International Journal of Circuit Theory and Applications,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Nov. 15, 2024
ABSTRACT
The
generated
mechanism
of
negative
emotion
involves
the
prefrontal
cortex,
hippocampus,
and
amygdala
in
brain.
Based
on
biological
mechanism,
this
paper
proposes
a
memristive
neural
network
circuit
inhibition
with
self‐repair
memory.
proposed
consists
five
modules:
memory
(hippocampus)
module,
(prefrontal
cortex)
damage
detection
repair
output
(amygdala)
module.
module
does
not
respond
to
small
signal,
but
large
signal
will
cause
form
memories.
If
repeats,
receives
then
generates
an
signal.
When
is
small,
outputs
normally.
applied
for
first
time,
becomes
damaged
abnormal.
As
memristor
has
exceeded
its
threshold,
generate
restore
memristance
after
receiving
again,
from
help
appropriately.
PSPICE
simulation
results
show
that
hippocampus
memories
learning
transmits
them
cortex
inhibits
can
be
robots,
which
flexibly
alter
intensity
expressed
by
robots.
Micromachines,
Journal Year:
2024,
Volume and Issue:
15(2), P. 253 - 253
Published: Feb. 8, 2024
Memristor
devices
have
diverse
physical
models
depending
on
their
structure.
In
addition,
the
properties
of
memristors
are
described
using
complex
differential
equations.
Therefore,
it
is
necessary
to
integrate
various
memristor
into
an
unified
physics-based
model.
this
paper,
we
propose
a
physics-informed
neural
network
(PINN)-based
compact
PINNs
can
solve
equations
intuitively
and
with
ease.
This
methodology
used
conduct
analysis.
The
weight
bias
extracted
from
PINN
implemented
in
Verilog-A
circuit
simulator
predict
device
characteristics.
accuracy
proposed
model
verified
two
devices.
results
show
that
be
extensively
models.
Sensors,
Journal Year:
2025,
Volume and Issue:
25(5), P. 1533 - 1533
Published: March 1, 2025
Neuromorphic
devices
emulating
the
temperature-sensing
capabilities
of
biological
thermoreceptors
hold
significant
promise
for
neuron-like
artificial
sensory
systems.
In
this
work,
Bi2Se3-based
threshold-switching
memristors
were
presented
in
constructing
neuron
circuits,
leveraging
its
exceptional
attributes,
such
as
high
switching
ratio
(>106),
low
threshold
voltage,
and
thermoelectric
response.
The
spiking
oscillation
response
to
resistance
temperature
variations
was
analyzed
using
Hspice
simulation
memristor
model
based
on
on/off
states,
voltage
(Vth),
(Vhold).
These
results
show
great
potential
enabling
biorealistic
thermoreception
applications.
The European Physical Journal Special Topics,
Journal Year:
2025,
Volume and Issue:
unknown
Published: May 23, 2025
Abstract
This
study
investigates
synchronization
dynamics
in
functional
networks
derived
from
magnetic
resonance
imaging
data
for
control
and
body
dysmorphic
disorder
groups,
emphasizing
the
importance
of
understanding
altered
brain
psychiatric
conditions.
First,
we
analyze
routes
to
synchronization,
identifying
explosive
continuous
transitions,
with
exhibiting
a
predominantly
route,
which
may
underlie
their
pathological
symptoms.
Second,
explore
chimera
states
intermediate
coupling
regime,
capturing
coexistence
synchronized
desynchronized
regions
as
network
transitions
heterogeneous
activity
complete
synchronization.
Finally,
evaluate
effect
sinusoidal
square
pulse
stimulation
on
levels
network,
targeting
specific
regions.
These
results
explain
reformed
tuning
neuromodulation
strategies.
IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems,
Journal Year:
2024,
Volume and Issue:
unknown, P. 1 - 1
Published: Jan. 1, 2024
Most
existing
visual
object
tracking
approaches
are
implemented
based
on
von
Neumann
computation
systems,
which
inevitably
have
the
problems
of
high
latency.Additionally,
remote
server
processing
video
resources
requires
a
large
amount
data
transmission
over
Internet,
limits
real-time
performance.The
integration
technology
into
electronic
devices
has
become
new
trend.However,
current
algorithm
complexity,
making
it
difficult
to
design
circuits
implement
corresponding
functions.In
this
paper,
memristorbased
attention
network
and
its
proposed
achieve
online
under
parallel
computing.Memristors
used
construct
encoding
record
changes
target
in
historical
frames,
adjust
signals
realtime
during
process,
avoiding
latency
problem
architecture.Inspired
by
working
process
γ-GABAergic
interneuron
tripartite
synapse,
we
propose
an
allocation
module
selectively
allocate
values.Combining
Winner-Take-All
principle,
localization
circuit
optimal
zone
selection
for
track
location
target.Finally,
experiments
analyses
OTB-100,
NFS,
VOT-RTb2022
benchmark
datasets
verify
that
memristor-based
promising
performance
achieves
speed
1000
FPS,
demonstrating
superior
performance.
