bioRxiv (Cold Spring Harbor Laboratory),
Год журнала:
2024,
Номер
unknown
Опубликована: Май 8, 2024
Abstract
Recent
theories
suggest
a
new
brain
pathway
dedicated
to
processing
social
movement
is
involved
in
understanding
emotions
from
biological
motion,
beyond
the
well-known
ventral
and
dorsal
pathways.
However,
how
this
functions
as
network
that
computes
dynamic
motion
signals
for
perceptual
behavior
unchartered.
Here,
we
used
generative
model
of
important
facial
movements
participants
(N
=
10)
categorized
“happy,”
“surprise,”
“fear,”
“anger,”
“disgust,”
“sad”
while
recorded
their
MEG
responses.
Using
representational
interaction
measures
(between
features,
t
source,
behavioral
responses),
reveal
per
participant
functional
extending
occipital
cortex
superior
temporal
gyrus.
Its
sources
selectively
represent,
communicate
compose
disambiguate
emotion
categorization
behavior,
swiftly
filters
out
task-irrelevant
identity-defining
face
shape
features.
Our
findings
complex
categorize
individual
participants.
Forests,
Год журнала:
2023,
Номер
14(3), С. 483 - 483
Опубликована: Фев. 27, 2023
Due
to
the
complexity
and
difficulty
of
forest
resource
ground
surveys,
remote-sensing-based
methods
assess
resources
effectively
plan
management
measures
are
particularly
important,
as
they
provide
effective
means
explore
changes
in
over
long
time
periods.
The
objective
this
study
was
monitor
spatiotemporal
trends
wood
carbon
stocks
standing
forests
southeastern
Xiaoxinganling
Mountains
by
using
Landsat
remote
sensing
data
collected
between
1989
2021.
Various
indicators
for
predicting
were
constructed
based
on
Google
Earth
Engine
(GEE)
platform.
We
initially
used
a
multiple
linear
regression
model,
deep
neural
network
model
convolutional
exploring
stocks.
Finally,
we
chose
because
it
provided
more
robust
predictions
stock
pixel-by-pixel
basis
hence
mapping
spatial
distribution
variable.
Savitzky–Golay
filter
smoothing
applied
predicted
annual
average
observe
overall
trend,
autocorrelation
analysis
conducted.
Sen’s
slope
Mann–Kendall
statistical
test
It
found
that
59.5%
area
showed
an
increasing
while
40.5%
decreasing
trend
past
33
years,
future
development
plotted
combining
results
with
Hurst
exponent.
Current Biology,
Год журнала:
2024,
Номер
34(15), С. 3392 - 3404.e5
Опубликована: Июль 18, 2024
Highlights•Occipital
cortex
represents
both
task-relevant
and
irrelevant
features
before
120
ms•Only
advance
to
the
temporal
region•During
121–150
ms,
occipital
representations
reduce
lower-dimensional
manifolds•These
manifolds
then
transform
into
from
161
350
msSummaryTo
interpret
our
surroundings,
brain
uses
a
visual
categorization
process.
Current
theories
models
suggest
that
this
process
comprises
hierarchy
of
different
computations
transforms
complex,
high-dimensional
inputs
(i.e.,
manifolds)
in
support
multiple
behaviors.
Here,
we
tested
hypothesis
by
analyzing
these
transformations
reflected
dynamic
MEG
source
activity
while
individual
participants
actively
categorized
same
stimuli
according
tasks:
face
expression,
gender,
pedestrian
vehicle
type.
Results
reveal
three
transformation
stages
guided
pre-frontal
cortex.
At
stage
1
(high-dimensional,
50–120
ms),
sources
represent
task-irrelevant
stimulus
features;
higher
ventral/dorsal
regions,
whereas
halt
at
occipital-temporal
junction.
2
(121–150
feature
manifolds,
which
underlying
behavior
over
3
(161–350
ms).
Our
findings
shed
light
on
how
brain's
network
mechanisms
specific
behaviors.Graphical
abstract
Nature Communications,
Год журнала:
2024,
Номер
15(1)
Опубликована: Май 17, 2024
Revealing
how
the
mind
represents
information
is
a
longstanding
goal
of
cognitive
science.
However,
there
currently
no
framework
for
reconstructing
broad
range
mental
representations
that
humans
possess.
Here,
we
ask
participants
to
indicate
what
they
perceive
in
images
made
random
visual
features
deep
neural
network.
We
then
infer
associations
between
semantic
their
responses
and
images.
This
allows
us
reconstruct
multiple
concepts,
both
those
supplied
by
other
concepts
extrapolated
from
same
space.
validate
these
reconstructions
separate
further
generalize
our
approach
predict
behavior
new
stimuli
task.
Finally,
individual
observers
enables
large-scale
investigation
conceptual
representations.
