
,
SS
D,
an
d
Fas
ter
R
-
C
NN
as sh
o
wn
in
T
ab
le
1
.
T
ab
le
1
.
C
o
m
p
a
r
ati
v
e
an
aly
s
is
M
o
d
e
l
A
c
c
u
r
a
c
y
(
%)
FPS
La
t
e
n
c
y
(
ms)
R
o
b
u
st
n
e
ss
(
%)
Y
O
LO
v
4
9
5
.
4
25
40
8
8
.
5
SSD
9
4
.
2
20
50
8
6
.
3
F
a
st
e
r
R
-
C
N
N
9
6
.
1
15
65
8
7
.
9
P
r
o
p
o
se
d
9
7
.
3
30
33
8
9
.
8
5.
DIS
CU
SS
I
O
N
5
.
1
.
I
nte
rpre
t
a
t
io
n
o
f
re
s
ults a
nd
t
heir
im
pli
ca
t
io
ns
T
h
e
ex
p
er
im
en
tal
r
esu
lts
d
em
o
n
s
tr
ate
th
at
th
e
p
r
o
p
o
s
ed
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
it
io
n
s
y
s
tem
ac
h
iev
es h
ig
h
ac
cu
r
ac
y
an
d
s
p
ee
d
,
m
ak
in
g
it su
itab
le
f
o
r
r
ea
l
-
wo
r
ld
ap
p
licatio
n
s
s
u
ch
as a
u
to
n
o
m
o
u
s
v
eh
icles
an
d
ADAS
as
s
h
o
wn
in
Fig
u
r
e
4
.
T
h
e
s
y
s
tem
’
s
ab
ilit
y
to
p
r
o
ce
s
s
v
id
eo
f
r
am
es
at
3
0
FP
S
with
an
ac
c
u
r
ac
y
o
f
9
7
.
3
%
o
n
th
e
GT
SR
B
d
atase
t
h
ig
h
lig
h
ts
its
ef
f
ec
ti
v
en
ess
in
d
etec
tin
g
an
d
class
if
y
in
g
t
r
af
f
ic
s
ig
n
s
u
n
d
er
d
iv
er
s
e
c
o
n
d
i
t
i
o
n
s
.
5
.
2
.
Str
eng
t
hs
a
nd
lim
it
a
t
io
ns
T
h
e
s
tr
en
g
th
s
o
f
th
e
s
tu
d
y
ar
e
:
i)
h
ig
h
ac
cu
r
ac
y
an
d
r
ea
l
-
ti
m
e
p
r
o
c
e
s
s
i
n
g
;
ii)
lig
h
tweig
h
t
d
esig
n
f
o
r
d
ep
lo
y
m
e
n
t
o
n
em
b
e
d
d
ed
s
y
s
tem
s
; a
n
d
iii)
r
o
b
u
s
t
ag
ain
s
t
p
ar
tial
o
cc
lu
s
io
n
s
an
d
v
is
u
al
ly
s
im
ilar
s
ig
n
s
.
Fu
th
er
m
o
r
e,
th
er
e
a
r
e
s
ev
er
al
lim
itatio
n
s
o
f
th
e
s
tu
d
y
ar
e:
i)
p
er
f
o
r
m
an
ce
d
r
o
p
s
in
p
o
o
r
lig
h
tin
g
an
d
w
e
a
t
h
e
r
;
ii)
d
e
p
e
n
d
e
n
c
y
on
b
e
n
c
h
m
a
r
k
d
a
t
as
e
ts
lim
its
r
ea
l
-
wo
r
ld
g
en
er
aliza
b
ilit
y
;
an
d
iii)
s
tr
u
g
g
les
with
m
o
tio
n
b
lu
r
in
d
y
n
am
ic
s
c
e
n
a
r
i
o
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
C
o
m
p
u
t Sci
I
n
f
T
ec
h
n
o
l
I
SS
N:
2722
-
3
2
2
1
A
n
o
ve
l a
p
p
r
o
a
c
h
fo
r
r
ea
l
-
time
tr
a
ffic sig
n
r
ec
o
g
n
itio
n
fr
a
mewo
r
k
(
K
s
h
a
tr
a
p
a
l S
in
g
h
)
229
Fig
u
r
e
4
.
E
x
p
er
im
e
n
tal
r
esu
lts
5
.
3
.
P
o
t
ent
ia
l
a
re
a
s
f
o
r
im
pro
v
e
m
ent
Sev
er
al
p
o
ten
tial
im
p
r
o
v
em
en
ts
ar
e
id
en
tifie
d
as
f
o
llo
ws:
i)
i
m
p
r
o
v
e
p
r
ep
r
o
ce
s
s
in
g
f
o
r
lo
w
-
lig
h
t
an
d
ad
v
er
s
e
co
n
d
itio
n
s
;
ii)
e
x
p
an
d
d
atasets
to
in
clu
d
e
r
ar
e
a
n
d
d
iv
er
s
e
s
ig
n
s
;
iii)
a
d
d
r
ess
m
o
tio
n
b
lu
r
u
s
in
g
tem
p
o
r
al
d
ata
f
r
o
m
co
n
s
ec
u
t
iv
e
f
r
am
es
;
iv
)
o
p
tim
ize
m
o
d
els
f
o
r
b
etter
ef
f
icien
cy
o
n
ed
g
e
d
e
v
ices
;
an
d
v
)
e
x
p
l
o
r
e
m
u
lti
-
m
o
d
al
ap
p
r
o
a
ch
es c
o
m
b
in
in
g
ca
m
er
a
a
n
d
s
e
n
s
o
r
d
ata.
6.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
p
r
o
p
o
s
ed
an
d
ev
al
u
ated
a
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
s
y
s
tem
th
at
co
m
b
in
es
h
ig
h
ac
cu
r
ac
y
with
ef
f
icien
t
p
er
f
o
r
m
an
ce
.
T
h
e
s
y
s
tem
ac
h
iev
e
d
a
9
7
.
3
%
ac
cu
r
ac
y
r
ate
o
n
b
en
ch
m
ar
k
d
atasets
an
d
m
ain
tain
ed
r
ea
l
-
tim
e
p
r
o
ce
s
s
in
g
s
p
ee
d
s
o
f
3
0
FP
S,
d
em
o
n
s
tr
atin
g
its
ap
p
licab
ilit
y
in
d
y
n
am
ic
en
v
ir
o
n
m
en
ts
.
R
o
b
u
s
tn
ess
test
in
g
h
ig
h
lig
h
te
d
its
ef
f
ec
tiv
e
n
ess
in
d
a
y
lig
h
t
co
n
d
itio
n
s
,
th
o
u
g
h
ch
allen
g
es
r
em
ain
u
n
d
er
ad
v
er
s
e
lig
h
tin
g
an
d
wea
th
e
r
.
T
h
e
r
esear
c
h
co
n
tr
ib
u
tes
t
o
ad
v
a
n
cin
g
th
e
f
ield
b
y
:
i)
d
ev
el
o
p
in
g
a
s
ca
lab
le
an
d
lig
h
tweig
h
t
ar
ch
itectu
r
e
o
p
tim
ized
f
o
r
r
ea
l
-
tim
e
d
ep
lo
y
m
e
n
t
;
ii)
d
em
o
n
s
tr
atin
g
a
b
alan
ce
b
etw
ee
n
ac
cu
r
ac
y
an
d
ef
f
icien
cy
,
w
h
ich
is
cr
itical
f
o
r
r
eso
u
r
ce
-
c
o
n
s
tr
ain
ed
d
ev
i
ce
s
;
an
d
iii)
h
i
g
h
lig
h
tin
g
s
tr
ateg
ies
to
en
h
a
n
ce
m
o
d
el
r
o
b
u
s
tn
ess
,
in
clu
d
in
g
d
ata
au
g
m
en
tatio
n
an
d
a
r
ch
itect
u
r
e
o
p
tim
izatio
n
.
T
h
e
s
ig
n
if
ican
ce
o
f
r
ea
l
-
tim
e
tr
af
f
ic
s
ig
n
r
ec
o
g
n
itio
n
lies
in
it
i
s
ab
ilit
y
to
im
p
r
o
v
e
r
o
ad
s
af
ety
an
d
en
ab
le
au
to
n
o
m
o
u
s
n
a
v
ig
atio
n
.
W
h
ile
th
is
r
esear
ch
ac
h
iev
ed
p
r
o
m
is
in
g
r
esu
lts
,
it
also
id
e
n
tifie
d
k
ey
ar
ea
s
f
o
r
im
p
r
o
v
em
e
n
t,
s
u
c
h
as
h
a
n
d
lin
g
ad
v
er
s
e
co
n
d
itio
n
s
,
e
x
p
an
d
in
g
d
atasets
,
an
d
in
teg
r
atin
g
m
u
lti
-
m
o
d
al
ap
p
r
o
ac
h
es.
Fu
tu
r
e
ad
v
an
ce
m
en
ts
in
h
ar
d
war
e
ac
ce
ler
ati
o
n
,
m
o
d
el
o
p
tim
izatio
n
,
an
d
r
ea
l
-
wo
r
ld
d
ataset
co
llectio
n
ar
e
ex
p
ec
te
d
to
e
n
h
an
ce
th
e
r
elia
b
ilit
y
an
d
s
ca
lab
ilit
y
o
f
T
SR
s
y
s
tem
s
,
p
av
in
g
t
h
e
way
f
o
r
wid
esp
r
ea
d
ad
o
p
tio
n
in
in
tel
lig
en
t tr
an
s
p
o
r
tatio
n
n
etwo
r
k
s
a
n
d
s
m
ar
t c
ities
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
t
r
ib
u
to
r
R
o
les
T
a
x
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
i
d
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Ksh
atr
ap
al
Sin
g
h
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
3
2
2
1
C
o
m
p
u
t Sci
I
n
f
T
ec
h
n
o
l
,
Vo
l.
7
,
No
.
2
,
J
u
ly
20
26
:
224
-
2
3
0
230
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
t
h
at
s
u
p
p
o
r
t
t
h
e
f
in
d
in
g
s
o
f
th
is
s
tu
d
y
a
r
e
av
ailab
l
e
f
r
o
m
th
e
co
r
r
esp
o
n
d
in
g
a
u
th
o
r
,
[
KS
]
,
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
J.
S
t
a
l
l
k
a
m
p
,
M
.
S
c
h
l
i
p
s
i
n
g
,
J
.
S
a
l
m
e
n
,
a
n
d
C
.
I
g
e
l
,
“
M
a
n
v
s.
c
o
m
p
u
t
e
r
:
b
e
n
c
h
mar
k
i
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
f
o
r
t
r
a
f
f
i
c
si
g
n
r
e
c
o
g
n
i
t
i
o
n
,
”
N
e
u
r
a
l
N
e
t
w
o
r
k
s
,
v
o
l
.
3
2
,
p
p
.
3
2
3
–
3
3
2
,
2
0
1
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
n
e
t
.
2
0
1
2
.
0
2
.
0
1
6
.
[
2
]
J.
R
e
d
m
o
n
a
n
d
A
.
F
a
r
h
a
d
i
,
“
Y
O
LO
v
3
:
a
n
i
n
c
r
e
me
n
t
a
l
i
mp
r
o
v
e
m
e
n
t
,
”
a
r
Xi
v
p
re
p
r
i
n
t
a
rXi
v
:
1
8
0
4
.
0
2
7
6
7
,
A
p
r
.
2
0
1
8
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
a
r
x
i
v
.
o
r
g
/
a
b
s
/
1
8
0
4
.
0
2
7
6
7
[
3
]
Z.
Z
h
u
,
D
.
Li
a
n
g
,
S
.
Z
h
a
n
g
,
X
.
H
u
a
n
g
,
B
.
L
i
,
a
n
d
S
.
H
u
,
“
Tr
a
f
f
i
c
-
si
g
n
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ssi
f
i
c
a
t
i
o
n
i
n
t
h
e
w
i
l
d
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
I
EEE
C
o
m
p
u
t
e
r
S
o
c
i
e
t
y
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
V
i
si
o
n
a
n
d
Pa
t
t
e
rn
Re
c
o
g
n
i
t
i
o
n
,
2
0
1
6
,
p
p
.
2
1
1
0
–
2
1
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
.
2
0
1
6
.
2
3
2
.
[
4
]
K
.
H
e
,
X
.
Zh
a
n
g
,
S
.
R
e
n
,
a
n
d
J.
S
u
n
,
“
D
e
e
p
r
e
s
i
d
u
a
l
l
e
a
r
n
i
n
g
f
o
r
i
ma
g
e
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
I
EEE
C
o
m
p
u
t
e
r
S
o
c
i
e
t
y
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r V
i
si
o
n
a
n
d
P
a
t
t
e
r
n
Re
c
o
g
n
i
t
i
o
n
,
2
0
1
6
,
p
p
.
7
7
0
–
7
7
8
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
.
2
0
1
6
.
9
0
.
[
5
]
S
.
H
o
u
b
e
n
,
J.
S
t
a
l
l
k
a
m
p
,
J.
S
a
l
me
n
,
M
.
S
c
h
l
i
p
si
n
g
,
a
n
d
C
.
I
g
e
l
,
“
D
e
t
e
c
t
i
o
n
o
f
t
r
a
f
f
i
c
si
g
n
s i
n
r
e
a
l
-
w
o
r
l
d
i
m
a
g
e
s
:
t
h
e
g
e
r
m
a
n
t
r
a
f
f
i
c
si
g
n
d
e
t
e
c
t
i
o
n
b
e
n
c
h
m
a
r
k
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
i
n
t
C
o
n
f
e
r
e
n
c
e
o
n
N
e
u
r
a
l
N
e
t
w
o
rks
,
2
0
1
3
,
d
o
i
:
1
0
.
1
1
0
9
/
I
JC
N
N
.
2
0
1
3
.
6
7
0
6
8
0
7
.
[
6
]
W
.
Li
u
e
t
a
l
.
,
“
S
S
D
:
si
n
g
l
e
s
h
o
t
m
u
l
t
i
b
o
x
d
e
t
e
c
t
o
r
,
”
i
n
E
u
r
o
p
e
a
n
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
p
u
t
e
r V
i
si
o
n
,
S
p
r
i
n
g
e
r
,
Ed
.
,
2
0
1
6
,
p
p
.
2
1
–
3
7
.
[
7
]
G
.
H
u
a
n
g
,
Z.
L
i
u
,
L
.
v
a
n
d
e
r
M
a
a
t
e
n
,
a
n
d
K
.
Q
.
W
e
i
n
b
e
r
g
e
r
,
“
D
e
n
sel
y
c
o
n
n
e
c
t
e
d
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
t
w
o
r
k
s,
”
i
n
P
ro
c
e
e
d
i
n
g
s
o
f
t
h
e
I
EEE
C
o
m
p
u
t
e
r
S
o
c
i
e
t
y
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
s
i
o
n
a
n
d
P
a
t
t
e
r
n
Re
c
o
g
n
i
t
i
o
n
,
2
0
1
7
,
p
p
.
4
7
0
0
–
4
7
0
8
.
[
8
]
D
.
C
i
r
e
ş
a
n
,
U
.
M
e
i
e
r
,
J.
M
a
s
c
i
,
a
n
d
J.
S
c
h
m
i
d
h
u
b
e
r
,
“
M
u
l
t
i
-
c
o
l
u
m
n
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
f
o
r
t
r
a
f
f
i
c
si
g
n
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
N
e
u
r
a
l
N
e
t
w
o
rks
,
v
o
l
.
3
2
,
p
p
.
3
3
3
–
3
3
8
,
2
0
1
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
n
e
t
.
2
0
1
2
.
0
2
.
0
2
3
.
[
9
]
D
.
Ta
b
e
r
n
i
k
a
n
d
D
.
S
k
o
c
a
j
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
l
a
r
g
e
-
sc
a
l
e
t
r
a
f
f
i
c
-
si
g
n
d
e
t
e
c
t
i
o
n
a
n
d
r
e
c
o
g
n
i
t
i
o
n
,
”
I
E
EE
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
t
e
l
l
i
g
e
n
t
T
r
a
n
sp
o
r
t
a
t
i
o
n
S
y
st
e
m
s
,
v
o
l
.
2
1
,
n
o
.
4
,
p
p
.
1
4
2
7
–
1
4
4
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
TS.2
0
1
9
.
2
9
1
3
5
8
8
.
[
1
0
]
Y
.
Z
h
u
a
n
d
W
.
Q
.
Y
a
n
,
“
Tr
a
f
f
i
c
si
g
n
r
e
c
o
g
n
i
t
i
o
n
b
a
s
e
d
o
n
d
e
e
p
l
e
a
r
n
i
n
g
,
”
Mu
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
8
1
,
n
o
.
1
3
,
p
p
.
1
7
7
7
9
–
1
7
7
9
1
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
0
4
2
-
022
-
1
2
1
6
3
-
0.
[
1
1
]
Y
u
Z
h
a
n
g
a
n
d
Q
i
a
n
g
Y
a
n
g
,
“
A
s
u
r
v
e
y
o
n
m
u
l
t
i
-
t
a
s
k
l
e
a
r
n
i
n
g
,
”
I
E
EE
T
ra
n
sa
c
t
i
o
n
s
o
n
K
n
o
w
l
e
d
g
e
a
n
d
D
a
t
a
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
3
4
,
n
o
.
1
2
,
p
p
.
5
5
8
6
–
5
6
0
9
,
2
0
2
1
.
[
1
2
]
X
.
R
.
L
i
m,
C
.
P
.
Le
e
,
K
.
M
.
L
i
m,
T
.
S
.
O
n
g
,
A
.
A
l
q
a
h
t
a
n
i
,
a
n
d
M
.
A
l
i
,
“
R
e
c
e
n
t
a
d
v
a
n
c
e
s
i
n
t
r
a
f
f
i
c
si
g
n
r
e
c
o
g
n
i
t
i
o
n
:
a
p
p
r
o
a
c
h
e
s
a
n
d
d
a
t
a
se
t
s,
”
S
e
n
s
o
rs
,
v
o
l
.
2
3
,
n
o
.
1
0
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
3
1
0
4
6
7
4
.
[
1
3
]
W
.
L
i
,
H
.
S
o
n
g
,
a
n
d
P
.
W
a
n
g
,
“
F
i
n
e
l
y
c
r
a
f
t
e
d
f
e
a
t
u
r
e
s
f
o
r
t
r
a
f
f
i
c
si
g
n
r
e
c
o
g
n
i
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
C
i
rc
u
i
t
s,
S
y
st
e
m
s
a
n
d
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
,
v
o
l
.
1
6
,
p
p
.
1
5
9
–
1
7
0
,
2
0
2
2
,
d
o
i
:
1
0
.
4
6
3
0
0
/
9
1
0
6
.
2
0
2
2
.
1
6
.
2
0
.
[1
4
]
K
.
A
l
a
w
a
j
i
a
n
d
R
.
H
e
d
j
a
r
,
“
C
o
m
p
a
r
i
so
n
s
t
u
d
y
o
f
t
r
a
f
f
i
c
si
g
n
s
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
a
r
c
h
i
t
e
c
t
u
r
e
s
,
”
i
n
2
0
2
2
1
3
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
I
n
f
o
rm
a
t
i
o
n
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
S
y
st
e
m
s
,
I
C
I
C
S
2
0
2
2
,
2
0
2
2
,
p
p
.
4
4
2
–
4
4
7
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
C
S
5
5
3
5
3
.
2
0
2
2
.
9
8
1
1
2
1
6
.
[1
5
]
D
.
G
.
Le
e
,
“
F
a
s
t
d
r
i
v
a
b
l
e
a
r
e
a
s
e
s
t
i
m
a
t
i
o
n
w
i
t
h
m
u
l
t
i
-
t
a
s
k
l
e
a
r
n
i
n
g
f
o
r
r
e
a
l
-
t
i
me
a
u
t
o
n
o
mo
u
s
d
r
i
v
i
n
g
a
ssi
s
t
a
n
t
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
(
S
w
i
t
zer
l
a
n
d
)
,
v
o
l
.
1
1
,
n
o
.
2
2
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
1
2
2
1
0
7
1
3
.
[1
6
]
C.
-
Y
.
W
a
n
g
,
A
.
B
o
c
h
k
o
v
sk
i
y
,
a
n
d
H
.
-
Y
.
M
.
L
i
a
o
,
“
Y
O
LO
v
7
:
t
r
a
i
n
a
b
l
e
b
a
g
-
of
-
f
r
e
e
b
i
e
s
se
t
s
n
e
w
st
a
t
e
-
of
-
t
h
e
-
a
r
t
f
o
r
r
e
a
l
-
t
i
me
o
b
j
e
c
t
d
e
t
e
c
t
o
r
s,
”
a
rXi
v
p
re
p
ri
n
t
a
rXi
v
:
2
2
0
7
.
0
2
6
9
6
,
Ju
l
.
2
0
2
2
.
[1
7
]
J.
Z
h
a
n
g
,
Z.
X
i
e
,
J.
S
u
n
,
X
.
Z
o
u
,
a
n
d
J.
W
a
n
g
,
“
A
c
a
sc
a
d
e
d
r
-
c
n
n
w
i
t
h
mu
l
t
i
s
c
a
l
e
a
t
t
e
n
t
i
o
n
a
n
d
i
m
b
a
l
a
n
c
e
d
samp
l
e
s
f
o
r
t
r
a
f
f
i
c
si
g
n
d
e
t
e
c
t
i
o
n
,
”
I
EE
E
A
c
c
e
ss
,
v
o
l
.
8
,
p
p
.
2
9
7
4
2
–
2
9
7
5
4
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
7
2
3
3
8
.
B
I
O
G
RAP
H
Y
O
F
AUTHO
R
K
sha
tr
a
p
a
l
S
in
g
h
is
c
u
rre
n
tl
y
wo
rk
in
g
a
s
De
a
n
a
n
d
P
ro
fe
ss
o
r
(CS
E)
a
t
KCC
In
stit
u
te
o
f
Tec
h
n
o
l
o
g
y
a
n
d
M
a
n
a
g
e
m
e
n
t
G
re
a
t
e
r
No
id
a
.
He
re
c
e
iv
e
d
h
is
Ba
c
h
e
lo
r
o
f
Tec
h
n
o
l
o
g
y
(BTec
h
)
i
n
t
h
e
d
o
m
a
in
o
f
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
fro
m
Uttar
P
ra
d
e
sh
Tec
h
n
ica
l
Un
i
v
e
rsity
(U
P
TU),
L
u
c
k
n
o
w,
I
n
d
ia
in
2
0
0
4
.
He
d
id
h
is
M
.
Tec
h
(Co
m
p
u
ter
En
g
i
n
e
e
rin
g
)
fr
o
m
M
a
h
a
rsh
i
Da
y
a
n
a
n
d
U
n
iv
e
rsit
y
(M
.
D.U.)
,
Ro
h
tak
,
I
n
d
ia
in
2
0
1
0
.
He
wa
s
a
wa
rd
e
d
h
is
P
h
.
D
.
i
n
CS
E
fro
m
Dr.
A.P
.
J.
Ab
d
u
l
Ka
lam
Tec
h
n
ica
l
Un
iv
e
rsit
y
(AK
TU),
Lu
c
k
n
o
w.
He
wa
s
a
ss
o
c
iate
d
wi
th
v
a
ri
o
u
s
o
rg
a
n
isa
ti
o
n
s
n
a
m
e
ly
G
r
e
a
ter
No
id
a
In
stit
u
te
o
f
Tec
h
n
o
l
o
g
y
,
KIET
G
ro
u
p
o
f
I
n
sti
tu
ti
o
n
s
G
h
a
z
iab
a
d
,
Kris
h
n
a
En
g
i
n
e
e
rin
g
Co
ll
e
g
e
G
h
a
z
iab
a
d
,
I.
T.
S
E
n
g
in
e
e
rin
g
Co
ll
e
g
e
G
re
a
te
r
No
id
a
,
Li
n
g
a
y
a
’s
Un
i
v
e
rsity
F
a
rid
a
b
a
d
,
S
o
m
a
n
y
G
ro
u
p
o
f
In
stit
u
t
io
n
s,
Re
wa
ri
a
n
d
m
a
n
y
m
o
re
.
He
h
a
s
m
o
re
th
a
n
2
0
+
y
e
a
rs
o
f
a
c
a
d
e
m
ic
a
n
d
8
y
e
a
rs
o
f
re
se
a
rc
h
e
x
p
e
rien
c
e
.
His
re
se
a
rc
h
in
tere
sts
in
c
l
u
d
e
d
a
ta
a
n
a
l
y
ti
c
s,
c
o
m
p
u
tati
o
n
a
l
b
i
o
lo
g
y
,
b
i
o
-
in
fo
rm
a
ti
c
s,
g
ra
p
h
t
h
e
o
ry
i
n
c
o
m
p
u
ter
sc
ien
c
e
a
n
d
d
istri
b
u
te
d
s
y
ste
m
s.
He
p
u
b
li
s
h
e
d
m
o
re
th
a
n
6
0
re
se
a
rc
h
p
a
p
e
rs,
p
a
ten
ts
a
n
d
c
h
a
p
ters
i
n
re
p
u
ted
j
o
u
r
n
a
ls
o
f
d
iffere
n
t
p
u
b
li
s
h
e
rs.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
m
e
k
p
si
n
g
h
1
@g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.