International Journal of e-Collaboration,
Journal Year:
2024,
Volume and Issue:
20(1), P. 1 - 16
Published: May 17, 2024
This
paper
proposes
a
neural
network
model
based
on
deep
learning,
which
can
better
analyze
the
impact
of
short
video
live
broadcast
sports
machinery
and
equipment.
Firstly,
this
U-Net-based
convolutional
as
backbone
paper,
mainly
realizes
sales.
Secondly,
dense
residual
module
Transformer
lightweight
module,
effectively
improve
global
modeling
ability
prediction
accuracy
model.
Finally,
through
large
number
experiments,
it
is
proved
that
U-Net
proposed
in
be
used
for
task
sales
equipment,
achieves
reasoning
speed.
IEEE Access,
Journal Year:
2024,
Volume and Issue:
12, P. 105165 - 105177
Published: Jan. 1, 2024
Underwater
target
detection
has
developed
greatly
in
recent
years.
However,
the
accuracy
of
underwater
is
limited
by
complex
environment.
Based
on
YOLOv7,
we
propose
an
object
algorithm
model
to
improve
precision
and
confidence
(BiFusion
Neck
module
a
MPDIoU
loss
function).
Compared
traditional
networks
module,
Bifusion
preserves
more
features
from
lower
layers
utilizing
output
P2
feature
layer.
Moreover,
function
was
improved
basis
IoU
introducing
Minimum
Point
Distance.
Finally,
LSKA
attention
mechanism
introduced
enhance
extraction
targets
at
different
scales.
The
experimental
results
demonstrate
that
BFD-YOLO
proposed
this
study
achieves
average
accuracy(mAP50)
84.8%
customized
dataset,
surpassing
performance
YOLOv7
11.5%
outperforming
other
tested
algorithms.
Furthermore,
exhibits
strong
various
datasets
demonstrates
superior
generalization
capabilities.
Electronics,
Journal Year:
2024,
Volume and Issue:
13(10), P. 1986 - 1986
Published: May 19, 2024
This
paper
examines
the
quantitative
and
qualitative
situation
of
current
fixed
mobile
Internet
its
expected
future.
It
provides
a
detailed
insight
into
past,
present,
future
along
with
development
technology
problems
that
have
arisen
in
accessing
using
broadband
Internet.
First,
number
users
penetration
rate
Internet,
various
types
services
different
countries,
ranking
countries
terms
mean
median
download
upload
data
speeds,
volume,
location
centers
world
are
presented.
The
second
task
introduces
details
twelve
performance
evaluation
metrics
for
access.
Third,
wired
wireless
technologies
introduced
compared
based
on
rate,
coverage,
type
infrastructure,
their
advantages
disadvantages.
Based
technical
functional
criteria,
fourth
work,
two
popular
platforms,
one
optical
fiber
other
5G
cellular
network,
general
Australia
particular.
Moreover,
this
has
look
at
Starlink
as
latest
satellite
candidate,
especially
rural
remote
areas.
fifth
outlines
emerging
Internet-based
applications
spotlight.
Sixthly,
it
focuses
three
world,
namely
digital
divide
due
to
qualities
available
new
technologies,
impact
social
interactions,
hacking
insecurity
Finally,
some
solutions
these
proposed.
Kurdistan Journal of Applied Research,
Journal Year:
2024,
Volume and Issue:
9(2), P. 23 - 41
Published: Sept. 22, 2024
The
transmission
of
real-time
videos
over
wireless
networks
is
prone
to
the
negative
consequences
packet
loss
and
delay,
which
can
have
a
potential
effect
on
video
quality
during
streaming.
These
impairments
lead
interruptions,
buffering,
degradation
visual
auditory
elements,
resulting
in
an
unsatisfactory
user
experience.
In
this
paper,
we
aim
address
challenges
associated
with
delay
parameters
networks,
propose
approach
alleviate
their
impact
transmission.
proposed
involves
utilizing
H.265/H.266
coding
standards.
For
Versatile
Video
Coding
(VVC),
patch
support
for
VVdeC
VVenC
FFmpeg
(Fast
Forward
Moving
Picture
Expert
Group)
added.
As
result,
used
encode,
stream
decode
all
videos.
Raw
2K
qualities
are
encoded
based
adaptive
quantization
(QP)
above-mentioned
codecs.
By
selecting
optimal
data
various
network
conditions,
enhances
Quality
Experience
(QoE)
end-users
while
minimizing
resource
usage
network.
Furthermore,
selects
codec
standards
according
bitrates
frame
rates.
Simulation
results
indicate
that
has
significant
improvement
streaming
satisfy
end
also
outperforms
other
related
work
by
gaining
PSNR
+12
dB
H.265
+13
H.266
when
1%.
