Radio
Frequency
Identification
(RFID),
as
one
of
the
core
technologies
in
field
Internet
Things
(IoT),
has
emerged
a
significant
medium
for
`passive
perception'
due
to
its
lightweight,
taggable,
and
easily
deployable
characteristics.
RFID
found
extensive
applications
people's
daily
production
life,
including
logistics
tracking,
target
detection,
item
identification.
Nonetheless,
systems
are
vulnerable
environmental
influences,
which
impact
system
performance.
This
paper
presents
two-level
weighted
multipath
interference
suppression
method
(TMS)
address
issue
systems.
Firstly,
RF
signal
propagation
model
is
established.
Secondly,
received
decomposed
acquire
reflection
signals
object
reflections.
Finally,
proposed
enhance
suppress
signal.
To
validate
effectiveness
method,
we
employed
experiment
detecting
concentration
white
wine
wine.
Subsequently,
extracted
`clean'
feature
values
inputted
them
into
CNN
based
on
hybrid
attention
mechanism
train
detection
model.
The
experimental
results
demonstrate
that
accuracy
reached
97.8\%,
while
96.8\%.
Compared
other
methods,
our
approach
exhibits
advantages
terms
enables
non-destructive
items.
Advanced Functional Materials,
Год журнала:
2024,
Номер
34(46)
Опубликована: Июль 18, 2024
Abstract
Ion‐conductive
elastomers
capable
of
damping
can
significantly
mitigate
the
interference
caused
by
mechanical
noise
during
data
acquisition
in
wearable
and
biomedical
devices.
However,
currently
available
often
lack
robust
properties
have
a
narrow
temperature
range
for
effective
damping.
Here,
precise
modulation
weak
to
strong
ion‐dipole
interactions
plays
crucial
role
bolstering
network
stability
tuning
relaxation
behavior
supramolecular
ion‐conductive
(SICEs).
The
SICEs
exhibit
impressive
properties,
including
modulus
13.2
MPa,
toughness
65.6
MJ
m
−3
,
fracture
energy
74.9
kJ
−2
.
Additionally,
they
demonstrate
remarkable
capabilities,
with
capacity
91.2%
peak
tan
δ
1.11.
Furthermore,
entropy‐driven
rearrangement
ensures
SICE
remain
stable
even
at
elevated
temperatures
(18–200
°C,
>
0.3),
making
it
most
thermally
resistant
elastomer
reported
date.
Moreover,
proves
filtering
out
various
noises
physiological
signal
detection
strain
sensing,
highlighting
its
vast
potential
flexible
electronics.
Bioengineering,
Год журнала:
2023,
Номер
10(11), С. 1324 - 1324
Опубликована: Ноя. 16, 2023
Accurate
and
real-time
gesture
recognition
is
required
for
the
autonomous
operation
of
prosthetic
hand
devices.
This
study
employs
a
convolutional
neural
network-enhanced
channel
attention
(CNN-ECA)
model
to
provide
unique
approach
surface
electromyography
(sEMG)
recognition.
The
introduction
ECA
module
improves
model's
capacity
extract
features
focus
on
critical
information
in
sEMG
data,
thus
simultaneously
equipping
sEMG-controlled
systems
with
characteristics
accurate
detection
control.
Furthermore,
we
suggest
preprocessing
strategy
extracting
envelope
signals
that
incorporates
Butterworth
low-pass
filtering
fast
Hilbert
transform
(FHT),
which
can
successfully
reduce
noise
interference
capture
essential
physiological
information.
Finally,
majority
voting
window
technique
adopted
enhance
prediction
results,
further
improving
accuracy
stability
model.
Overall,
our
multi-layered
network
model,
conjunction
signal
extraction
mechanisms,
offers
promising
innovative
control
hands,
allowing
precise
fine
motor
actions.
IEEE Transactions on Intelligent Transportation Systems,
Год журнала:
2023,
Номер
25(3), С. 3118 - 3127
Опубликована: Сен. 21, 2023
Autonomous
driving
has
gradually
become
a
research
hotspot
in
recent
years,
but
the
robustness
of
loopback
detection
complex
environments
such
as
dynamic
and
weak
textures
needs
to
be
improved.
A
semantic
method
is
proposed
based
on
instance
segmentation
visual
SLAM
make
sufficient
use
information
autonomous
driving.
The
combines
image
(Simultaneous
Localization
Mapping)
construct
system.
What's
more,
data
association
that
geometric
improve
traditional
by
using
increase
accuracy
detection.
result
experiment
TUM
public
dataset
shows
improved
higher
than
bag-of-words
all
four
datasets,
our
algorithm
can
effectively
system
general.
Concurrency and Computation Practice and Experience,
Год журнала:
2024,
Номер
36(18)
Опубликована: Май 28, 2024
Summary
Passenger
flow
prediction
is
an
important
part
of
daily
metro
operation,
and
its
accuracy
affects
the
deployment
train
resources
management.
Due
to
complex
spatiotemporal
correlation
characteristics
passenger
flow,
it
necessary
describe
improve
prediction.
However,
existing
models
mainly
construct
weight
matrix
based
on
static
graph
similarity
between
stations
when
describing
spatial
station
but
ignore
time‐varying
flow.
To
address
this
problem,
study
introduces
a
dynamic
multi‐graph
multidimensional
attention
model.
Specifically,
Graph
Convolutional
Neural
Network
combined
with
multigraph
extracts
features
Gated
Recurrent
Unit
temporal
The
can
obtain
data
by
assigning
weights
them.
Finally,
model
has
been
used
conduct
experiments
Beijing
datasets
time
granularity
10
15
min.
result
indicates
that
DGMANN
outperforms
state‐of‐the‐art
other
deep
learning
methods
in
In
addition,
effectiveness
key
submodules
verified
through
ablation
experiments.
