Science and Technology of Advanced Materials,
Journal Year:
2022,
Volume and Issue:
24(1)
Published: Nov. 25, 2022
Conventional
von
Neumann
architecture
is
insufficient
in
establishing
artificial
intelligence
(AI)
terms
of
energy
efficiency,
computing
memory
and
dynamic
learning.
Delightedly,
rapid
developments
neuromorphic
provide
a
new
paradigm
to
solve
this
dilemma.
Furthermore,
devices
that
can
realize
synaptic
plasticity
function
have
extraordinary
significance
for
system.
A
three-terminal
transistor
one
the
typical
representatives.
In
addition,
human
body
has
five
senses,
including
vision,
touch,
auditory
sense,
olfactory
sense
gustatory
providing
abundant
information
brain.
Inspired
by
perception
system,
system
will
give
vitality
intelligent
robots.
This
review
discusses
operation
mechanism,
application
transistors.
The
latest
progresses
systems
based
on
transistors
are
provided.
Finally,
opportunities
challenges
summarized.
Advanced Functional Materials,
Journal Year:
2022,
Volume and Issue:
32(22)
Published: Feb. 26, 2022
Abstract
Neuromorphic
circuits
emulating
the
bio‐brain
functionality
via
artificial
devices
have
achieved
a
substantial
scientific
leap
in
past
decade.
However,
even
with
advent
of
highly
advanced
bio‐inspired
algorithms,
intelligence
based
on
current
neuromorphic
is
lagging
behind
significantly
when
compared
naturally
evolved
biological
neural
circuits.
This
massive
and
intriguing
discrepancy
partly
due
to
incomprehensive
understanding
operating
mechanism,
which
relies
heavily
extremely
complexed
entangled
3D
hierarchical
networks.
Configuring
hardware
combined
computing
memory
functionalities,
coupled
compatible
progress
software
can
be
an
inevitable
route
surmount
limitation
encountered
by
2D
Herein,
referring
neuron
configuration
perspective
together
detailed
signal
generation
propagation
von
Neumann
state‐of‐the‐art
in‐memory
architecture,
development
perspectives
are
highlighted.
Advanced Functional Materials,
Journal Year:
2022,
Volume and Issue:
32(21)
Published: Feb. 15, 2022
Abstract
The
simulation
of
human
feelings,
perceptions,
and
actions
has
become
an
important
application
field
for
medical
treatment,
human–computer
interfaces,
intelligent
robots.
However,
the
need
various
functional
units
hinders
system
integration.
A
threshold‐switching
memristor
is
used
as
sympathetic
nerve
center
to
simulate
unconditioned
reflex.
High
current
nonlinearity
(10
6
)
achieved
by
adopting
a
nanocontact
structure,
high
flexibility
(bending
radius
≈5
mm)
obtained
combining
structure
with
flexible
polymer
film.
Furthermore,
integrating
sensor
artificial
muscle
made
ionic
polymer–metal
composite
allows
form
all‐flexible
complete
reflex
arc
can
provide
strain‐dependent
control
through
simple
logic,
indicating
great
potential
memristors
primary
neuromorphological
treatment.
IEEE Transactions on Circuits and Systems for Video Technology,
Journal Year:
2023,
Volume and Issue:
33(12), P. 7928 - 7942
Published: May 12, 2023
Video
sentiment
analysis
can
effectively
establish
the
relationship
between
emotion
state
and
multimodal
information,
while
still
suffer
from
intensive
computation
low
efficiency,
due
to
von
Neumann
computing
architecture.
Here,
we
present
a
brain-inspired
hierarchical
interactive
in-memory
(IMC)
system,
which
efficiently
solve
'von
bottleneck',
enabling
cross-modal
interactions
semantic
gap
elimination.
First,
1T1M
synapse
array
is
fabricated
using
cost-effective,
highly
stable,
flexible,
eco-friendly
carbon
materials,
offering
efficient
analog
multiply-accumulate
operations.
To
illustrate
complexity
of
proposed
IMC
three
modules
are
proposed:
1)
unimodal
extraction
module,
2)
3)
output
module.
Furthermore,
system
validated
by
applying
it
video
analysis.
The
experimental
results
demonstrate
that
outperforms
existing
state-of-the-art
methods
with
high
computational
efficiency
good
robustness.
This
work
opens
up
new
way
achieve
deep
integration
nanomaterials,
learning,
modern
electronics
into
IMC.
Advanced Functional Materials,
Journal Year:
2023,
Volume and Issue:
33(10)
Published: Jan. 15, 2023
Abstract
Over
the
past
several
years,
a
variety
of
hardware‐based
artificial
sensory
systems,
including
skin,
electronic
noses,
and
retinas,
have
attracted
considerable
research
interest
in
advanced
intelligence
systems.
The
integration
sensing
computing
functions
single
or
multiple
connected
self‐adaptive
field‐effect
transistor
(FET)‐structured
devices
to
implement
olfactory
systems
for
in‐sensor
has
recently
increasing
attention.
In
this
review,
development
status
FET‐based
gas
is
focused
on.
mechanisms
FET
devices,
gas‐recognition
materials,
strategies
improving
performance,
into
system
are
discussed.
Finally,
further
their
great
potential
next‐generation
intelligent
discussed
broad
fields
such
as
environmental
monitoring,
health
care,
military
industries.
Advanced Materials,
Journal Year:
2023,
Volume and Issue:
35(35)
Published: April 29, 2023
Technologies
based
on
the
fusion
of
gas
sensors
and
neuromorphic
computing
to
mimic
olfactory
system
have
immense
potential.
However,
implementation
systems
remains
in
a
state
infancy
because
conventional
lack
necessary
functions.
Therefore,
this
study
proposes
hysteretic
"chemi-memristive
sensor"
oxygen
vacancy
chemi-memristive
dynamics
that
differ
from
sensors.
After
memristive
switching
operation,
redox
reaction
with
external
molecules
is
enhanced,
resulting
generation
elimination
vacancies
induce
rapid
current
changes.
In
addition,
pre-generated
enhance
post-sensing
properties.
fast
responses,
short
recovery
times,
response
are
achieved
by
proposed
sensor
at
room
temperature.
Based
advantageous
functionality
sensor,
device-level
can
monitor
history
input
stimuli
experimentally
demonstrated
as
potential
application.
Moreover,
analog
conductance
modulation
induced
oxidizing
reducing
gases
enables
conversion
into
synaptic
weights
hence
realization
typical
functionalities
without
an
additional
device
or
circuit.
The
represents
advance
bioinspired
technology
adopted
creating
artificial
intelligence
systems.
Advanced Functional Materials,
Journal Year:
2024,
Volume and Issue:
34(36)
Published: April 22, 2024
Abstract
The
superior
recognition
ability
and
excitatory–inhibitory
balance
of
the
olfactory
system
has
important
applications
in
efficient
recognition,
analysis,
processing
data.
In
this
study,
transistor
synaptic
devices
are
prepared
utilizing
poly‐diketo‐pyrrolopyrrole‐selenophene
polymer
(PTDPPSe‐5Si)
with
excellent
electrical
properties
as
active
layer,
dual‐gas
pulses
applied
for
first
time
to
simulate
excitatory
inhibitory
behaviors
system.
Basic
successfully
simulated,
such
excitatory/inhibitory
postsynaptic
currents
(EPSC/IPSC),
long‐term
potentiation/depression
(LTP/LTD).
regulation
biomimetic
olfaction
is
simulated.
This
working
mechanism
attributed
capture
release
carriers
channel
induced
by
gas's
electron‐donating
electron‐withdrawing
characteristics.
neuromotor
pathway
constructed
using
key
unit,
which
enables
integration
information
from
neurons
output
motor
neurons.
A
convolutional
neural
network
achieve
eight
common
laboratory
gas
types
concentrations
a
accuracy
over
97%.
simulated
exhibited
device
hold
significant
importance
development
artificial
networks,
intelligent
frameworks,
robots.
Advanced Materials,
Journal Year:
2024,
Volume and Issue:
36(28)
Published: May 15, 2024
The
rapid
and
responsive
capabilities
of
soft
robots
in
perceiving,
assessing,
reacting
to
environmental
stimuli
are
highly
valuable.
However,
many
existing
robots,
designed
mimic
humans
other
higher
animals,
often
rely
on
data
centers
for
the
modulation
mechanoelectrical
transduction
electromechanical
actuation.
This
reliance
significantly
increases
system
complexity
time
delays.
Herein,
drawing
inspiration
from
Venus
flytraps,
a
robot
employing
power
strategy
is
presented
active
stimulus
reaction,
eliminating
need
center.
achieves
through
Ni
Advanced Materials,
Journal Year:
2024,
Volume and Issue:
36(29)
Published: May 1, 2024
Neuromorphic
olfactory
systems
have
been
actively
studied
in
recent
years
owing
to
their
considerable
potential
electronic
noses,
robotics,
and
neuromorphic
data
processing
systems.
However,
conventional
gas
sensors
typically
the
ability
detect
hazardous
levels
but
lack
synaptic
functions
such
as
memory
recognition
of
accumulation,
which
are
essential
for
realizing
human-like
sensory
system.
In
this
study,
a
seamless
architecture
system
capable
detecting
memorizing
present
level
accumulation
status
nitrogen
dioxide
(NO
