I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
15
,
No
.
2
,
A
p
r
il
20
25
,
p
p
.
1
6
8
6
~
1
6
9
5
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
2
.
pp
1
6
8
6
-
1
6
9
5
1686
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Lo
w comp
lex
i
ty r
a
te
co
ntrol
for v
e
rsa
tile
v
ideo
codi
ng
wit
h
hy
brid
La
g
ra
ng
e
multiplier
Alex
a
nd
er
Vict
o
r
B
uk
it
,
Su
wa
di,
Wira
wa
n
D
e
p
a
r
t
me
n
t
o
f
E
l
e
c
t
r
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
F
a
c
u
l
t
y
o
f
I
n
t
e
l
l
i
g
e
n
t
E
l
e
c
t
r
i
c
a
l
a
n
d
I
n
f
o
r
m
a
t
i
c
s
T
e
c
h
n
o
l
o
g
y
,
I
n
s
t
i
t
u
t
T
e
k
n
o
l
o
g
i
S
e
p
u
l
u
h
N
o
p
e
m
b
e
r
,
S
u
r
a
b
a
y
a
,
I
n
d
o
n
e
si
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 2
3
,
2
0
2
4
R
ev
is
ed
Dec
2
,
2
0
2
4
Acc
ep
ted
Dec
1
4
,
2
0
2
4
Ve
rsa
ti
le
v
i
d
e
o
c
o
d
i
n
g
(V
VC)
h
a
s
a
n
o
ta
b
le
i
n
c
r
e
a
se
in
e
n
c
o
d
i
n
g
e
ff
icie
n
c
y
o
v
e
r
h
i
g
h
e
ff
ic
ie
n
c
y
v
i
d
e
o
c
o
d
i
n
g
(H
EV
C)
b
u
t
b
a
n
d
w
i
d
th
a
n
d
s
t
o
ra
g
e
a
re
li
m
it
e
d
i
n
re
a
l
-
w
o
rl
d
v
i
d
e
o
a
p
p
li
c
a
ti
o
n
s
,
s
o
ra
te
c
o
n
tr
o
l
is
c
r
u
c
ial
.
Ev
e
n
i
f
VV
C
s
h
o
ws
a
n
o
ta
b
le
i
n
c
re
a
s
e
i
n
e
n
c
o
d
in
g
c
a
p
a
c
i
t
y
o
v
e
r
HEV
C
,
t
h
e
ra
t
e
c
o
n
t
r
o
l
m
a
y
c
r
e
a
t
e
f
l
u
c
t
u
a
t
i
o
n
s
i
n
v
i
d
e
o
q
u
a
l
i
t
y
a
n
d
c
o
m
p
u
t
a
t
i
o
n
a
l
c
o
m
p
l
e
x
i
t
y
.
T
h
is
fl
u
c
t
u
a
ti
o
n
c
a
n
h
a
v
e
a
si
g
n
i
fica
n
t
im
p
a
c
t
o
n
th
e
wa
tc
h
i
n
g
e
x
p
e
r
ie
n
c
e
,
p
a
r
t
ic
u
la
rl
y
i
n
lo
w
-
b
it
ra
te
se
tt
i
n
g
s
.
T
h
e
ra
te
c
o
n
t
r
o
l
v
a
ri
a
b
les
o
f
th
e
c
o
n
s
is
te
n
t
re
s
o
l
u
ti
o
n
a
n
d
t
h
e
d
e
c
re
a
se
d
re
so
l
u
t
i
o
n
a
r
e
p
r
o
p
o
se
d
i
n
o
r
d
e
r
t
o
m
a
k
e
th
e
ra
te
c
o
n
tr
o
l
m
e
t
h
o
d
h
y
b
ri
d
t
o
c
o
n
si
s
te
n
t
re
s
o
l
u
t
i
o
n
s
a
n
d
re
d
u
c
e
c
o
m
p
u
t
a
t
i
o
n
a
l
c
o
m
p
le
x
i
t
y
.
T
o
a
c
h
ie
v
e
o
p
ti
m
a
l
c
o
d
in
g
p
a
ra
m
e
ters
,
t
h
e
fra
m
e
s
wi
th
v
a
r
y
i
n
g
re
s
o
l
u
ti
o
n
s
La
g
ra
n
g
e
m
u
lt
i
p
l
ier
(
λ
),
s
k
i
p
c
o
d
i
n
g
t
re
e
u
n
i
t
(
CTU
)
a
n
d
q
u
a
n
t
i
z
a
t
i
o
n
p
a
r
a
m
e
t
e
r
(
Q
P
)
a
r
e
c
o
m
b
i
n
e
d
.
A
f
t
e
r
e
v
a
l
u
a
t
i
o
n
,
t
h
e
a
v
e
r
a
g
e
e
n
c
o
d
i
n
g
t
i
m
e
s
a
v
i
n
g
s
(
)
w
e
re
f
o
u
n
d
t
o
b
e
1
9
.
4
9
%
,
w
i
t
h
a
B
j
o
n
t
e
g
a
a
rd
d
e
l
t
a
b
i
t
r
a
t
e
(
BD
B
R
)
o
f
-
0
.
4
4
%
i
n
d
i
c
a
t
i
n
g
i
n
s
i
g
n
i
f
i
c
a
n
t
q
u
a
l
i
t
y
l
o
s
s
.
K
ey
w
o
r
d
s
:
L
ag
r
an
g
e
m
u
ltip
lier
L
o
w
c
o
m
p
lex
ity
Qu
an
tizatio
n
p
ar
am
ete
r
R
ate
c
o
n
tr
o
l
Ver
s
atile
v
id
eo
co
d
in
g
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
:
Su
wad
i
Dep
ar
tm
en
t o
f
E
lectr
ical
E
n
g
i
n
ee
r
in
g
,
Facu
lty
o
f
I
n
tellig
en
t
E
lectr
ical
an
d
I
n
f
o
r
m
atics T
ec
h
n
o
lo
g
y
,
I
n
s
titu
t T
ek
n
o
lo
g
i Sep
u
lu
h
No
p
em
b
er
Su
r
ab
ay
a,
I
n
d
o
n
esia
E
m
ail: su
wad
i@
ee
.
its
.
ac
.
id
NO
M
E
N
CL
A
T
UR
E
:
T
h
e
allo
ca
ted
b
it r
ate
:
T
h
e
b
it r
ate
ta
r
g
et
:
Dis
to
r
tio
n
,
:
Mo
d
el
p
a
r
am
eter
s
ass
o
ciate
d
with
th
e
s
o
u
r
ce
'
s
ch
ar
ac
ter
is
tic
,
:
Vid
eo
s
o
u
r
ce
-
r
elate
d
f
ac
t
o
r
s
,
u
p
d
ated
b
ased
o
n
th
e
b
itra
te
:
Qu
an
tizatio
n
p
a
r
am
eter
:
T
h
e
in
d
ex
o
f
R
-
D
m
o
d
el
p
ar
a
m
eter
:
T
h
e
in
d
ex
o
f
C
T
U
:
R
ate
-
d
is
to
r
tio
n
co
s
t
∗
:
T
h
e
o
p
tim
u
m
am
o
u
n
t L
a
g
r
an
g
e
m
u
ltip
lier
(
)
:
No
is
y
s
u
b
g
r
a
d
ien
t
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
in
cr
ea
s
in
g
p
o
p
u
lar
ity
o
f
v
id
eo
a
p
p
licatio
n
s
h
as
in
c
r
ea
s
ed
th
e
d
em
an
d
f
o
r
m
o
r
e
e
f
f
ec
tiv
e
v
id
eo
co
d
in
g
tech
n
o
lo
g
ies.
I
n
p
r
ac
tical
v
id
eo
co
d
in
g
a
p
p
licatio
n
s
,
r
ate
co
n
tr
o
l
is
cr
u
cial
b
ec
au
s
e
o
f
b
an
d
wid
th
a
n
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Lo
w
co
mp
lexity
r
a
te
co
n
tr
o
l fo
r
ve
r
s
a
tile v
id
eo
co
d
in
g
w
ith
h
yb
r
id
…
(
A
lexa
n
d
er
V
icto
r
B
u
kit
)
1687
s
to
r
ag
e
b
u
d
g
et
c
o
n
s
tr
ain
ts
.
R
ate
co
n
tr
o
l
is
a
f
u
n
d
am
en
tal
t
o
o
l
in
v
id
eo
co
d
in
g
an
d
is
v
it
al
to
r
ea
lis
tic
v
id
eo
co
d
in
g
s
y
s
tem
s
b
ec
au
s
e
o
f
b
a
n
d
wid
th
o
r
s
to
r
ag
e
lim
itatio
n
s
in
r
ea
l
-
wo
r
ld
ap
p
licatio
n
s
.
I
t
s
g
o
al
is
to
r
e
d
u
ce
d
is
to
r
tio
n
in
c
o
m
p
r
ess
ed
v
id
eo
s
wh
ile
ad
h
er
i
n
g
to
b
it
-
r
a
te
r
estrictio
n
s
.
R
ate
co
n
tr
o
l
(
R
C
)
,
o
n
e
o
f
th
e
f
u
n
d
am
e
n
tal
co
m
p
o
n
e
n
ts
o
f
v
er
s
atile
v
id
eo
c
o
d
in
g
(
VVC)
,
allo
ws
f
o
r
m
o
r
e
p
r
ec
is
e
b
i
t
allo
ca
tio
n
an
d
is
p
ar
ticu
lar
ly
im
p
o
r
ta
n
t
f
o
r
r
ed
u
cin
g
d
is
to
r
tio
n
[
1
]
–
[
3
]
.
C
o
n
s
e
q
u
en
tly
,
wh
en
th
e
b
its
o
f
tar
g
e
t
in
t
h
e
r
ate
co
n
tr
o
l
m
eth
o
d
s
ar
e
lo
w
e
n
o
u
g
h
at
th
e
f
r
am
e
lev
el,
d
o
wn
s
am
p
lin
g
-
b
ased
co
d
in
g
o
u
g
h
t to
b
e
ad
v
a
n
tag
eo
u
s
.
R
ate
co
n
tr
o
l
h
as
b
ee
n
ex
te
n
s
iv
ely
r
esear
ch
ed
o
v
er
t
h
e
last
f
ew
d
ec
ad
es
f
o
r
v
ar
io
u
s
ap
p
licatio
n
s
o
f
v
id
eo
c
o
d
in
g
s
tan
d
ar
d
s
.
T
h
e
f
u
n
d
am
e
n
tal
id
ea
o
f
b
it
allo
ca
tio
n
is
to
ass
ig
n
a
g
r
ea
ter
n
u
m
b
er
o
f
b
its
to
h
ig
h
-
co
m
p
lex
ity
f
r
am
es
an
d
a
le
s
s
er
n
u
m
b
er
to
lo
w
-
c
o
m
p
le
x
ity
f
r
am
es
to
en
s
u
r
e
u
n
if
o
r
m
v
is
u
al
q
u
ality
.
Dep
en
d
in
g
o
n
th
e
in
tr
icac
y
o
f
th
e
cu
r
r
e
n
t
f
r
am
e,
q
u
an
tizat
io
n
p
ar
am
eter
(
QP)
is
d
y
n
am
ically
ch
an
g
ed
f
o
r
ev
er
y
f
r
a
m
e.
T
h
e
r
ef
o
r
e
,
th
e
co
m
p
lex
ity
m
ea
s
u
r
e
m
en
t
m
o
d
el
f
o
r
ea
ch
f
r
am
e
s
h
o
u
ld
b
e
p
r
o
p
o
s
ed
b
ef
o
r
e
d
eter
m
in
in
g
QP
f
o
r
ea
ch
f
r
am
e.
Ach
iev
in
g
th
e
tar
g
et
b
itra
te
an
d
r
ed
u
cin
g
d
is
to
r
tio
n
ar
e
t
h
e
two
g
o
als
o
f
r
ate
m
an
ag
em
en
t.
Gen
er
ally
,
th
e
r
e
ar
e
t
h
r
ee
t
y
p
es
o
f
ty
p
ical
r
ate
co
n
tr
o
l
m
o
d
els
R
-
Q,
ρ
d
o
m
ai
n
,
an
d
R
-
λ
m
o
d
els.
S
tu
d
ies
[
4
]
,
[
5
]
in
tr
o
d
u
ce
d
th
e
R
-
λ
m
o
d
el,
an
d
it
was
in
c
o
r
p
o
r
ate
d
in
to
th
e
h
ig
h
ef
f
icien
cy
v
id
eo
co
d
in
g
(
HE
VC
)
s
tan
d
ar
d
s
o
f
twar
e
H
E
VC
test
m
o
d
el
(
HM
)
[
6
]
,
[
7
]
.
T
h
e
R
-
λ
m
o
d
el
o
f
f
e
r
s
a
f
av
o
r
a
b
le
tr
ad
e
-
o
f
f
b
etwe
en
th
e
ac
cu
r
ac
y
o
f
p
ar
a
m
eter
am
o
u
n
ts
a
n
d
p
r
ed
ictio
n
wh
en
co
m
p
ar
ed
to
class
ic
r
a
te
co
n
tr
o
l
m
o
d
els,
wh
ich
co
u
ld
h
el
p
im
p
r
o
v
e
p
ar
am
eter
s
tab
ilit
y
.
Fo
llo
win
g
th
at,
HE
VC
h
as
b
ee
n
th
e
s
u
b
ject
o
f
m
u
ch
r
esear
c
h
b
ased
o
n
th
e
R
-
λ
m
o
d
el
[
8
]
–
[
1
4
]
.
T
h
e
f
u
n
d
am
e
n
tal
p
r
e
m
is
e
o
f
th
e
Q
-
d
o
m
ain
r
ate
co
n
tr
o
l
tech
n
iq
u
es
[
1
5
]
–
[
1
7
]
is
th
at,
g
iv
en
th
e
co
n
s
tr
ain
ts
o
f
th
e
d
esire
d
b
it
r
ate
R
,
th
e
o
p
tim
al
co
d
in
g
p
er
f
o
r
m
a
n
ce
ca
n
b
e
attain
ed
with
a
ca
r
ef
u
lly
s
elec
ted
q
u
a
n
tizatio
n
p
ar
a
m
eter
Q.
T
h
er
e
ar
e
s
ev
er
al
m
o
d
es
s
ets
to
ch
o
o
s
e
f
r
o
m
in
th
e
v
id
eo
co
d
in
g
s
y
s
tem
,
wh
ich
r
es
u
lts
in
an
en
o
r
m
o
u
s
n
u
m
b
er
o
f
r
ate
an
d
d
is
to
r
tio
n
(
R
-
D)
d
ata
p
air
s
.
Ho
wev
er
,
th
e
g
o
al
o
f
r
ate
co
n
tr
o
l
is
to
s
elec
t
th
e
b
est
m
o
d
e
f
o
r
a
d
esire
d
b
it
r
at
e
to
m
in
im
ize
d
is
to
r
tio
n
at
th
e
s
p
ec
if
ied
b
it
r
ate.
I
t
ca
n
b
e
tr
an
s
f
o
r
m
ed
i
n
to
an
u
n
co
n
s
tr
ain
e
d
f
o
r
m
af
ter
b
ei
n
g
f
o
r
m
u
lated
b
y
a
co
n
s
tr
ain
ed
o
p
tim
izatio
n
p
r
o
b
lem
.
T
wo
-
p
ass
r
ate
co
n
tr
o
l
tech
n
iq
u
es
[
1
8
]
,
en
c
o
d
ed
th
e
s
eq
u
en
ce
twice.
Mo
tio
n
v
ec
t
o
r
s
,
th
e
n
u
m
b
er
o
f
d
is
cr
ete
co
s
in
e
tr
an
s
f
o
r
m
(
DC
T
)
co
ef
f
icien
ts
,
h
ea
d
er
le
n
g
th
s
,
t
h
e
n
u
m
b
er
o
f
r
esid
u
al
b
its
p
er
f
r
am
e,
a
n
d
o
th
e
r
s
tatis
tical
in
f
o
r
m
atio
n
wer
e
g
ath
er
ed
d
u
r
in
g
th
e
f
ir
s
t
r
u
n
a
n
d
watc
h
ed
h
o
w
th
e
s
eq
u
en
ce
m
an
if
ested
its
elf
.
Usi
n
g
th
e
p
ar
am
eter
s
co
llected
in
th
e
f
ir
s
t
p
ass
,
th
e
QP
i
s
im
p
r
o
v
ed
in
th
e
s
ec
o
n
d
p
ass
.
Mo
s
t
two
-
p
ass
tech
n
iq
u
es
ar
e
u
s
ed
f
o
r
n
o
n
-
r
ea
l
-
tim
e,
f
ix
ed
-
s
ized
s
to
r
ag
e
ap
p
licatio
n
s
s
u
ch
as
o
n
lin
e
v
id
eo
s
tr
ea
m
in
g
an
d
d
i
g
ital
v
id
eo
d
is
cs
(
DVDs
)
,
wh
er
ea
s
o
n
e
-
p
ass
tech
n
iq
u
es
ar
e
m
o
r
e
ap
p
r
o
p
r
iate
f
o
r
r
ea
l
-
tim
e
ap
p
licatio
n
s
.
Ho
wev
er
,
esp
ec
ially
wh
en
th
er
e
a
r
e
m
ajo
r
m
o
tio
n
s
o
r
s
ce
n
e
ch
an
g
es,
th
e
d
is
to
r
tio
n
o
f
th
e
co
-
lo
ca
ted
c
o
d
in
g
tr
ee
u
n
it
(
C
T
U)
m
ay
b
e
v
er
y
d
if
f
er
e
n
t
f
r
o
m
th
e
r
ea
l d
is
to
r
tio
n
o
f
th
e
C
T
U.
T
h
is
is
b
ec
au
s
e
th
e
ap
p
r
o
ac
h
i
s
s
till
s
o
lely
f
o
cu
s
ed
o
n
in
ter
-
C
T
U
r
ate
co
n
tr
o
l.
A
s
im
ilar
s
tu
d
y
b
y
Fu
et
a
l.
[
1
9
]
p
r
o
p
o
s
ed
a
r
ate
co
n
tr
o
l
w
ith
r
eso
lu
tio
n
ch
an
g
es
f
o
r
VV
C
.
Usi
n
g
a
λ
-
d
o
m
ain
r
ate
c
o
n
tr
o
l
a
p
p
r
o
ac
h
th
at
allo
ws
f
o
r
r
eso
lu
tio
n
m
o
d
if
icatio
n
s
,
a
f
r
am
e'
s
co
d
in
g
r
eso
lu
tio
n
is
ca
lcu
lated
b
ased
o
n
its
p
r
o
p
e
r
ties
.
T
h
e
r
ate
co
n
tr
o
l
v
ar
iab
l
es
f
o
r
th
e
lo
w
r
eso
lu
tio
n
an
d
h
ig
h
r
eso
lu
tio
n
a
r
e
ch
an
g
ed
in
d
ep
en
d
en
tly
in
o
r
d
er
to
ad
ap
t
th
e
r
ate
co
n
tr
o
l
tech
n
iq
u
e
to
m
ix
ed
r
eso
lu
tio
n
s
.
Hy
b
r
id
L
ap
lace
d
is
tr
ib
u
tio
n
-
b
ased
lo
w
co
m
p
lex
ity
r
ate
-
d
is
to
r
tio
n
o
p
tim
ize
d
q
u
an
tizatio
n
was
p
r
o
p
o
s
ed
b
y
C
u
i
et
a
l.
[
2
0
]
,
th
ey
s
u
g
g
ested
a
lo
w
-
co
m
p
lex
ity
r
ate
-
d
is
to
r
tio
n
o
p
tim
ized
q
u
an
tizatio
n
(
R
DOQ
)
ap
p
r
o
ac
h
b
y
u
s
in
g
a
h
y
b
r
id
L
ap
lace
d
is
tr
ib
u
tio
n
to
m
o
d
el
th
e
s
tatis
tics
o
f
th
e
tr
an
s
f
o
r
m
co
ef
f
icie
n
ts
.
T
h
is
way
,
d
ep
e
n
d
in
g
o
n
t
h
e
co
ef
f
icien
t
d
is
tr
ib
u
tio
n
,
d
is
t
o
r
tio
n
m
o
d
els
an
d
b
lo
c
k
lev
el
r
ate
ar
e
s
p
ec
if
ically
d
e
v
elo
p
ed
.
T
h
u
s
,
b
y
m
ax
im
izin
g
t
h
e
R
D
p
er
f
o
r
m
a
n
ce
o
f
th
e
en
tire
b
lo
ck
,
th
e
id
ea
l
q
u
an
tizatio
n
lev
els
m
a
y
b
e
f
o
u
n
d
d
ir
ec
tly
,
an
d
th
e
lab
o
r
io
u
s
R
D
co
s
t
co
m
p
u
t
atio
n
s
ca
n
ev
en
tu
ally
b
e
av
o
i
d
ed
.
I
n
a
r
elate
d
s
tu
d
y
,
L
iao
e
t
a
l.
[
2
1
]
s
u
g
g
ested
f
r
am
e
-
lev
el
r
ate
co
n
tr
o
l
in
f
l
ex
ib
le
v
id
e
o
c
o
d
in
g
u
s
in
g
co
n
ten
t
-
ad
ap
tiv
e
r
ate
-
d
is
to
r
tio
n
m
o
d
elin
g
.
Fo
r
o
n
e
-
p
ass
v
id
eo
co
d
in
g
,
th
ey
p
r
o
p
o
s
ed
h
ig
h
-
o
r
d
e
r
R
-
D
m
o
d
els
an
d
co
r
r
esp
o
n
d
in
g
f
r
am
e
-
le
v
el
R
C
alg
o
r
ith
m
s
.
Fin
ally
,
th
e
q
u
ad
r
atic
R
-
D
m
o
d
el
is
in
co
r
p
o
r
ated
in
to
f
r
am
e
-
lev
el
R
C
in
th
e
VVC
te
s
t
m
o
d
el
(
VT
M)
an
d
a
co
n
ten
t
-
ad
a
p
tiv
e
m
o
d
el
s
elec
tio
n
is
m
ad
e
b
etwe
en
th
e
f
ir
s
t
-
o
r
d
er
an
d
s
ec
o
n
d
-
o
r
d
er
R
-
D
m
o
d
els.
An
R
D
p
ar
am
eter
u
p
d
atin
g
te
ch
n
iq
u
e
was
g
iv
en
b
y
L
i
et
a
l
.
in
[
2
2
]
,
wh
er
e
th
ey
also
e
x
a
m
in
ed
th
e
q
u
ality
d
ep
en
d
en
cy
b
etwe
en
f
r
am
es.
A
f
r
am
e
-
lev
el
r
ec
u
r
s
iv
e
B
ay
esian
esti
m
atio
n
(
R
B
E
)
b
ased
co
n
s
tan
t
b
it
-
r
ate
(
C
B
R
)
co
n
tr
o
l
tech
n
iq
u
e
was
p
r
esen
ted
in
[
2
3
]
.
As
it
s
tan
d
s
,
th
e
m
o
d
el
o
f
h
y
p
er
b
o
lic
R
-
D
i
s
h
ig
h
ly
ac
cu
r
ate,
th
u
s
cr
ea
tin
g
a
n
ew
v
er
s
io
n
o
f
t
h
e
m
o
d
el
will
n
o
t
lik
ely
in
cr
ea
s
e
ac
c
u
r
ac
y
m
u
c
h
f
u
r
t
h
er
.
T
h
e
R
B
E
,
u
tili
zin
g
u
p
d
ate
d
s
tag
es
an
d
a
lter
n
atin
g
p
r
e
d
ictio
n
,
c
o
u
ld
as
s
ess
b
it
r
ates
an
d
d
is
tr
ib
u
te
tar
g
et
b
it
r
ates
b
ased
o
n
th
e
d
is
to
r
tio
n
v
a
r
iatio
n
s
o
f
th
e
r
ec
en
tly
en
co
d
ed
f
r
a
m
es.
T
h
eir
m
eth
o
d
o
lo
g
y
r
ed
u
ce
s
t
h
e
m
u
lti
-
o
b
jectiv
e
o
p
tim
izatio
n
ch
allen
g
e
to
a
s
in
g
le
-
o
b
jectiv
e
is
s
u
e
aim
ed
at
m
in
im
izin
g
av
er
ag
e
d
is
to
r
tio
n
an
d
q
u
ality
v
ar
iatio
n
with
in
t
h
e
lim
ited
b
it
r
ate.
Su
b
s
eq
u
en
tl
y
,
a
d
u
al
-
p
h
ase
ap
p
r
o
ac
h
g
r
o
u
n
d
e
d
in
th
e
D
-
λ
m
o
d
el
was
s
u
g
g
ested
to
g
et
th
e
id
ea
l
r
eso
lu
tio
n
.
T
h
is
ap
p
r
o
ac
h
y
ield
s
a
s
o
lu
tio
n
th
at
ass
ig
n
s
th
e
b
it
r
ate
to
ea
ch
C
T
U
o
n
ce
f
o
r
all.
Un
f
o
r
tu
n
ately
,
th
e
ap
p
r
o
ac
h
n
ec
ess
itates
tw
o
-
p
ass
co
d
in
g
,
an
d
its
h
ig
h
co
s
t
o
f
co
m
p
le
x
ity
p
r
ev
en
ts
it f
r
o
m
b
ein
g
ap
p
lied
to
r
ea
l
-
tim
e
r
eq
u
ir
em
en
ts
in
R
C
cir
cu
m
s
tan
ce
s
.
I
n
th
is
s
tu
d
y
,
we
p
r
o
p
o
s
e
n
e
w
lo
w
co
m
p
lex
ity
r
ate
c
o
n
tr
o
l
m
eth
o
d
h
y
b
r
i
d
,
co
n
t
r
ib
u
tin
g
to
r
ed
u
ce
co
m
p
u
tatio
n
al
co
m
p
lex
ity
an
d
co
n
s
is
ten
t
r
eso
lu
tio
n
s
.
T
h
e
r
ate
co
n
tr
o
l
m
o
d
e
d
ec
is
io
n
s
tr
ateg
ies
u
s
in
g
tex
tu
r
e
d
etail
s
ce
n
ar
io
s
'
th
r
esh
o
ld
v
a
lu
es.
T
o
r
e
d
u
ce
c
o
m
p
u
tin
g
r
eq
u
ir
em
en
ts
an
d
ex
p
ed
ite
th
e
r
ate
co
n
tr
o
l
m
o
d
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
2
,
Ap
r
il
20
25
:
1
6
8
6
-
1
6
9
5
1688
d
eter
m
in
atio
n
p
r
o
ce
s
s
,
an
e
f
f
icien
t
ter
m
in
atio
n
an
d
b
y
p
ass
tech
n
iq
u
e
is
em
p
lo
y
ed
.
T
h
e
av
er
ag
e
e
n
co
d
i
n
g
tim
e
s
av
in
g
s
(
)
wer
e
f
o
u
n
d
to
b
e
1
9
.
4
9
%,
with
a
B
jo
n
teg
aa
r
d
d
elta
b
it
r
ate
(
B
DB
R
)
o
f
-
0
.
4
4
%
in
d
icatin
g
in
s
ig
n
if
ican
t
q
u
ality
lo
s
s
.
2.
L
AG
RANG
E
M
U
L
T
I
P
L
I
E
R
Acc
o
r
d
in
g
to
r
ate
-
d
is
to
r
tio
n
t
h
eo
r
y
,
th
e
r
e
is
an
o
p
tim
al
co
m
p
r
ess
io
n
r
ate,
R
(
D)
,
g
iv
en
a
n
ac
ce
p
tab
le
d
is
to
r
tio
n
D.
T
h
is
f
u
n
ctio
n
is
k
n
o
wn
as
th
e
r
ate
-
d
is
to
r
tio
n
(
R
-
D)
f
u
n
ctio
n
,
an
d
it
is
co
n
v
ex
an
d
g
r
ad
u
ally
d
ec
r
ea
s
in
g
.
A
d
ec
e
n
t
v
id
e
o
en
co
d
er
m
u
s
t
d
eter
m
in
e
th
e
o
p
ti
m
al
s
et
o
f
p
ar
am
eter
s
to
r
ed
u
ce
d
is
to
r
tio
n
wh
ile
ad
h
er
in
g
to
a
s
p
ec
if
ic
b
it r
ate
r
estrictio
n
.
T
h
e
f
o
llo
win
g
allo
c
atio
n
o
f
r
eso
u
r
ce
s
p
r
o
b
lem
ca
n
b
e
r
eso
lv
e
d
b
y
a
n
o
p
tim
al
r
ate
-
co
n
tr
o
l sy
s
tem
,
a
s
s
u
m
in
g
th
at
N
co
d
in
g
u
n
its
w
ill m
ak
e
u
p
th
e
f
u
ll v
i
d
eo
[
2
4
]
.
mi
n
{(
,
)}
∑
=
1
(
,
)
∑
(
,
)
≤
=
1
(
1
)
T
h
e
en
c
o
d
er
cr
ea
tes
a
b
its
tr
ea
m
b
y
co
m
p
r
ess
in
g
th
e
in
p
u
t
C
T
U
p
ix
els.
T
h
e
λ
-
v
alu
e
is
u
p
d
ated
u
s
in
g
th
e
p
r
o
d
u
ce
d
b
it
n
u
m
b
er
R
an
d
d
is
to
r
tio
n
D.
T
h
e
v
alu
es
R
an
d
D
ar
e
r
eg
ar
d
ed
as
r
an
d
o
m
v
ar
iab
les
s
in
ce
th
ey
v
ar
y
f
r
o
m
C
T
U
to
C
T
U.
T
h
e
f
o
llo
win
g
is
s
u
e
is
r
eso
lv
ed
b
y
r
ef
o
r
m
u
latin
g
th
e
r
ate
co
n
tr
o
l
p
r
o
ce
d
u
r
e.
Gen
er
ally
,
λ
is
u
s
ed
b
y
v
id
e
o
en
c
o
d
er
s
t
o
ad
j
u
s
t
R
an
d
D.
Fo
r
s
im
p
le
im
p
lem
en
tatio
n
s
,
v
id
eo
en
c
o
d
er
s
u
s
u
ally
u
s
e
em
p
ir
ically
p
r
e
d
eter
m
in
ed
λ
c
o
ef
f
icien
ts
f
o
r
QP
v
alu
es,
alth
o
u
g
h
λ
ca
n
b
e
d
eter
m
in
ed
th
eo
r
etica
lly
.
A
r
ate
co
n
tr
o
lle
r
u
s
es
th
e
R
-
QP
-
λ
r
elatio
n
to
ca
lcu
late
a
QP
b
ased
o
n
th
e
in
ten
d
ed
b
it
r
ates.
Nex
t,
th
e
QP
is
u
s
ed
in
th
e
en
co
d
er
in
t
h
e
v
id
e
o
co
m
p
r
ess
io
n
p
r
o
ce
s
s
.
I
n
th
is
en
co
d
e
r
,
th
e
R
DO
co
m
p
u
tes
th
e
jo
in
t
R
-
D
co
s
t
b
etwe
en
R
an
d
D
u
s
in
g
λ
to
id
en
tify
th
e
b
est
en
co
d
i
n
g
m
o
d
es
a
n
d
en
co
d
in
g
s
ettin
g
s
.
Fin
ally
,
th
e
en
co
d
er
a
d
d
s
r
esid
u
es
(
n
o
n
-
tex
tu
r
e
b
its
)
an
d
v
ar
i
o
u
s
p
ar
am
eter
v
alu
es
(
tex
tu
r
e
b
it
s
)
to
th
e
o
u
tp
u
t
b
it
-
s
tr
ea
m
s
.
I
t
is
n
o
ticed
th
at
th
e
λ
em
p
lo
y
ed
i
n
HE
VC
is
n
ea
r
ly
id
en
tical
t
o
VVC.
Fo
r
p
r
a
ctica
l
p
u
r
p
o
s
es,
th
e
q
u
an
tizatio
n
p
ar
am
eter
f
ac
to
r
(
)
is
d
ef
in
ed
as 0
.
8
5
.
T
h
e
s
lice
-
ty
p
e
in
VT
M
d
ef
in
es
λ
as
(
2
)
:
=
.
2
(
−
12
)
/
3
(
2
)
I
t
is
o
b
s
er
v
ed
th
at
QP
m
o
d
els
λ
in
d
ep
en
d
en
t
o
f
t
h
e
v
id
e
o
c
o
n
ten
t.
Fu
r
th
e
r
m
o
r
e
,
f
o
r
a
p
p
r
o
p
r
iate
b
it
allo
ca
tio
n
in
h
ier
ar
ch
ical
f
o
r
e
ca
s
tin
g
s
tr
u
ctu
r
es,
it
is
g
en
er
ally
ex
p
ec
ted
th
at
a
h
ig
h
e
r
lay
er
'
s
λ
v
alu
e
will
b
e
lar
g
er
th
a
n
its
lo
wer
lay
er
'
s
(
i.
e.
,
a
h
ig
h
er
la
y
er
'
s
QP
v
alu
e
will
b
e
lar
g
er
t
h
an
a
l
o
wer
la
y
er
'
s
)
.
I
n
th
is
s
tu
d
y
,
we
ap
p
r
o
x
im
ate
th
e
λ
v
alu
e
u
s
in
g
th
e
s
to
ch
asti
c
s
u
b
g
r
ad
ien
t
ap
p
r
o
ac
h
an
d
(
)
(
n
o
is
y
s
u
b
g
r
a
d
ien
t)
.
T
h
e
ℎ
iter
atio
n
u
p
d
ates th
e
v
alu
e
o
f
λ
by
(
3
)
:
(
+
1
)
=
(
)
+
(
)
(
3
)
T
h
is
s
tu
d
y
p
r
esen
ts
a
u
n
if
o
r
m
q
u
ality
c
r
iter
io
n
f
o
r
th
e
r
ate
c
o
n
tr
o
l
m
o
d
el,
an
d
th
e
R
-
D
o
p
tim
izatio
n
p
r
o
ce
d
u
r
e
y
ield
s
a
g
en
er
ic
o
p
tim
al
s
o
lu
tio
n
with
th
e
leas
t
p
er
ce
p
tu
al
d
is
to
r
tio
n
v
ar
iati
o
n
r
estrictio
n
.
T
h
e
m
ajo
r
ity
o
f
R
-
D
o
p
tim
izatio
n
s
tr
ateg
ies
ar
e
co
m
p
atib
le
with
th
e
s
o
lu
tio
n
.
Ultim
ately
,
co
m
p
r
eh
en
s
iv
e
test
s
ar
e
ca
r
r
ied
o
u
t
to
g
et
h
er
with
in
-
d
ep
th
an
aly
s
is
to
d
em
o
n
s
tr
ate
t
h
e
ef
f
ec
tiv
e
n
ess
o
f
th
e
s
u
g
g
es
ted
alg
o
r
ith
m
.
T
h
e
an
aly
s
is
lead
s
to
th
e
p
r
o
p
o
s
al
o
f
th
e
R
-
λ
r
ate
co
n
tr
o
l
m
o
d
el.
Fin
d
in
g
an
ac
cu
r
ate
esti
m
atio
n
f
o
r
th
e
i
d
ea
l
R
-
D
cu
r
v
e
is
im
p
o
r
tan
t
f
o
r
λ
v
alu
e
p
r
ed
ictio
n
.
R
an
d
D
s
tan
d
f
o
r
th
e
co
n
s
u
m
ed
b
its
an
d
d
is
to
r
tio
n
,
r
esp
ec
tiv
ely
.
T
h
e
co
n
s
u
m
ed
b
its
in
th
is
ca
s
e
o
u
g
h
t
to
b
e
clo
s
e
to
t
h
e
tar
g
et
b
its
.
T
h
e
L
ag
r
an
g
e
m
u
ltip
lier
,
o
r
λ
,
r
ep
r
esen
ts
th
e
R
-
D
c
u
r
v
e'
s
s
lo
p
e.
T
h
e
m
o
d
e
ch
o
ice
will
b
e
g
u
i
d
ed
b
y
λ
o
n
ce
it
h
as
b
ee
n
estab
lis
h
ed
.
T
h
e
d
e
r
iv
atio
n
o
f
λ
is
cr
u
cial
in
th
e
λ
-
d
o
m
ain
r
ate
co
n
tr
o
l
tech
n
iq
u
es.
T
h
e
lin
k
b
etwe
en
R
an
d
D
is
ch
ar
ac
ter
ized
b
y
th
e
h
y
p
er
b
o
lic
f
u
n
ctio
n
in
[
2
5
]
–
[
2
9
]
.
Sin
ce
λ
r
ep
r
esen
ts
th
e
R
-
D
cu
r
v
e'
s
s
lo
p
e,
it m
ay
b
e
wr
itte
n
as
(
4
)
:
=
−
=
.
−
−
1
≜
(
4
)
wh
er
e
α
a
n
d
β
ar
e
th
e
v
id
e
o
s
o
u
r
ce
-
r
elate
d
f
ac
to
r
s
.
T
o
g
et
a
m
o
r
e
ac
cu
r
ate
esti
m
ate
o
f
th
e
b
itra
te,
α
an
d
β
g
et
u
p
d
ates a
cc
o
r
d
in
g
to
th
e
b
itra
t
e
an
d
λ
o
f
th
e
c
o
d
ed
f
r
am
es.
Sin
ce
th
e
v
ar
iatio
n
in
q
u
ality
in
a
co
m
p
r
ess
ed
v
id
eo
s
ig
n
al
h
as
a
s
u
b
s
tan
tial
ef
f
ec
t
o
n
h
u
m
an
v
is
u
al
p
er
ce
p
tio
n
,
co
n
s
is
ten
t
q
u
ality
m
an
ag
e
m
en
t
s
ee
k
s
t
o
m
in
i
m
ize
th
e
d
is
to
r
tio
n
v
ar
iatio
n
o
v
er
v
i
d
eo
f
r
am
es
with
in
th
e
co
n
s
tr
ain
ts
o
f
f
r
am
e
r
ate,
b
an
d
wid
th
,
an
d
d
elay
r
eq
u
ir
ed
.
B
ased
o
n
th
e
p
r
e
v
io
u
s
ly
p
r
o
v
id
ed
r
ate
m
o
d
els
an
d
d
is
to
r
tio
n
,
th
e
o
p
tim
al
b
it
allo
ca
tio
n
p
r
o
b
lem
f
o
r
th
e
v
id
eo
co
d
in
g
s
y
s
tem
r
ep
lacin
g
an
d
f
r
o
m
th
e
s
tan
d
p
o
in
t
o
f
m
i
n
im
izin
g
av
er
a
g
e
(
MI
NAVE
)
d
is
to
r
tio
n
ca
n
b
e
p
h
r
ased
as
(
5
)
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Lo
w
co
mp
lexity
r
a
te
co
n
tr
o
l fo
r
ve
r
s
a
tile v
id
eo
co
d
in
g
w
ith
h
yb
r
id
…
(
A
lexa
n
d
er
V
icto
r
B
u
kit
)
1689
1
.
.
∗
=
1
.
.
∗
,
1
∑
(
)
=
1
∑
=
1
(
)
≤
(
5
)
wh
er
e
∗
is
th
e
o
p
tim
u
m
am
o
u
n
t
o
f
.
T
h
e
tech
n
iq
u
e
p
r
o
p
o
s
es
to
lo
o
s
en
th
e
c
o
n
s
tr
ain
t
∑
(
)
≤
=
1
with
th
e
in
eq
u
ality
o
f
g
e
o
m
et
r
ic
an
d
ar
ith
m
etic
m
ea
n
s
,
b
as
ed
o
n
th
e
o
b
s
er
v
atio
n
th
at
t
h
e
b
u
d
g
et
is
a
p
o
wer
f
u
n
ctio
n
co
n
ce
r
n
in
g
th
e
v
ar
iab
le
.
I
t
m
u
s
t,
h
o
wev
er
,
tak
e
in
to
ac
co
u
n
t
th
e
f
r
a
m
e
b
u
d
g
et
a
n
d
b
u
f
f
er
s
tate;
as
a
r
esu
lt,
it
m
u
s
t
m
o
d
if
y
∗
to
en
s
u
r
e
th
at
∑
(
)
≤
=
1
in
th
e
ac
tu
al
en
co
d
e
r
v
id
eo
co
d
in
g
.
T
h
e
en
co
d
er
ca
n
c
h
o
o
s
e
th
e
b
est
r
eso
lu
tio
n
b
y
co
m
b
i
n
in
g
d
o
wn
s
am
p
lin
g
an
d
c
o
d
in
g
d
is
to
r
tio
n
s
,
wh
ich
ar
e
im
p
ac
ted
b
y
f
r
am
e
s
p
atial
f
ea
tu
r
es,
u
s
in
g
th
e
co
d
in
g
r
eso
lu
tio
n
in
VVC.
B
ef
o
r
e
en
co
d
in
g
,
th
e
f
r
am
e
to
b
e
co
d
e
d
is
d
o
wn
a
n
d
u
p
-
s
am
p
le
d
.
T
h
e
u
p
-
s
am
p
led
f
r
am
e
is
th
en
co
m
p
a
r
ed
to
th
e
o
r
ig
in
al
f
r
am
e
to
estab
lis
h
th
e
d
o
wn
s
am
p
lin
g
d
is
to
r
tio
n
.
W
h
en
th
e
d
is
to
r
tio
n
o
f
d
o
wn
s
am
p
lin
g
is
b
elo
w
a
p
r
eset
th
r
esh
o
ld
,
th
e
d
o
wn
s
am
p
lin
g
-
b
ased
co
d
in
g
is
s
u
b
s
eq
u
en
tly
ap
p
lied
.
B
ased
o
n
t
h
e
s
tatis
tical
d
ata
an
aly
s
is
an
d
ea
r
lier
r
esear
ch
[
1
9
]
,
th
e
t
h
r
esh
o
ld
ca
n
b
e
d
eter
m
in
ed
u
s
in
g
QP.
=
0
.
175
.
−
3
.
5
(
6
)
E
v
en
with
g
r
ea
t a
ttem
p
ts
to
i
m
p
r
o
v
e
r
ate
co
n
tr
o
l o
f
R
-
λ
b
a
s
e,
th
e
p
ar
am
eter
s
ar
e
s
till
ca
lcu
lated
b
y
co
m
b
in
in
g
b
o
th
s
k
ip
an
d
n
o
n
-
s
k
ip
b
l
o
ck
s
,
wh
ich
r
esu
lts
in
a
n
im
p
r
ec
is
e
esti
m
atio
n
o
f
t
h
e
V
VC
s
tan
d
ar
d
.
Sk
ip
in
f
o
r
m
atio
n
ex
t
r
ac
tio
n
ca
n
h
e
lp
with
b
it
allo
ca
tio
n
at
th
e
C
T
U
lev
el
in
a
d
d
itio
n
to
h
el
p
in
g
with
p
ar
a
m
eter
esti
m
atio
n
an
d
th
e
R
-
D
cu
r
v
e.
=
(
7
)
=
−
∑
−
1
=
1
−
1
(
8
)
T
h
e
av
er
ag
e
b
u
d
g
et
allo
tted
f
o
r
ev
er
y
C
T
U
is
to
b
e
th
e
tar
g
et
b
its
(
5
)
.
W
h
en
th
e
s
tar
tin
g
λ
v
alu
e
is
n
o
t
n
ea
r
th
e
id
ea
l
v
alu
e,
th
e
f
lu
ctu
atin
g
r
esu
lt
b
its
m
ay
d
iv
er
g
e
f
r
o
m
th
e
in
ten
d
ed
.
As
a
r
esu
lt,
we
ad
ju
s
t
th
e
tar
g
et
b
its
f
o
r
t
h
e
ℎ
C
T
U
d
y
n
am
ically
an
d
ass
ig
n
th
e
r
em
ain
in
g
b
it
b
u
d
g
et
to
th
e
last
C
T
U
(
6
)
.
R
m
m
u
s
t
in
co
r
p
o
r
ate
th
e
b
its
f
r
o
m
th
e
h
ea
d
er
in
f
o
r
m
atio
n
p
r
o
d
u
ce
d
b
y
t
h
e
ℎ
C
T
U
a
s
wel
l
as
th
e
tex
tu
r
e
r
esid
u
al
to
esti
m
ate
th
e
tar
g
et
b
it
with
an
y
d
e
g
r
ee
o
f
ac
cu
r
ac
y
.
E
v
er
y
o
n
e
o
f
t
h
e
6
4
p
o
s
s
ib
le
QPs
is
ex
am
in
ed
to
s
ee
wh
ich
is
th
e
b
est
QP
at
th
e
lo
west
.
T
ested
ar
e
th
e
f
o
ll
o
win
g
2
5
v
i
d
eo
s
eq
u
en
ce
s
:
C
lass
A1
,
A2
,
B
,
C
,
D,
E
,
an
d
F.
T
h
e
g
o
al
o
f
s
elec
tin
g
th
ese
s
eq
u
en
ce
s
w
as
to
co
v
er
th
e
en
tire
s
p
ec
tr
u
m
o
f
m
o
tio
n
ac
tiv
ity
ch
ar
ac
ter
is
tics
,
r
an
g
in
g
f
r
o
m
l
o
w
to
h
ig
h
.
T
h
er
e
ar
e
s
ev
en
c
o
n
s
tan
t
λ
v
alu
es
f
o
r
ea
ch
s
er
ie
s
,
r
an
g
in
g
f
r
o
m
5
t
o
4
0
0
.
W
e
s
ea
r
ch
f
o
r
an
d
k
ee
p
t
h
e
b
est
QP
v
alu
e
f
o
r
ea
c
h
f
ix
e
d
λ
v
alu
e.
All
d
is
tr
ib
u
tio
n
s
,
it
tu
r
n
s
o
u
t,
ce
n
ter
o
n
th
eir
p
ea
k
s
.
W
e
g
ath
er
t
h
e
m
o
s
t lik
ely
QPs
an
d
th
eir
λ
v
alu
e
s
f
o
r
ev
er
y
test
v
id
eo
s
eq
u
en
c
e.
3.
M
E
T
H
O
D
Pro
v
id
e
r
ate
co
n
tr
o
l
f
o
r
v
i
d
eo
co
d
in
g
c
o
n
s
is
ts
o
f
two
p
ar
ts
:
th
e
r
ate
co
n
tr
o
l
m
o
d
el
a
n
d
th
e
tar
g
et
b
it
allo
ca
tio
n
.
I
n
th
e
f
o
r
m
e
r
,
p
a
r
am
eter
s
ar
e
ad
ju
s
ted
b
y
r
ate
co
n
tr
o
l
d
esig
n
s
to
g
et
an
ac
t
u
al
n
u
m
b
e
r
o
f
co
d
ed
b
its
th
at
is
as
clo
s
e
a
s
p
o
s
s
ib
l
e
to
th
e
tar
g
et
b
it
b
u
d
g
et
f
o
r
b
ase
co
d
in
g
u
n
its
,
f
r
am
es,
an
d
GOPs
;
in
th
e
latter
,
th
e
tar
g
et
b
itra
te
is
as
s
ig
n
ed
th
r
o
u
g
h
th
e
g
r
o
u
p
o
f
p
ictu
r
e
s
(
GOP
)
lev
el
to
th
e
f
r
am
e
an
d
C
T
U
lev
els
i
n
alig
n
m
en
t
with
a
m
in
im
u
m
am
o
u
n
t
o
f
d
is
to
r
tio
n
.
T
ar
g
et
b
it
im
p
lem
e
n
tatio
n
aim
s
t
o
co
n
v
e
r
t
tar
g
et
b
it
allo
ca
tio
n
v
alu
e
in
to
q
u
an
tiza
tio
n
p
ar
am
eter
QP
v
alu
e;
QP
d
eter
m
in
es
b
o
th
th
e
co
m
p
r
e
s
s
io
n
r
atio
an
d
th
e
f
in
al
en
co
d
in
g
r
ate.
Fo
llo
win
g
th
e
q
u
an
tizatio
n
p
ar
am
eter
'
s
v
alu
e
d
eter
m
in
atio
n
,
t
h
e
r
elev
a
n
t
p
ar
am
eter
s
m
u
s
t
b
e
u
p
d
ated
a
n
d
ad
ju
s
ted
.
T
a
r
g
et
b
it
im
p
lem
en
tatio
n
is
f
in
is
h
ed
at
th
e
p
ictu
r
e
an
d
C
T
U
lay
er
s
,
an
d
th
e
im
p
lem
en
tatio
n
co
n
ce
p
ts
ar
e
e
s
s
en
tially
th
e
s
am
e.
T
h
e
L
ag
r
an
g
e
m
u
ltip
lier
was
alter
ed
to
im
p
r
o
v
e
th
e
q
u
alit
y
d
ev
iatio
n
b
ased
o
n
th
e
d
is
to
r
tio
n
s
o
f
ea
r
lier
C
T
Us.
Ho
wev
er
,
esp
e
cially
wh
en
th
er
e
ar
e
n
o
tab
le
m
o
tio
n
s
o
r
s
ce
n
e
c
h
an
g
es,
t
h
e
d
is
to
r
tio
n
o
f
th
e
co
-
lo
ca
ted
C
T
U
m
ay
d
if
f
er
g
r
ea
tly
f
r
o
m
th
e
r
ea
l
d
is
to
r
tio
n
o
f
th
e
cu
r
r
e
n
t
C
T
U.
T
h
is
is
b
e
ca
u
s
e
th
e
ap
p
r
o
ac
h
is
s
till
s
o
lely
f
o
cu
s
ed
o
n
t
h
e
i
n
ter
-
C
T
U
r
ate
c
o
n
tr
o
l,
a
two
-
p
ass
r
ate
co
n
tr
o
l
tech
n
iq
u
e
th
a
t
u
s
es
d
o
wn
s
am
p
led
v
id
eo
c
o
d
in
g
to
p
r
o
v
id
e
co
n
s
i
s
ten
t
q
u
ality
.
T
h
e
L
a
g
r
an
g
e
m
u
ltip
lier
is
ap
p
lied
af
ter
th
e
f
ir
s
t
p
ass
to
m
in
im
ize
q
u
ality
f
lu
ct
u
atio
n
s
.
I
t
is
e
v
id
en
t
th
at
wh
ile
th
ese
c
u
r
r
e
n
t
c
o
n
s
is
ten
t
q
u
ality
-
o
r
ie
n
ted
alg
o
r
ith
m
s
ca
n
ai
d
in
s
m
o
o
th
in
g
th
e
q
u
ality
o
f
co
m
p
r
ess
ed
v
id
eo
,
th
e
m
ajo
r
ity
o
f
th
em
f
o
cu
s
ed
o
n
th
e
v
ar
iatio
n
in
q
u
ality
b
etwe
en
in
ter
f
r
am
es r
ath
er
th
an
th
e
v
ar
iatio
n
in
q
u
ality
i
n
s
id
e
a
s
in
g
l
e
f
r
am
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
2
,
Ap
r
il
20
25
:
1
6
8
6
-
1
6
9
5
1690
R
ate
co
n
t
r
o
l
m
a
y
ca
u
s
e
v
a
r
ia
tio
n
s
in
v
id
e
o
q
u
ali
ty
s
i
n
c
e
t
h
e
s
u
r
r
o
u
n
d
i
n
g
a
r
e
as
a
r
e
q
u
a
n
ti
ze
d
i
n
t
o
d
i
f
f
er
e
n
t
q
u
alit
y
le
v
els.
T
h
is
ca
n
h
a
v
e
a
s
i
g
n
if
i
ca
n
t
i
m
p
a
ct
o
n
t
h
e
w
atc
h
i
n
g
e
x
p
e
r
ie
n
ce
,
p
a
r
ti
cu
la
r
l
y
in
l
o
w
b
it
r
a
te
s
ett
in
g
s
.
B
as
ed
o
n
t
h
e
ca
l
cu
lat
io
n
o
f
f
r
am
e
co
m
p
le
x
i
ty
,
w
e
s
u
g
g
est
a
f
ast
a
n
d
a
cc
u
r
at
e
t
ec
h
n
iq
u
e
f
o
r
all
o
c
ati
n
g
b
its
at
th
e
f
r
am
e
l
a
y
er
in
VVC
r
a
te
r
e
g
u
l
ati
o
n
.
T
h
e
ta
r
g
et
b
i
t c
a
n
b
e
ass
i
g
n
e
d
t
o
e
ac
h
f
r
am
e
i
n
a
GO
P
ac
c
o
r
d
in
g
to
th
e
p
r
ed
ict
e
d
c
o
m
p
l
e
x
it
y
.
C
o
m
p
ar
in
g
th
e
s
u
g
g
est
ed
al
g
o
r
it
h
m
t
o
o
t
h
e
r
a
v
ai
la
b
l
e
te
ch
n
i
q
u
es
,
ex
p
e
r
i
m
e
n
t
al
r
esu
lts
d
em
o
n
s
t
r
ate
t
h
at
i
t
m
a
y
g
r
e
atl
y
r
ed
u
ce
t
h
e
q
u
alit
y
v
a
r
ia
ti
o
n
i
n
t
h
e
f
r
a
m
e
l
ay
e
r
.
Fig
u
r
e
1
s
h
o
ws
th
e
p
r
o
p
o
s
ed
m
eth
o
d
alg
o
r
ith
m
f
lo
wch
ar
t
s
o
f
th
e
r
ate
co
n
tr
o
l
m
o
d
e
d
ec
is
io
n
in
VT
M.
T
h
e
v
alu
e
o
f
L
ag
r
an
g
e
m
u
ltip
lier
(
λ
)
is
ca
lcu
lated
f
r
o
m
(
2
)
an
d
t
h
e
p
r
o
jecte
d
R
-
D
p
ar
am
eter
s
,
C
,
R
an
d
K,
co
m
e
f
r
o
m
(
4
)
.
T
h
e
v
a
lu
es
o
f
C
a
n
d
K
ar
e
t
h
e
m
o
d
el
p
ar
am
eter
s
ass
o
ciate
d
t
o
th
e
ch
ar
ac
ter
is
tic
o
f
t
h
e
in
p
u
t.
W
h
en
th
e
r
e
ar
e
s
ev
er
al
co
d
in
g
s
tag
es
an
d
f
r
am
es
wit
h
m
u
ltip
le
m
ix
e
d
s
k
ip
C
T
U,
t
h
e
f
r
am
e
le
v
el
with
b
it
allo
ca
tio
n
n
ee
d
s
to
b
e
tak
e
n
in
to
co
n
s
id
er
atio
n
m
o
r
e
th
o
r
o
u
g
h
ly
.
E
ac
h
C
T
U
is
in
itially
th
o
u
g
h
t
o
f
as
th
e
s
tan
d
ar
d
,
o
r
n
o
n
-
s
k
ip
b
lo
c
k
.
I
n
itial
v
alu
es
ar
e
p
r
o
v
id
ed
f
o
r
th
ese
two
p
ar
am
eter
s
,
an
d
th
ey
will
b
e
ad
ju
s
ted
co
n
tin
u
o
u
s
ly
in
th
e
n
ex
t
s
tep
.
I
f
th
e
ar
ea
f
alls
b
elo
w
a
p
r
ed
ef
in
ed
th
r
esh
o
ld
(
T
h
)
,
th
is
p
ar
ti
cu
lar
C
T
U
will
b
e
m
an
ag
ed
as
a
s
tan
d
ar
d
C
T
U.
T
h
e
s
k
ip
C
T
U
wo
u
ld
b
e
in
t
er
p
r
eted
o
th
er
wis
e.
Af
ter
th
e
cu
r
r
e
n
t
f
r
am
e
h
as
co
m
p
leted
e
n
co
d
i
n
g
,
d
is
to
r
tio
n
,
th
e
p
ix
els,
a
n
d
s
p
e
n
t
b
its
i
n
th
e
s
tan
d
ar
d
C
T
U
will
p
ar
ti
cip
ate
in
th
e
f
r
a
m
e
-
lev
el
R
-
D
v
ar
iab
le
u
p
d
ate.
T
h
e
d
ata
in
s
k
ip
C
T
U
w
ill n
o
t
b
e
u
s
ed
.
Fig
u
r
e
1
.
Pro
p
o
s
ed
m
et
h
o
d
al
g
o
r
ith
m
f
lo
wch
ar
ts
o
f
t
h
e
r
ate
co
n
tr
o
l m
o
d
e
d
ec
is
io
n
in
VT
M
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Lo
w
co
mp
lexity
r
a
te
co
n
tr
o
l fo
r
ve
r
s
a
tile v
id
eo
co
d
in
g
w
ith
h
yb
r
id
…
(
A
lexa
n
d
er
V
icto
r
B
u
kit
)
1691
W
h
en
th
e
m
o
d
e
d
ec
is
io
n
is
to
s
k
ip
C
T
U,
u
p
d
atin
g
th
e
λ
v
al
u
e
ca
n
b
e
p
r
e
-
elim
in
ated
w
h
e
n
a
C
T
U
is
class
if
ied
as
n
o
n
-
SKI
P.
ℎ
is
eq
u
al
to
0
.
0
0
1
an
d
ℎ
is
0
.
9
[
3
0
]
a
n
d
ca
n
b
e
m
o
d
if
ied
to
p
r
o
d
u
c
e
d
if
f
e
r
en
t
o
u
tco
m
es
[
3
1
]
.
W
h
en
th
e
am
o
u
n
t
o
f
b
it
p
er
p
ix
el
(
)
as
m
ea
s
u
r
ed
b
y
R
ex
ce
ed
s
an
u
p
p
e
r
th
r
esh
o
ld
ℎ
o
r
f
alls
b
elo
w
a
lo
wer
th
r
esh
o
l
d
ℎ
,
th
e
ar
ea
r
atio
o
f
Sk
ip
C
T
U
in
th
e
co
d
ed
C
T
U.
Fo
r
th
e
test
,
s
ev
er
al
v
id
eo
s
eq
u
en
ce
s
f
r
o
m
C
lass
es A1
,
A2
,
B
,
C
,
D,
E
,
an
d
F a
r
e
s
elec
ted
,
co
n
tain
in
g
all
f
r
am
es
at
d
ef
au
lt
f
r
am
e
r
ates.
T
h
e
f
ix
QP
m
eth
o
d
r
ef
er
s
to
f
o
u
r
f
ix
e
d
QP
2
2
,
2
7
,
3
2
an
d
3
7
th
at
ar
e
s
et
to
d
o
v
id
eo
c
o
d
in
g
o
n
s
eq
u
e
n
c
es
with
o
u
t
r
ate
co
n
tr
o
l
in
itia
lly
wh
ile
tak
in
g
lo
w
b
itra
te
co
n
s
id
er
atio
n
in
to
ac
co
u
n
t.
T
h
e
Fix
QP
-
b
ased
c
o
m
p
r
ess
io
n
b
itra
tes
ar
e
th
en
estab
lis
h
ed
as
th
e
d
esire
d
b
itr
ates
f
o
r
e
v
er
y
r
ate
co
n
tr
o
l
m
eth
o
d
u
s
ed
in
th
is
s
tu
d
y
.
C
o
m
p
ar
is
o
n
s
with
o
th
e
r
ap
p
r
o
ac
h
es
ar
e
an
ch
o
r
ed
b
y
t
h
e
VT
M
1
4
.
2
r
at
e
co
n
tr
o
l
m
eth
o
d
'
s
p
er
f
o
r
m
an
ce
.
I
t
s
h
o
u
ld
b
e
n
o
ted
th
at
th
e
d
ef
au
lt
tech
n
iq
u
es
f
o
r
r
ate
c
o
n
tr
o
l
ar
e
s
till
f
r
am
e
lev
el
an
d
GOP
lev
el.
T
h
e
n
ex
t
s
u
b
s
ec
tio
n
s
co
m
p
ar
e
a
n
d
p
r
o
v
i
d
e
d
etailed
ex
p
er
im
en
t
al
d
ata
f
o
r
all
r
ate
co
n
tr
o
l
alg
o
r
ith
m
s
,
i
n
clu
d
in
g
th
e
q
u
ality
v
ar
ia
n
ce
,
th
e
o
b
j
ec
tiv
e
an
d
p
er
ce
p
t
u
al
R
-
D
ef
f
icac
y
,
an
d
th
e
R
C
ac
cu
r
ac
y
at
d
i
f
f
er
en
t le
v
els,
an
d
th
e
s
u
b
jectiv
e
lev
el
o
f
p
e
r
f
o
r
m
an
ce
with
co
d
i
n
g
co
m
p
lex
it
y
.
W
h
en
α
is
f
ix
ed
,
a
b
ig
g
er
v
alu
e
co
r
r
esp
o
n
d
s
to
a
f
aster
a
p
p
r
o
ac
h
o
f
to
th
e
o
p
tim
u
m
.
Ho
we
v
er
,
if
λ
v
ar
ies
m
o
r
e
ab
o
u
t
th
e
co
n
v
er
g
en
t
v
alu
e,
th
e
c
o
d
in
g
b
ec
o
m
es
less
ef
f
ec
tiv
e.
W
e
ad
ap
tiv
ely
v
ar
y
th
e
v
alu
e
o
f
λ
in
th
e
VT
M
r
ef
e
r
en
ce
s
o
f
t
war
e'
s
en
co
d
in
g
p
r
o
ce
s
s
to
ex
p
lo
r
e
th
e
im
p
licatio
n
s
o
f
v
ar
io
u
s
s
tep
s
izes.
Alg
o
r
ith
m
1
e
x
p
lain
s
th
e
lo
w
co
m
p
lex
ity
r
ate
co
n
tr
o
l f
o
r
th
e
VVC m
eth
o
d
.
Alg
o
r
ith
m
1
.
Ps
eu
d
o
co
d
e
o
f
l
o
w
co
m
p
lex
ity
r
ate
co
n
t
r
o
l
f
o
r
VVC
Input:
GOP, Frame, CTU;
Output: Mode decision, QP, CU,
λ
;
Step l.
Prefetch CTU;
Step 2.
Calculate
=
.
2
(
−
12
)
/
3
;
Step 3.
Calculate C, R and K from (4);
Step 4.
if skip CTU jump to end;
Step 5.
Update λ
(n+1)
=
λ
(n)
+α
n
g
(n)
Step 6.
if current
decision = BestRD
COST
then;
Step 7.
if
∑
(
,
)
≤
=
1
then;
Step 8.
end if
Step 9.
jump to step 3
Step 10.
return mode decision, QP,
λ
;
Step 11. end
Fo
r
ac
cu
r
ate
R
D
co
s
t
ca
lcu
la
tio
n
,
co
m
p
lex
ity
r
ed
u
ctio
n
n
ec
ess
itate
s
th
e
cr
ea
tio
n
o
f
p
r
ec
is
e
r
ate
-
d
is
to
r
tio
n
m
o
d
els.
Mu
c
h
less
atten
tio
n
h
as
r
ec
en
tly
b
ee
n
p
a
id
to
o
p
tim
izin
g
th
e
q
u
an
tizati
o
n
p
r
o
ce
s
s
,
d
esp
ite
th
e
f
ac
t
th
at
n
u
m
er
o
u
s
p
r
ec
is
e
r
ate
d
is
to
r
tio
n
m
o
d
els
h
av
e
b
ee
n
p
u
t
o
u
t
f
o
r
th
e
f
ast
m
o
d
e
d
eter
m
in
atio
n
an
d
r
ate
co
n
tr
o
l
d
o
m
ai
n
s
.
T
h
r
o
u
g
h
th
e
estab
lis
h
m
en
t
o
f
s
p
ec
iall
y
cr
ea
ted
r
ate
an
d
d
is
to
r
tio
n
m
o
d
els
b
ased
o
n
th
e
p
r
o
p
o
s
ed
m
o
d
el
ap
p
r
o
ac
h
,
a
lo
w
co
m
p
lex
ity
alg
o
r
ith
m
is
p
r
o
p
o
s
ed
in
th
is
s
tu
d
y
.
C
o
n
s
eq
u
en
tly
,
with
k
n
o
wled
g
e
o
f
th
e
d
is
to
r
tio
n
an
d
r
ate
m
o
d
els
as
f
u
n
ctio
n
s
o
f
th
e
r
esid
u
al
s
tatis
tics
an
d
q
u
an
tizatio
n
p
ar
am
eter
s
,
it
is
p
o
s
s
ib
le
to
s
u
cc
ess
f
u
lly
av
o
id
th
e
co
m
p
u
tatio
n
ally
in
ten
s
iv
e
s
ea
r
ch
f
o
r
th
e
o
p
tim
al
q
u
an
tized
co
ef
f
icien
ts
.
Fig
u
r
e
2
illu
s
tr
at
es
th
e
v
is
u
al
co
m
p
ar
is
o
n
s
o
f
Fo
u
r
Peo
p
le
s
eq
u
en
ce
b
etwe
en
th
e
r
ec
o
n
s
tr
u
cted
f
r
am
es
o
b
tain
ed
u
s
in
g
th
e
d
ef
au
lt
r
ate
co
n
tr
o
l
in
VT
M
1
4
.
2
in
Fig
u
r
e
2
(
a)
an
d
af
ter
r
ec
o
n
s
tr
u
cted
u
s
in
g
o
u
r
r
ate
co
n
tr
o
l
ap
p
r
o
ac
h
in
Fig
u
r
e
2
(
b
)
.
T
h
e
co
m
p
ar
is
o
n
o
f
s
u
b
jectiv
e
v
is
u
als
s
h
o
ws
in
s
ig
n
if
ican
t
q
u
ality
lo
s
s
an
d
B
DB
R
as
s
h
o
wn
in
T
ab
le
1
.
(
a)
(
b
)
Fig
u
r
e
2
.
T
h
e
co
m
p
ar
is
o
n
o
f
s
u
b
jectiv
e
v
is
u
al
o
f
Fo
u
r
Peo
p
le
s
eq
u
en
ce
(
a)
d
ef
a
u
lt VT
M
an
d
(
b
)
o
u
r
p
r
o
p
o
s
ed
m
eth
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
2
,
Ap
r
il
20
25
:
1
6
8
6
-
1
6
9
5
1692
T
ab
le
1
.
T
h
e
r
esu
lt o
f
t
h
e
p
r
o
p
o
s
ed
m
eth
o
d
C
l
a
s
s
S
e
q
u
e
n
c
e
P
r
o
p
o
se
d
m
e
t
h
o
d
B
D
B
R
(
%)
(
%)
(
%)
A1
Ta
n
g
o
1
-
0
.
0
9
1
7
.
4
8
0
.
0
7
8
F
o
o
d
M
a
r
k
e
t
4
0
.
2
8
1
7
.
7
9
0
.
0
9
3
C
a
m
p
f
i
r
e
0
.
4
1
2
2
.
4
7
0
.
5
3
A2
C
a
t
R
o
b
o
t
-
0
.
0
3
2
1
7
.
5
8
0
.
0
7
6
D
a
y
l
i
g
h
t
R
o
a
d
2
0
.
3
5
1
8
.
5
1
0
.
0
6
7
P
a
r
k
R
u
n
n
i
n
g
3
-
0
.
0
9
1
6
.
4
2
0
.
0
2
5
B
K
i
m
o
n
o
-
1
.
3
6
2
1
.
1
1
0
.
1
2
P
a
r
k
S
c
e
n
e
0
.
0
8
2
0
.
1
3
0
.
0
9
5
C
a
c
t
u
s
-
0
.
5
4
1
9
.
8
2
0
.
1
2
B
a
s
k
e
t
b
a
l
l
D
r
i
v
e
0
.
1
7
3
1
6
.
8
7
0
.
0
9
B
Q
Te
r
r
a
c
e
0
.
2
8
2
0
.
2
1
0
.
1
6
6
C
B
a
s
k
e
t
b
a
l
l
D
r
i
l
l
-
0
.
9
8
1
8
.
7
9
0
.
3
2
B
Q
M
a
l
l
-
0
.
3
1
2
1
.
6
5
0
.
3
2
P
a
r
t
y
S
c
e
n
e
-
1
.
2
1
2
1
.
8
9
0
.
1
8
R
a
c
e
H
o
r
s
e
sC
0
.
1
3
1
6
.
3
4
0
.
2
1
D
B
a
s
k
e
t
b
a
l
l
P
a
ss
-
0
.
3
4
1
7
.
4
3
0
.
6
2
B
Q
S
q
u
a
r
e
-
0
.
8
7
2
1
.
7
9
0
.
4
8
B
l
o
w
i
n
g
B
u
b
b
l
e
s
-
0
.
1
1
2
2
.
1
9
0
.
3
5
R
a
c
e
H
o
r
s
e
s
-
0
.
0
9
1
4
.
9
8
0
.
8
7
E
F
o
u
r
P
e
o
p
l
e
-
0
.
3
3
2
2
.
9
8
0
.
5
5
Jo
h
n
y
-
0
.
7
6
2
2
.
8
7
0
.
6
4
5
K
r
i
s
t
e
n
A
n
d
S
a
n
d
r
a
-
1
.
2
4
1
8
.
1
3
0
.
7
8
F
B
a
s
k
e
t
b
a
l
l
D
r
i
l
l
Te
x
t
-
0
.
8
9
1
6
.
2
1
0
.
3
9
S
l
i
d
e
E
d
i
t
i
n
g
-
0
.
1
6
2
0
.
9
5
0
.
3
5
S
l
i
d
e
S
h
o
w
-
0
.
8
8
2
2
.
7
8
0
.
6
5
A
v
e
r
a
g
e
-
0
.
4
4
1
9
.
4
9
0
.
3
2
7
4.
RE
SU
L
T
AND
DI
SCUS
SI
O
N
A
co
m
p
ar
is
o
n
with
th
e
r
ate
co
n
tr
o
l
m
eth
o
d
f
o
u
n
d
in
VT
M
1
4
.
2
is
d
o
n
e
in
o
r
d
er
to
c
o
n
f
ir
m
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
s
u
g
g
est
ed
r
ate
co
n
tr
o
l
s
tr
ateg
y
.
T
h
e
C
T
C
f
o
r
lo
w
d
elay
(
L
D)
an
d
all
in
tr
a
(
AI
)
co
n
f
ig
u
r
atio
n
s
ar
e
u
s
ed
f
o
r
th
e
s
im
u
latio
n
s
.
As
d
o
wn
s
am
p
lin
g
-
b
ased
co
d
in
g
is
o
f
ten
ad
v
an
tag
eo
u
s
f
o
r
l
o
w
b
itra
tes an
d
h
ig
h
-
r
eso
lu
tio
n
s
eq
u
en
ce
s
with
test
QPs
s
et
to
{
2
2
,
2
7
,
3
2
,
3
7
}.
T
o
attain
th
e
g
o
al
b
itra
te
u
s
ed
in
o
u
r
r
esear
ch
,
th
e
test
s
eq
u
en
ce
s
ar
e
en
c
o
d
ed
u
s
in
g
t
h
e
o
r
ig
in
al
VT
M
1
4
.
2
with
o
u
t
th
e
p
r
o
p
o
s
ed
r
ate
co
n
tr
o
l
m
o
d
e
b
y
th
e
af
o
r
em
en
tio
n
ed
test
cir
cu
m
s
tan
ce
s
.
T
h
e
r
esu
ltin
g
tar
g
et
b
itra
te
is
th
en
u
tili
ze
d
a
s
th
e
in
p
u
t
tar
g
et
b
itra
te
f
o
r
"Pr
o
p
o
s
ed
R
C
"
an
d
"R
C
in
VT
M
-
1
4
.
2
"
af
ter
b
ein
g
r
o
u
n
d
ed
to
t
h
e
n
ea
r
est
in
teg
er
.
Sin
ce
th
e
o
b
jectiv
e
o
f
r
a
te
co
n
tr
o
l
is
to
ac
h
iev
e
th
e
d
esire
d
b
itra
te
an
d
m
in
im
ize
d
is
to
r
tio
n
,
th
e
r
ate
c
o
n
tr
o
l
alg
o
r
ith
m
is
ev
al
u
ated
u
s
in
g
b
o
th
b
itra
te
er
r
o
r
an
d
R
-
D
p
er
f
o
r
m
a
n
ce
.
I
n
th
is
ca
s
e,
th
e
b
itra
te
er
r
o
r
ca
n
b
e
co
m
p
u
ted
u
s
in
g
:
=
|
−
|
×
100%
(
9
)
wh
er
e
is
th
e
d
esig
n
ated
tar
g
e
t
b
itra
te
an
d
is
th
e
e
n
co
d
er
'
s
o
u
tp
u
t
b
itra
te.
I
n
th
is
p
ap
e
r
,
VT
M
1
4
.
2
was
test
ed
o
n
a
PC
r
u
n
n
in
g
L
i
n
u
x
Ub
u
n
t
u
2
0
.
0
4
.
4
L
T
S
with
an
I
n
tel®
C
o
r
eT
M
i7
-
6
7
0
0
0
T
C
o
r
e
@
2
.
8
0
GHz
p
r
o
ce
s
s
o
r
.
T
h
ir
t
y
f
r
am
es
ar
e
ex
am
in
ed
f
o
r
ea
c
h
s
eq
u
en
ce
i
n
th
e
test
s
,
wh
ich
ar
e
co
n
d
u
ct
ed
u
s
in
g
th
e
J
VE
T
co
m
m
o
n
test
co
n
d
itio
n
s
.
T
ab
l
e
1
d
is
p
lay
s
th
e
r
esu
lt
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
.
T
h
e
d
ef
au
lt
t
est
co
n
d
itio
n
s
u
s
ed
in
th
is
p
ap
er
ar
e
co
m
m
o
n
test
co
n
d
itio
n
s
with
q
u
an
tizatio
n
p
ar
am
eter
(
QP)
:
2
2
,
2
7
,
3
2
,
3
7
,
an
d
all
in
tr
a
(
AI
)
.
T
h
e
co
d
in
g
p
er
f
o
r
m
an
ce
is
m
ea
s
u
r
ed
u
s
in
g
B
DB
R
,
an
d
t
h
e
en
co
d
in
g
tim
e
s
av
in
g
s
(
)
ar
e
co
m
p
ar
ed
to
th
e
o
r
ig
in
al
VT
M
p
r
o
g
r
am
(
)
an
d
th
e
s
u
g
g
ested
tech
n
iq
u
e
(
)
.
T
h
e
f
o
llo
win
g
p
r
o
v
id
es
th
e
av
er
ag
e
e
n
co
d
i
n
g
tim
e
s
av
in
g
s
,
wh
ich
is
a
m
ea
s
u
r
e
o
f
co
m
p
lex
ity
r
ed
u
ctio
n
:
T
s
av
i
n
g
=
T
i
m
e
V
T
M
−
T
i
m
e
pr
o
po
s
ed
T
i
m
e
V
T
M
×
100%
(
1
0
)
R
ate
co
n
tr
o
l
will
in
v
ar
iab
ly
r
esu
lt
in
a
h
ea
v
y
co
m
p
u
tatio
n
al
co
s
t.
T
h
u
s
,
we
ex
am
i
n
e
o
u
r
m
eth
o
d
'
s
co
n
v
er
g
en
ce
s
p
ee
d
f
r
o
m
t
h
e
p
er
s
p
ec
tiv
e
o
f
u
tili
ze
d
tim
e
.
No
tab
ly
,
o
u
r
in
v
esti
g
atio
n
co
n
f
ir
m
s
th
at
ea
ch
r
ep
etitio
n
o
n
ly
n
ee
d
s
a
s
m
all
am
o
u
n
t
o
f
p
r
o
ce
s
s
in
g
tim
e
.
T
o
p
u
t
it
s
u
cc
in
ctly
,
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
ca
n
o
p
tim
ize
b
it
allo
ca
tio
n
at
lo
w
en
co
d
in
g
co
m
p
lex
ity
c
o
s
ts
.
T
h
u
s
,
b
y
in
c
o
r
p
o
r
atin
g
t
h
e
m
o
r
e
p
r
ec
is
e
R
-
D
esti
m
atio
n
,
o
u
r
s
y
s
tem
ca
n
all
o
ca
te
b
its
o
p
tim
ally
with
m
in
i
m
al
ad
d
itio
n
al
co
m
p
lex
ity
c
o
s
t.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Lo
w
co
mp
lexity
r
a
te
co
n
tr
o
l fo
r
ve
r
s
a
tile v
id
eo
co
d
in
g
w
ith
h
yb
r
id
…
(
A
lexa
n
d
er
V
icto
r
B
u
kit
)
1693
Du
e
to
ap
p
r
o
p
r
iate
b
it
allo
ca
tio
n
an
d
p
r
ec
is
e
Q
-
λ
r
elatio
n
s
h
ip
,
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
c
an
ac
h
iev
e
1
9
.
4
9
%
an
d
-
0
.
4
4
%
B
DB
R
s
av
in
g
s
wh
en
co
m
p
a
r
ed
t
o
th
e
VVC
r
ate
co
n
tr
o
l
tech
n
iq
u
e.
T
h
ese
f
in
d
in
g
s
o
f
f
er
s
ig
n
if
ican
t
p
r
o
o
f
o
f
th
e
s
u
g
g
ested
r
ate
co
n
tr
o
l
alg
o
r
ith
m
'
s
av
er
ag
e
ef
f
icac
y
i
n
co
m
p
ar
is
o
n
to
th
e
in
itial
R
DOQ
im
p
lem
en
tatio
n
in
VT
M.
T
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
o
r
ig
in
al
VT
M
a
n
d
o
u
r
r
ec
o
m
m
en
d
e
d
m
ajo
r
m
o
d
e
s
elec
tio
n
ar
e
co
m
p
ar
ed
i
n
th
is
s
u
b
s
ec
tio
n
.
T
o
ass
ess
th
e
o
u
tco
m
es
an
d
t
h
e
co
d
in
g
ti
m
e
r
ed
u
ctio
n
r
ate
in
r
elatio
n
to
th
e
a
n
ch
o
r
test
f
in
d
in
g
s
,
we
em
p
lo
y
e
d
th
e
B
DB
R
an
d
.
Sin
ce
it
r
eq
u
ir
es
as
litt
le
v
id
e
o
co
d
in
g
tim
e
as
p
o
s
s
ib
le
to
m
in
im
ize
d
elay
tim
e,
it
m
ay
p
r
o
f
it
f
r
o
m
a
co
n
s
id
er
ab
le
d
ec
r
ea
s
e
in
%
co
m
p
ar
ed
to
a
d
ec
r
ea
s
e
in
B
DB
R
th
at
wa
s
m
o
d
est
en
o
u
g
h
to
b
e
ap
p
lied
t
o
o
n
lin
e
an
d
r
ea
l
-
tim
e
v
id
e
o
co
m
m
u
n
icatio
n
.
T
ab
le
1
an
d
Fi
g
u
r
e
2
d
is
p
lay
th
e
test
r
esu
lts
u
s
in
g
th
e
s
u
g
g
ested
m
eth
o
d
.
T
h
e
av
e
r
ag
e
B
DB
R
%
was
-
0
.
4
4
%
an
d
th
e
a
v
er
ag
e
d
r
o
p
in
%
was
1
9
.
4
9
%.
T
ab
le
1
d
em
o
n
s
tr
ates
th
at
th
e
m
eth
o
d
m
ay
,
o
n
av
er
ag
e
,
s
av
e
1
9
.
4
9
%
o
f
th
e
en
co
d
in
g
tim
e
at
VT
M
1
4
.
2
with
ju
s
t
a
s
m
all
d
ec
r
ea
s
e
in
s
eq
u
en
ce
te
s
t
q
u
ality
.
T
o
s
u
m
u
p
,
all
QPs
h
av
e
u
n
if
o
r
m
p
r
o
ce
d
u
r
es,
a
n
d
th
e
av
e
r
ag
e
tim
e
s
av
in
g
s
o
v
er
VT
M
1
4
.
2
is
s
ig
n
if
ican
t.
T
h
e
r
esu
lts
in
d
icate
th
at
o
u
r
R
C
alg
o
r
ith
m
p
er
f
o
r
m
s
s
im
ilar
ly
to
th
e
r
ate
co
n
tr
o
l
o
f
th
e
a
n
ch
o
r
in
g
VT
M
1
4
.
2
p
latf
o
r
m
,
with
a
o
f
0
.
3
7
2
%.
B
ec
au
s
e
th
e
tim
e
s
av
in
g
s
an
d
B
DB
R
p
r
o
ce
s
s
ar
e
d
ir
ec
tly
as
s
o
ciate
d
with
th
e
tim
e
s
av
i
n
g
s
ac
r
o
s
s
th
e
en
tire
en
co
d
i
n
g
p
r
o
ce
s
s
,
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
e
n
ab
les
th
e
s
y
s
tem
to
s
av
e
a
f
air
am
o
u
n
t
o
f
tim
e
d
u
r
in
g
th
e
VT
M
1
4
.
2
en
co
d
in
g
p
r
o
ce
s
s
.
Fo
r
th
e
R
ac
eHo
r
s
e
s
eq
u
e
n
ce
,
th
e
r
ec
o
m
m
en
d
e
d
m
eth
o
d
l
o
wer
ed
th
e
e
n
co
d
i
n
g
tim
e
b
y
at
least
1
4
.
9
8
%,
an
d
f
o
r
th
e
Fo
u
r
Peo
p
le
s
eq
u
en
ce
,
it
r
ed
u
ce
d
it
b
y
b
etter
th
a
n
2
2
.
9
8
%.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
p
r
o
v
i
d
es
less
f
o
r
th
e
f
ast
-
m
o
tio
n
m
o
v
ie
s
eq
u
e
n
c
e
"RaceH
o
r
s
e,
"
b
u
t
it
p
er
f
o
r
m
s
well
f
o
r
v
id
eo
s
lik
e
"Fo
u
r
Peo
p
le,
"
wh
i
ch
in
clu
d
e
a
lo
t o
f
s
m
o
o
th
s
ec
tio
n
s
an
d
s
lo
w
m
o
tio
n
.
T
h
is
d
em
o
n
s
tr
ates th
at
th
e
p
r
o
p
o
s
ed
m
eth
o
d
is
ef
f
icien
t
a
n
d
u
n
i
v
er
s
al
in
r
ed
u
cin
g
co
d
i
n
g
co
m
p
lex
ity
an
d
m
ay
b
e
u
s
ed
to
an
y
s
eq
u
en
ce
.
Giv
en
th
at
it
ca
n
s
h
o
r
te
n
th
e
e
n
co
d
in
g
tim
e
an
d
p
r
o
d
u
ce
a
p
ar
titi
o
n
p
r
ed
ictio
n
r
esu
lt
th
at
i
s
alm
o
s
t
id
en
tical
to
th
e
o
r
ig
in
al
m
eth
o
d
o
lo
g
y
,
th
e
m
eth
o
d
p
r
o
p
o
s
ed
in
th
is
wo
r
k
ca
n
d
r
asti
ca
lly
cu
t
th
e
e
n
co
d
in
g
tim
e
wh
ile
m
ain
tain
in
g
en
c
o
d
in
g
q
u
ality
.
W
e
co
m
p
ar
e
th
e
p
r
o
p
o
s
ed
al
g
o
r
ith
m
t
o
o
th
e
r
s
tate
-
of
-
th
e
-
ar
t
wo
r
k
s
,
in
clu
d
in
g
t
h
e
Fu
et
a
l.
(
T
F)
alg
o
r
ith
m
i
n
[
1
9
]
,
C
u
i
et
a
l.
(
J
C
)
alg
o
r
ith
m
in
[
2
0
]
a
n
d
L
iao
et
a
l.
(
J
L
)
alg
o
r
ith
m
in
[
2
1
]
,
t
o
p
r
o
v
id
e
an
o
b
jectiv
e
co
m
p
ar
is
o
n
.
T
h
e
VT
M
v
id
eo
c
o
d
in
g
s
tan
d
ar
d
an
d
t
h
e
s
eq
u
en
ce
s
(
C
lass
A1
,
A2
,
B
,
C
,
D,
an
d
E
)
ar
e
u
s
ed
to
test
ea
ch
m
eth
o
d
.
T
h
e
av
er
ag
e
tim
e
s
av
in
g
s
(
%)
av
er
ag
e
is
th
e
m
etr
ic
u
s
ed
to
ev
alu
ate
th
e
alg
o
r
ith
m
'
s
p
er
f
o
r
m
a
n
ce
.
T
h
e
v
alu
e
o
f
,
as
d
is
p
lay
ed
in
Fig
u
r
e
3
,
in
d
icate
s
th
e
r
atio
o
f
r
ed
u
ctio
n
in
co
m
p
u
tatio
n
co
m
p
le
x
ity
.
Fig
u
r
e
3
.
T
h
e
co
m
p
ar
is
o
n
o
f
t
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
e
x
p
er
i
m
en
tal
av
er
ag
e
r
esu
lts
with
th
e
p
r
ev
io
u
s
m
eth
o
d
T
h
e
s
u
g
g
ested
alg
o
r
ith
m
o
u
tp
er
f
o
r
m
s
t
h
e
c
u
r
r
en
t
alg
o
r
ith
m
s
b
y
a
s
m
all
m
a
r
g
in
.
First,
o
u
r
alg
o
r
ith
m
o
u
tp
er
f
o
r
m
ed
th
e
alg
o
r
ith
m
i
n
[
1
9
]
in
ter
m
s
o
f
lo
wer
in
g
t
h
e
av
er
a
g
e
c
o
d
in
g
co
m
p
lex
it
y
.
Ad
d
itio
n
ally
,
o
u
r
alg
o
r
ith
m
'
s
en
co
d
in
g
tim
e
wa
s
8
.
4
9
%
s
h
o
r
ter
.
Nex
t,
o
u
r
s
tr
ateg
y
d
ec
r
ea
s
ed
th
e
e
n
co
d
in
g
tim
e
b
y
1
.
3
1
%
in
co
m
p
ar
is
o
n
to
[
2
0
]
.
Nex
t,
o
u
r
ap
p
r
o
ac
h
r
esu
lted
in
a
1
2
.
4
9
%
r
ed
u
ctio
n
in
en
co
d
i
n
g
tim
e
wh
en
co
m
p
ar
ed
to
[
2
1
]
.
T
o
th
e
b
est
o
f
o
u
r
k
n
o
wled
g
e,
th
e
lo
w
c
o
m
p
lex
ity
r
ate
co
n
tr
o
l
f
o
r
VVC
with
h
y
b
r
i
d
L
ag
r
a
n
g
e
m
u
ltip
lier
is
b
ein
g
p
r
o
p
o
s
ed
f
o
r
th
e
f
ir
s
t
tim
e
in
liter
atu
r
e.
E
ac
h
u
n
iq
u
e
s
eq
u
en
ce
is
co
n
s
is
ten
t,
th
u
s
th
er
e
will
b
e
v
ar
iatio
n
s
in
th
e
c
o
m
p
lex
it
y
r
ed
u
ctio
n
.
T
h
is
s
tr
ateg
y
wil
l
p
er
f
o
r
m
b
etter
with
s
tead
y
s
eq
u
en
ce
s
an
d
less
well
with
r
ap
id
l
y
c
h
an
g
in
g
s
e
q
u
en
ce
s
.
I
t
is
e
x
p
ec
ted
th
at
t
h
is
m
eth
o
d
will
ap
p
l
y
th
ese
r
esu
lts
co
n
s
is
ten
tly
to
all
v
id
eo
en
c
o
d
in
g
s
th
at
c
o
m
p
l
y
with
VT
M.
0
5
10
15
20
TF
JC
Pro
pos
ed
JL
(%)
Th
e
se
l
e
c
te
d
state
-
of
-
th
e
-
ar
t works
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
2
,
Ap
r
il
20
25
:
1
6
8
6
-
1
6
9
5
1694
T
h
e
lo
s
s
y
en
co
d
e
r
m
o
d
u
les
lik
e
q
u
an
tizatio
n
,
ar
e
th
e
s
o
u
r
c
e
o
f
th
e
co
d
in
g
d
is
to
r
tio
n
,
wh
i
ch
is
m
o
s
tly
d
eter
m
in
ed
b
y
th
e
co
d
in
g
s
ettin
g
s
.
I
t seem
s
th
at
s
ig
n
if
ican
t d
o
wn
s
am
p
lin
g
d
is
to
r
tio
n
im
p
lies
a
s
ig
n
if
ican
t lo
s
s
o
f
h
ig
h
-
f
r
e
q
u
en
c
y
in
f
o
r
m
atio
n
d
u
r
in
g
th
e
r
esam
p
lin
g
p
r
o
ce
s
s
.
T
h
e
en
co
d
er
lik
ely
f
av
o
r
s
f
u
ll
-
r
eso
lu
tio
n
co
d
in
g
in
th
is
s
itu
atio
n
.
On
th
e
o
th
er
h
an
d
,
wh
e
n
th
e
d
o
wn
s
am
p
lin
g
d
is
to
r
tio
n
is
lo
w,
d
o
wn
s
am
p
lin
g
-
b
ased
co
d
in
g
is
p
r
o
b
a
b
ly
g
o
in
g
to
p
er
f
o
r
m
b
etter
.
T
h
u
s
,
th
e
id
ea
l
co
d
in
g
r
eso
lu
tio
n
is
ascer
tain
ed
b
y
u
tili
zin
g
th
e
d
o
wn
s
am
p
lin
g
d
is
to
r
tio
n
.
5.
CO
NCLU
SI
O
N
E
x
p
er
im
en
tal
v
alid
atio
n
an
d
p
u
b
licatio
n
o
f
th
e
lo
w
co
m
p
l
ex
ity
r
ate
co
n
tr
o
l
f
o
r
VVC
w
ith
h
y
b
r
id
L
ag
r
an
g
e
m
u
ltip
lier
h
av
e
b
ee
n
co
m
p
leted
.
T
h
e
VT
M
1
4
.
2
(
VVC
tes
t
m
o
d
e)
s
tan
d
ar
d
is
u
s
ed
to
ce
r
tify
r
ea
l
-
tim
e
v
id
eo
co
m
m
u
n
icatio
n
.
T
h
e
o
p
tim
izatio
n
ca
n
r
ed
u
ce
th
e
co
d
in
g
tim
e
(
t%)
av
e
r
ag
e
b
y
1
9
.
4
9
%
b
y
u
s
in
g
th
e
p
r
o
p
o
s
ed
m
eth
o
d
.
T
h
is
is
a
s
ig
n
if
ican
t
d
ec
r
ea
s
e
co
m
p
ar
ed
to
th
e
d
r
o
p
p
in
g
B
DB
R
av
e
r
ag
e
lev
el,
wh
ic
h
is
s
o
litt
le
as to
b
e
in
s
ig
n
if
ican
t (
-
0
.
4
4
%).
T
h
e
test
r
esu
lts
d
em
o
n
s
tr
ate
t
h
at
th
e
co
d
e
n
ee
d
ed
f
o
r
r
ea
l
-
ti
m
e
v
id
eo
tr
an
s
m
is
s
io
n
m
ay
b
e
co
m
p
leted
in
a
s
h
o
r
ter
a
m
o
u
n
t
o
f
tim
e,
wh
ich
is
p
r
etty
g
o
o
d
.
I
t
is
ex
p
ec
ted
th
at
th
e
s
am
e
o
u
tco
m
es
wo
u
ld
h
o
ld
tr
u
e
f
o
r
o
th
er
s
eq
u
en
ce
s
as
well,
o
p
e
n
in
g
u
p
f
u
r
th
er
ap
p
licatio
n
o
p
tio
n
s
.
T
h
ese
s
p
ec
if
ic
f
i
n
d
in
g
s
o
f
f
e
r
p
r
o
m
is
in
g
m
eth
o
d
s
f
o
r
f
u
r
th
er
r
esear
c
h
an
d
im
p
lem
e
n
tatio
n
,
p
a
r
ticu
la
r
ly
in
co
m
p
lex
ity
r
e
d
u
ctio
n
v
id
eo
co
d
i
n
g
ef
f
o
r
ts
an
d
r
ate
co
n
tr
o
l e
n
h
an
ce
m
e
n
t.
ACK
NO
WL
E
DG
E
M
E
NT
S
T
h
e
Au
th
o
r
s
wo
u
ld
lik
e
to
ac
k
n
o
wled
g
e
th
e
s
u
p
p
o
r
t
p
r
o
v
id
ed
b
y
th
e
E
lectr
ical
E
n
g
in
ee
r
in
g
Dep
ar
tm
en
t,
Facu
lty
o
f
I
n
tell
ig
en
t
E
lectr
ical
an
d
I
n
f
o
r
m
at
ics
T
ec
h
n
o
lo
g
y
,
I
T
S
o
n
r
ese
ar
ch
f
ac
ilit
ies
an
d
f
in
an
cial
s
u
p
p
o
r
t d
u
r
in
g
t
h
e
co
m
p
letio
n
o
f
th
is
r
esear
ch
.
RE
F
E
R
E
NC
E
S
[
1
]
H
.
Y
u
a
n
,
Q
.
W
a
n
g
,
Q
.
L
i
u
,
J
.
H
u
o
,
a
n
d
P
.
L
i
,
“
H
y
b
r
i
d
d
i
s
t
o
r
t
i
o
n
-
b
a
sed
r
a
t
e
-
d
i
st
o
r
t
i
o
n
o
p
t
i
m
i
z
a
t
i
o
n
a
n
d
r
a
t
e
c
o
n
t
r
o
l
f
o
r
H
.
2
6
5
/
H
EV
C
,
”
I
E
EE
T
r
a
n
sa
c
t
i
o
n
s
o
n
C
o
n
s
u
m
e
r
El
e
c
t
ro
n
i
c
s
,
v
o
l
.
6
7
,
n
o
.
2
,
p
p
.
9
7
–
1
0
6
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
E.
2
0
2
1
.
3
0
6
5
6
3
6
.
[
2
]
Y
.
Li
,
Z
.
C
h
e
n
,
“
R
a
t
e
c
o
n
t
r
o
l
f
o
r
V
V
C
,
”
D
o
c
u
m
e
n
t
J
VET
-
K
0
3
9
0
,
J
u
l
y
.
2
0
1
8
.
A
c
c
e
ss
e
d
:
Ja
n
8
,
2
0
2
5
.
[
O
n
l
i
n
e
]
,
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
j
v
e
t
-
e
x
p
e
r
t
s.
o
r
g
/
d
o
c
_
e
n
d
_
u
ser/
d
o
c
u
me
n
t
s
/
1
1
_
Lj
u
b
l
j
a
n
a
/
w
g
1
1
/
JV
ET
-
K
0
3
9
0
-
v
5
.
z
i
p
[
3
]
Z.
L
i
u
,
Y
.
L
i
,
Z
.
C
h
e
n
,
“
A
d
a
p
t
i
v
e
λ
r
a
t
i
o
e
st
i
ma
t
i
o
n
f
o
r
r
a
t
e
c
o
n
t
r
o
l
i
n
V
V
C
,
”
D
o
c
u
m
e
n
t
J
V
ET
-
L
0
2
4
1
,
O
c
t
.
2
0
1
8
.
A
c
c
e
ss
e
d
:
J
a
n
8
,
2
0
2
5
.
[
O
n
l
i
n
e
]
,
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
j
v
e
t
-
e
x
p
e
r
t
s.
o
r
g
/
d
o
c
_
e
n
d
_
u
s
e
r
/
d
o
c
u
m
e
n
t
s
/
1
2
_
M
a
c
a
o
/
w
g
1
1
/
JV
E
T
-
L0
2
4
1
-
v
4
.
z
i
p
[
4
]
B
.
Li
,
H
.
Li
,
L.
Li
,
a
n
d
J
.
Zh
a
n
g
,
“
R
a
t
e
c
o
n
t
r
o
l
b
y
R
-
λ
m
o
d
e
l
f
o
r
H
EV
C
,
”
i
n
J
o
i
n
t
C
o
l
l
a
b
o
r
a
t
i
v
e
T
e
a
m
o
n
V
i
d
e
o
C
o
d
i
n
g
(
J
C
T
-
VC
)
o
f
I
T
U
-
T
S
G
1
6
WP
3
a
n
d
I
S
O
/
I
EC
J
T
C
1
/
S
C
2
9
/
WG1
1
,
J
C
T
V
C
-
K
0
1
0
3
,
2
0
1
2
.
A
c
c
e
ss
e
d
:
Ja
n
8
,
2
0
2
5
.
[
O
n
l
i
n
e
]
,
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
m
p
e
g
.
e
x
p
e
r
t
/
j
c
t
/
f
i
l
e
s/
J
C
TV
C
-
K
0
1
0
3
-
v
2
.
z
i
p
[
5
]
R
.
S
o
n
g
a
n
d
Y
.
Z
h
a
n
g
,
“
O
p
t
i
m
i
z
e
d
r
a
t
e
c
o
n
t
r
o
l
a
l
g
o
r
i
t
h
m o
f
h
i
g
h
-
e
f
f
i
c
i
e
n
c
y
v
i
d
e
o
c
o
d
i
n
g
b
a
s
e
d
o
n
r
e
g
i
o
n
o
f
i
n
t
e
r
e
st
,
”
J
o
u
rn
a
l
o
f
El
e
c
t
r
i
c
a
l
a
n
d
C
o
m
p
u
t
e
r
En
g
i
n
e
e
r
i
n
g
,
v
o
l
.
2
0
2
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
0
/
4
3
5
4
2
9
6
.
[
6
]
B
.
Li
,
H
.
Li
,
L.
L
i
,
a
n
d
J.
Z
h
a
n
g
,
“
A
d
a
p
t
i
v
e
b
i
t
a
l
l
o
c
a
t
i
o
n
f
o
r
R
-
λ
m
o
d
e
l
r
a
t
e
c
o
n
t
r
o
l
i
n
H
M
,
”
i
n
J
o
i
n
t
C
o
l
l
a
b
o
ra
t
i
v
e
T
e
a
m
o
n
Vi
d
e
o
C
o
d
i
n
g
(
J
C
T
-
VC
)
o
f
I
T
U
-
T
S
G
1
6
WP
3
a
n
d
I
S
O
/
I
EC
J
T
C
1
/
S
C
2
9
/
WG1
1
,
J
C
T
VC
-
M
0
0
3
6
,
2
0
1
3
.
A
c
c
e
s
sed
:
J
a
n
8
,
2
0
2
5
.
[
O
n
l
i
n
e
]
,
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
m
p
e
g
.
e
x
p
e
r
t
/
j
c
t
/
f
i
l
e
s
/
JC
TV
C
-
M
0
0
3
6
-
v
4
.
z
i
p
[
7
]
H
.
W
e
i
,
W
.
Z
h
o
u
,
X
.
Z
h
o
u
,
a
n
d
Z.
D
u
a
n
,
“
S
u
b
-
f
r
a
me
l
a
y
e
r
r
a
t
e
d
i
s
t
o
r
t
i
o
n
o
p
t
i
m
i
z
a
t
i
o
n
f
o
r
h
i
g
h
e
f
f
i
c
i
e
n
c
y
v
i
d
e
o
c
o
d
i
n
g
,
”
I
EE
E
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
5
3
1
1
6
–
5
3
1
3
2
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
8
1
1
4
3
.
[
8
]
L
.
L
i
,
B
.
L
i
,
H
.
L
i
,
a
n
d
C
.
W
.
C
h
e
n
,
“
λ
-
d
o
m
a
i
n
o
p
t
i
m
a
l
b
i
t
a
l
l
o
c
a
t
i
o
n
a
l
g
o
r
i
t
h
m
f
o
r
h
i
g
h
e
f
f
i
c
i
e
n
c
y
v
i
d
e
o
c
o
d
i
n
g
,
”
I
E
E
E
T
r
a
n
s
a
c
t
i
o
n
s
o
n
C
i
r
c
u
i
t
s
a
n
d
S
y
s
t
e
m
s
f
o
r
V
i
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
8
,
n
o
.
1
,
p
p
.
1
3
0
–
1
4
2
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
S
V
T
.
2
0
1
6
.
2
5
9
8
6
7
2
.
[
9
]
C
.
H
.
K
u
o
,
Y
.
L.
S
h
i
h
,
a
n
d
S
.
C
.
Y
a
n
g
,
“
R
a
t
e
c
o
n
t
r
o
l
v
i
a
a
d
j
u
st
me
n
t
o
f
l
a
g
r
a
n
g
e
mu
l
t
i
p
l
i
e
r
f
o
r
v
i
d
e
o
c
o
d
i
n
g
,
”
I
E
EE
T
r
a
n
sa
c
t
i
o
n
s
o
n
C
i
rc
u
i
t
s
a
n
d
S
y
st
e
m
s
f
o
r
Vi
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
6
,
n
o
.
1
1
,
p
p
.
2
0
6
9
–
2
0
7
8
,
2
0
1
6
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
S
V
T.
2
0
1
5
.
2
5
0
1
9
2
1
.
[
1
0
]
F
.
R
a
u
f
me
h
r
,
M
.
R
.
S
a
l
e
h
i
,
a
n
d
E
.
A
b
i
r
i
,
“
A
f
r
a
me
-
l
e
v
e
l
M
LP
-
b
a
s
e
d
b
i
t
-
r
a
t
e
c
o
n
t
r
o
l
l
e
r
f
o
r
r
e
a
l
-
t
i
me
v
i
d
e
o
t
r
a
n
sm
i
ssi
o
n
u
si
n
g
V
V
C
st
a
n
d
a
r
d
,
”
J
o
u
rn
a
l
o
f
R
e
a
l
-
T
i
m
e
I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
v
o
l
.
1
8
,
n
o
.
3
,
p
p
.
7
5
1
–
7
6
3
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
5
5
4
-
0
2
0
-
0
1
0
1
8
-
2.
[
1
1
]
M
.
W
a
n
g
,
K
.
N
.
N
g
a
n
,
a
n
d
H
.
Li
,
“
L
o
w
-
d
e
l
a
y
r
a
t
e
c
o
n
t
r
o
l
f
o
r
c
o
n
s
i
st
e
n
t
q
u
a
l
i
t
y
u
si
n
g
d
i
st
o
r
t
i
o
n
-
b
a
se
d
l
a
g
r
a
n
g
e
mu
l
t
i
p
l
i
e
r
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
m
a
g
e
Pro
c
e
ss
i
n
g
,
v
o
l
.
2
5
,
n
o
.
7
,
p
p
.
2
9
4
3
–
2
9
5
5
,
2
0
1
6
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
P
.
2
0
1
6
.
2
5
5
2
6
4
6
.
[
1
2
]
S
.
L
i
,
M
.
X
u
,
Z.
W
a
n
g
,
a
n
d
X
.
S
u
n
,
“
O
p
t
i
mal
b
i
t
a
l
l
o
c
a
t
i
o
n
f
o
r
C
TU
l
e
v
e
l
r
a
t
e
c
o
n
t
r
o
l
i
n
H
EV
C
,
”
I
EEE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
C
i
r
c
u
i
t
s
a
n
d
S
y
s
t
e
m
s
f
o
r V
i
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
7
,
n
o
.
1
1
,
p
p
.
2
4
0
9
–
2
4
2
4
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
S
V
T.
2
0
1
6
.
2
5
8
9
8
7
8
.
[
1
3
]
Z.
C
h
e
n
a
n
d
X
.
P
a
n
,
“
A
n
o
p
t
i
mi
z
e
d
r
a
t
e
c
o
n
t
r
o
l
f
o
r
l
o
w
-
d
e
l
a
y
H
.
2
6
5
/
H
EV
C
,
”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
v
o
l
.
2
8
,
n
o
.
9
,
p
p
.
4
5
4
1
–
4
5
5
2
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
P
.
2
0
1
9
.
2
9
1
1
1
8
0
.
[
1
4
]
X
.
P
a
n
a
n
d
Z.
C
h
e
n
,
“
M
u
l
t
i
-
l
a
y
e
r
q
u
a
n
t
i
z
a
t
i
o
n
c
o
n
t
r
o
l
f
o
r
q
u
a
l
i
t
y
-
c
o
n
s
t
r
a
i
n
e
d
H
.
2
6
5
/
H
EV
C
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
m
a
g
e
Pro
c
e
ssi
n
g
,
v
o
l
.
2
6
,
n
o
.
7
,
p
p
.
3
4
3
7
–
3
4
4
8
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
P
.
2
0
1
6
.
2
6
2
7
8
1
8
.
[
1
5
]
T.
C
h
i
a
n
g
a
n
d
Y
.
Q
.
Zh
a
n
g
,
“
A
n
e
w
r
a
t
e
c
o
n
t
r
o
l
sc
h
e
m
e
u
si
n
g
q
u
a
d
r
a
t
i
c
r
a
t
e
d
i
st
o
r
t
i
o
n
mo
d
e
l
,
”
I
EE
E
T
r
a
n
s
a
c
t
i
o
n
s
o
n
C
i
rc
u
i
t
s
a
n
d
S
y
s
t
e
m
s
f
o
r V
i
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
7
,
n
o
.
1
,
p
p
.
2
4
6
–
2
5
0
,
1
9
9
7
,
d
o
i
:
1
0
.
1
1
0
9
/
7
6
.
5
5
4
4
3
9
.
[
1
6
]
S
.
M
a
,
W
.
G
a
o
,
a
n
d
Y
.
Lu
,
“
R
a
t
e
-
d
i
st
o
r
t
i
o
n
a
n
a
l
y
si
s
f
o
r
H
.
2
6
4
/
A
V
C
v
i
d
e
o
c
o
d
i
n
g
a
n
d
i
t
s
a
p
p
l
i
c
a
t
i
o
n
t
o
r
a
t
e
c
o
n
t
r
o
l
,
”
I
E
EE
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
i
rc
u
i
t
s
a
n
d
S
y
s
t
e
m
s
f
o
r
V
i
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
5
,
n
o
.
1
2
,
p
p
.
1
5
3
3
–
1
5
4
3
,
2
0
0
5
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
S
V
T.
2
0
0
5
.
8
5
7
3
0
0
.
[
1
7
]
D
.
K
.
K
w
o
n
,
M
.
Y
.
S
h
e
n
,
a
n
d
C
.
C
.
J
a
y
K
u
o
,
“
R
a
t
e
c
o
n
t
r
o
l
f
o
r
H
.
2
6
4
v
i
d
e
o
w
i
t
h
e
n
h
a
n
c
e
d
r
a
t
e
a
n
d
d
i
s
t
o
r
t
i
o
n
m
o
d
e
l
s
,
”
I
E
E
E
T
r
a
n
s
a
c
t
i
o
n
s
o
n
C
i
r
c
u
i
t
s
a
n
d
S
y
s
t
e
m
s
f
o
r
V
i
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
7
,
n
o
.
5
,
p
p
.
5
1
7
–
5
2
8
,
2
0
0
7
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
S
V
T
.
2
0
0
7
.
8
9
4
0
5
3
.
[
1
8
]
C
.
H
.
K
u
o
,
L.
C
.
C
h
a
n
g
,
H
.
C
.
C
h
i
u
,
a
n
d
B
.
D
.
L
i
u
,
“
Ef
f
i
c
i
e
n
t
t
w
o
-
p
a
ss
r
a
t
e
c
o
n
t
r
o
l
sc
h
e
me
b
a
se
d
o
n
a
d
j
u
st
i
n
g
d
i
s
t
r
i
b
u
t
i
o
n
o
f
d
i
s
c
r
e
t
e
c
o
s
i
n
e
t
r
a
n
sf
o
r
m
c
o
e
f
f
i
c
i
e
n
t
s
,
”
I
ET I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
v
o
l
.
4
,
n
o
.
3
,
p
p
.
2
1
1
–
2
2
2
,
2
0
1
0
,
d
o
i
:
1
0
.
1
0
4
9
/
i
e
t
-
i
p
r
.
2
0
0
9
.
0
0
8
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Lo
w
co
mp
lexity
r
a
te
co
n
tr
o
l fo
r
ve
r
s
a
tile v
id
eo
co
d
in
g
w
ith
h
yb
r
id
…
(
A
lexa
n
d
er
V
icto
r
B
u
kit
)
1695
[
1
9
]
T.
F
u
,
K
.
Z
h
a
n
g
,
L
.
Z
h
a
n
g
,
S
.
W
a
n
g
,
a
n
d
S
.
M
a
,
“
R
a
t
e
c
o
n
t
r
o
l
w
i
t
h
r
e
so
l
u
t
i
o
n
c
h
a
n
g
e
s
f
o
r
V
V
C
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
-
I
EEE
I
n
t
e
r
n
a
t
i
o
n
a
l
S
y
m
p
o
si
u
m
o
n
C
i
rc
u
i
t
s
a
n
d
S
y
s
t
e
m
s
,
2
0
2
3
,
v
o
l
.
2
0
2
3
-
M
a
y
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
S
C
A
S
4
6
7
7
3
.
2
0
2
3
.
1
0
1
8
1
9
5
1
.
[
2
0
]
J
.
C
u
i
,
S
.
W
a
n
g
,
S
.
W
a
n
g
,
X
.
Z
h
a
n
g
,
S
.
M
a
,
a
n
d
W
.
G
a
o
,
“
H
y
b
r
i
d
l
a
p
l
a
c
e
d
i
s
t
r
i
b
u
t
i
o
n
-
b
a
s
e
d
l
o
w
c
o
m
p
l
e
x
i
t
y
r
a
t
e
-
d
i
s
t
o
r
t
i
o
n
o
p
t
i
m
i
z
e
d
q
u
a
n
t
i
z
a
t
i
o
n
,
”
I
E
E
E
T
r
a
n
s
a
c
t
i
o
n
s
o
n
I
m
a
g
e
P
r
o
c
e
s
s
i
n
g
,
v
o
l
.
2
6
,
n
o
.
8
,
p
p
.
3
8
0
2
–
3
8
1
6
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
T
I
P
.
2
0
1
7
.
2
7
0
3
1
1
2
.
[
2
1
]
J.
Li
a
o
,
L.
L
i
,
D
.
L
i
u
,
a
n
d
H
.
Li
,
“
C
o
n
t
e
n
t
-
a
d
a
p
t
i
v
e
r
a
t
e
-
d
i
s
t
o
r
t
i
o
n
mo
d
e
l
i
n
g
f
o
r
f
r
a
me
-
l
e
v
e
l
r
a
t
e
c
o
n
t
r
o
l
i
n
v
e
r
sat
i
l
e
v
i
d
e
o
c
o
d
i
n
g
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
Mu
l
t
i
m
e
d
i
a
,
v
o
l
.
2
6
,
p
p
.
6
8
6
4
–
6
8
7
9
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
TM
M
.
2
0
2
4
.
3
3
5
8
0
6
3
.
[
2
2
]
Y
.
Li
,
Z.
Li
u
,
Z.
C
h
e
n
,
a
n
d
S
.
L
i
u
,
“
R
a
t
e
c
o
n
t
r
o
l
f
o
r
v
e
r
sa
t
i
l
e
v
i
d
e
o
c
o
d
i
n
g
,
”
Pr
o
c
e
e
d
i
n
g
s
-
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
m
a
g
e
Pro
c
e
ssi
n
g
,
I
C
I
P
,
v
o
l
.
2
0
2
0
-
O
c
t
o
b
,
p
p
.
1
1
7
6
–
1
1
8
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
P
4
0
7
7
8
.
2
0
2
0
.
9
1
9
1
1
2
5
.
[
2
3
]
M
.
H
.
H
y
u
n
,
B
.
Le
e
,
a
n
d
M
.
K
i
m,
“
A
f
r
a
me
-
l
e
v
e
l
c
o
n
s
t
a
n
t
b
i
t
-
r
a
t
e
c
o
n
t
r
o
l
u
si
n
g
r
e
c
u
r
s
i
v
e
b
a
y
e
s
i
a
n
e
st
i
ma
t
i
o
n
f
o
r
v
e
r
sa
t
i
l
e
v
i
d
e
o
c
o
d
i
n
g
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
2
2
7
2
5
5
–
2
2
7
2
6
9
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
3
0
4
6
0
4
3
.
[
2
4
]
Z.
G
.
L
i
e
t
a
l
.
,
“
A
d
a
p
t
i
v
e
r
a
t
e
c
o
n
t
r
o
l
f
o
r
H
.
2
6
4
,
”
J
o
u
r
n
a
l
o
f
V
i
s
u
a
l
C
o
m
m
u
n
i
c
a
t
i
o
n
a
n
d
I
m
a
g
e
R
e
p
r
e
se
n
t
a
t
i
o
n
,
v
o
l
.
1
7
,
n
o
.
2
,
p
p
.
3
7
6
–
4
0
6
,
2
0
0
6
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
v
c
i
r
.
2
0
0
5
.
0
4
.
0
0
4
.
[
2
5
]
B
.
L
i
,
H
.
Li
,
L.
L
i
,
a
n
d
J.
Zh
a
n
g
,
“
λ
d
o
ma
i
n
r
a
t
e
c
o
n
t
r
o
l
a
l
g
o
r
i
t
h
m
f
o
r
h
i
g
h
e
f
f
i
c
i
e
n
c
y
v
i
d
e
o
c
o
d
i
n
g
,
”
I
EEE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
I
m
a
g
e
Pro
c
e
ssi
n
g
,
v
o
l
.
2
3
,
n
o
.
9
,
p
p
.
3
8
4
1
–
3
8
5
4
,
2
0
1
4
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
P
.
2
0
1
4
.
2
3
3
6
5
5
0
.
[
2
6
]
X
.
W
e
i
,
M
.
Z
h
o
u
,
H
.
W
a
n
g
,
H
.
Y
a
n
g
,
L
.
C
h
e
n
,
a
n
d
S
.
K
w
o
n
g
,
“
R
e
c
e
n
t
a
d
v
a
n
c
e
s
i
n
r
a
t
e
c
o
n
t
r
o
l
:
f
r
o
m
o
p
t
i
mi
z
a
t
i
o
n
t
o
i
mp
l
e
m
e
n
t
a
t
i
o
n
a
n
d
b
e
y
o
n
d
,
”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
i
rc
u
i
t
s
a
n
d
S
y
st
e
m
s
f
o
r
Vi
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
3
4
,
n
o
.
1
,
p
p
.
1
7
–
3
3
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
TC
S
V
T
.
2
0
2
3
.
3
2
8
7
5
6
1
.
[
2
7
]
Y
.
M
a
o
,
M
.
W
a
n
g
,
S
.
W
a
n
g
,
a
n
d
S
.
K
w
o
n
g
,
“
H
i
g
h
e
f
f
i
c
i
e
n
c
y
r
a
t
e
c
o
n
t
r
o
l
f
o
r
v
e
r
sa
t
i
l
e
v
i
d
e
o
c
o
d
i
n
g
b
a
se
d
o
n
c
o
mp
o
s
i
t
e
c
a
u
c
h
y
d
i
s
t
r
i
b
u
t
i
o
n
,
”
I
E
EE
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
i
rc
u
i
t
s
a
n
d
S
y
st
e
m
s
f
o
r
Vi
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
3
2
,
n
o
.
4
,
p
p
.
2
3
7
1
–
2
3
8
4
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
S
V
T.
2
0
2
1
.
3
0
9
3
3
1
5
.
[
2
8
]
F
.
L
i
u
a
n
d
Z.
C
h
e
n
,
“
M
u
l
t
i
-
o
b
j
e
c
t
i
v
e
o
p
t
i
m
i
z
a
t
i
o
n
o
f
q
u
a
l
i
t
y
i
n
V
V
C
r
a
t
e
c
o
n
t
r
o
l
f
o
r
l
o
w
-
d
e
l
a
y
v
i
d
e
o
c
o
d
i
n
g
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
m
a
g
e
Pro
c
e
ss
i
n
g
,
v
o
l
.
3
0
,
p
p
.
4
7
0
6
–
4
7
1
8
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
P
.
2
0
2
1
.
3
0
7
2
2
2
5
.
[
2
9
]
H
.
E
v
e
r
e
t
t
I
I
I
,
“
G
e
n
e
r
a
l
i
z
e
d
l
a
g
r
a
n
g
e
mu
l
t
i
p
l
i
e
r
met
h
o
d
f
o
r
s
o
l
v
i
n
g
p
r
o
b
l
e
ms
o
f
o
p
t
i
m
u
m
a
l
l
o
c
a
t
i
o
n
o
f
r
e
so
u
r
c
e
s,
”
O
p
e
r
a
t
i
o
n
s
Re
se
a
rc
h
,
v
o
l
.
1
1
,
n
o
.
3
,
p
p
.
3
9
9
–
4
1
7
,
1
9
6
3
,
d
o
i
:
1
0
.
1
2
8
7
/
o
p
r
e
.
1
1
.
3
.
3
9
9
.
[
3
0
]
Y
.
H
.
Li
n
,
C
.
Y
.
C
h
e
n
,
a
n
d
C
.
W
.
Ta
n
g
,
“
V
V
C
b
a
se
d
r
a
t
e
c
o
n
t
r
o
l
u
s
i
n
g
S
K
I
P
C
TU
p
r
e
d
i
c
t
o
r
,
”
i
n
2
0
2
2
I
E
EE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
n
su
m
e
r E
l
e
c
t
r
o
n
i
c
s
-
Asi
a
,
I
C
C
E
-
As
i
a
2
0
2
2
,
2
0
2
2
,
p
p
.
1
–
4
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
E
-
A
si
a
5
7
0
0
6
.
2
0
2
2
.
9
9
5
4
8
0
2
.
[
3
1
]
Y
.
W
a
n
g
,
P
.
D
a
i
,
J
.
Z
h
a
o
,
a
n
d
Q
.
Z
h
a
n
g
,
“
F
a
st
C
U
p
a
r
t
i
t
i
o
n
d
e
c
i
si
o
n
a
l
g
o
r
i
t
h
m
f
o
r
V
V
C
i
n
t
r
a
c
o
d
i
n
g
u
s
i
n
g
a
n
M
ET
-
C
N
N
,
”
El
e
c
t
r
o
n
i
c
s
(
S
w
i
t
zer
l
a
n
d
)
,
v
o
l
.
1
1
,
n
o
.
1
9
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s
1
1
1
9
3
0
9
0
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Ale
x
a
n
d
e
r
Vic
t
o
r
Bu
k
it
re
c
e
iv
e
d
t
h
e
b
a
c
h
e
lo
r
d
e
g
re
e
i
n
e
le
c
tri
c
a
l
e
n
g
i
n
e
e
rin
g
fro
m
Un
iv
e
rsity
o
f
S
u
m
a
tera
Uta
ra
,
M
e
d
a
n
,
In
d
o
n
e
sia
,
th
e
m
a
ste
r
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
,
tele
c
o
m
m
u
n
ica
ti
o
n
,
fro
m
In
stit
u
t
Tek
n
o
lo
g
i
S
e
p
u
lu
h
No
p
e
m
b
e
r,
S
u
ra
b
a
y
a
,
In
d
o
n
e
sia
in
2
0
1
7
.
C
u
rre
n
tl
y
,
h
e
is
c
o
m
p
leti
n
g
h
is
P
h
.
D.
in
tel
e
c
o
m
m
u
n
ica
ti
o
n
,
a
t
I
n
stit
u
t
Tek
n
o
lo
g
i
S
e
p
u
l
u
h
No
p
e
m
b
e
r
,
S
u
ra
b
a
y
a
,
I
n
d
o
n
e
sia
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
tele
c
o
m
m
u
n
ica
ti
o
n
p
lan
n
in
g
,
v
id
e
o
c
o
d
i
n
g
a
n
d
o
p
t
imiz
a
ti
o
n
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
7
0
2
2
2
1
1
0
0
7
@s
tu
d
e
n
t.
it
s.a
c
.
id
.
S
u
wa
d
i
re
c
e
iv
e
d
t
h
e
b
a
c
h
e
lo
r
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
i
n
e
e
rin
g
fro
m
I
n
stit
u
t
Tek
n
o
lo
g
i
S
e
p
u
lu
h
N
o
p
e
m
b
e
r,
S
u
ra
b
a
y
a
,
In
d
o
n
e
sia
in
1
9
9
2
,
t
h
e
m
a
ste
r
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
fro
m
I
n
stit
u
t
Tek
n
o
l
o
g
i
Ba
n
d
u
n
g
,
In
d
o
n
e
sia
,
in
1
9
9
9
,
a
n
d
t
h
e
d
o
c
t
o
ra
l
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
fro
m
I
n
sti
tu
t
Te
k
n
o
lo
g
i
S
e
p
u
l
u
h
No
p
e
m
b
e
r,
S
u
ra
b
a
y
a
,
In
d
o
n
e
sia
i
n
2
0
1
2
.
C
u
rre
n
tl
y
,
h
e
is
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
t
h
e
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
In
stit
u
t
Tek
n
o
l
o
g
i
S
e
p
u
l
u
h
No
p
e
m
b
e
r,
S
u
ra
b
a
y
a
,
I
n
d
o
n
e
sia
.
Hi
s
re
se
a
rc
h
in
tere
sts
in
c
l
u
d
e
d
ig
it
a
l
c
o
m
m
u
n
ica
ti
o
n
s,
u
n
d
e
rwa
ter
a
c
o
u
stic
c
o
m
m
u
n
ica
ti
o
n
,
d
i
g
i
t
a
l
sig
n
a
l
p
r
o
c
e
ss
in
g
,
ima
g
e
a
n
d
v
id
e
o
p
ro
c
e
ss
in
g
,
s
o
ftwa
re
d
e
fin
e
d
ra
d
io
a
n
d
b
ro
a
d
b
a
n
d
wire
l
e
ss
c
o
m
m
u
n
ica
ti
o
n
s.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
s
u
wa
d
i@e
e
.
it
s.a
c
.
id
.
Wira
wa
n
re
c
e
iv
e
d
t
h
e
B.
E
.
d
e
g
re
e
in
tele
c
o
m
m
u
n
ica
ti
o
n
e
n
g
in
e
e
rin
g
f
ro
m
In
stit
u
t
Tek
n
o
lo
g
i
S
e
p
u
lu
h
N
o
p
e
m
b
e
r
(IT
S
),
S
u
ra
b
a
y
a
,
In
d
o
n
e
sia
,
in
1
9
8
7
,
th
e
DEA
i
n
sig
n
a
l
a
n
d
ima
g
e
p
ro
c
e
ss
in
g
fr
o
m
Eco
le
S
u
p
é
rieu
re
e
n
S
c
ien
c
e
s
In
fo
rm
a
ti
q
u
e
s,
S
o
p
h
ia
An
t
ip
o
li
s,
a
n
d
th
e
Dr.
d
e
g
re
e
in
ima
g
e
p
ro
c
e
s
sin
g
fro
m
Tele
c
o
m
P
a
risTec
h
(
p
re
v
io
u
sly
Eco
le
Na
ti
o
n
a
le
S
u
p
é
rieu
re
d
e
s
Télé
c
o
m
m
u
n
ica
ti
o
n
s),
P
a
ris,
F
ra
n
c
e
,
in
1
9
9
6
a
n
d
2
0
0
3
,
re
sp
e
c
ti
v
e
ly
.
S
in
c
e
1
9
8
9
h
e
h
a
s
b
e
e
n
wi
th
IT
S
a
s
le
c
tu
re
r
in
t
h
e
El
e
c
tri
c
a
l
En
g
in
e
e
ri
n
g
De
p
a
rtme
n
t.
His
re
se
a
rc
h
in
tere
sts
li
e
in
th
e
g
e
n
e
ra
l
a
re
a
o
f
sta
ti
stica
l
sig
n
a
l
p
ro
c
e
ss
in
g
,
a
n
d
re
c
e
n
tl
y
f
o
c
u
se
s
m
o
re
o
n
u
n
d
e
rwa
ter
a
c
o
u
stic
c
o
m
m
u
n
ica
ti
o
n
a
n
d
n
e
two
r
k
in
g
a
n
d
o
n
v
a
rio
u
s
a
sp
e
c
ts
o
f
wire
les
s
se
n
so
r
n
e
two
rk
s.
He
is a
m
e
m
b
e
r
o
f
IEE
E.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
wira
wa
n
@its.
a
c
.
id
.
Evaluation Warning : The document was created with Spire.PDF for Python.