arXiv (Cornell University),
Год журнала:
2023,
Номер
unknown
Опубликована: Янв. 1, 2023
The
Attention
Restoration
Theory
(ART)
presents
a
theoretical
framework
with
four
essential
indicators
(being
away,
extent,
fascinating,
and
compatibility)
for
comprehending
urban
natural
restoration
quality.
However,
previous
studies
relied
on
non-sequential
data
non-spatial
dependent
methods,
which
overlooks
the
impact
of
spatial
structure
defined
here
as
positional
relationships
between
scene
entities
past
methods
also
make
it
challenging
to
measure
quality
an
scale.
In
this
work,
spatial-dependent
graph
neural
networks
(GNNs)
approach
is
proposed
reveal
relation
Specifically,
we
constructed
two
different
types
graphs
at
street
city
levels.
street-level
graphs,
using
sequential
view
images
(SVIs)
road
segments
capture
position
entities,
were
used
represent
structure.
city-level
graph,
modeling
topological
roads
non-Euclidean
structures
embedding
features
(including
Perception-features,
Spatial-features,
Socioeconomic-features),
was
results
demonstrate
that:
1)
GNNs
model
outperforms
traditional
(Acc
=
0.735,
F1
0.732);
2)
portrayed
through
SVIs
significantly
influences
quality;
3)
spaces
same
exhibited
distinct
patterns.
This
study
clarifies
association
quality,
providing
new
perspective
improve
well-being
in
future.
Land,
Год журнала:
2025,
Номер
14(3), С. 610 - 610
Опубликована: Март 13, 2025
The
mental
health
of
university
students
has
received
much
attention
due
to
the
various
pressures
studies,
life,
and
employment.
Several
studies
have
confirmed
that
campus
public
spaces
contain
multiple
restorative
potentials.
Yet,
space
is
still
not
ready
meet
students’
new
need
for
percetions.
Renewal
practices
integrate
multi-issues
are
becoming
more
important,
further
clarification
measurement
methods
optimization
pathways
also
needed.
This
study
applied
semantic
segmentation
technique
deep
learning
model
extract
feature
indicators
outdoor
based
on
street
view
image
(SVI)
data.
subjective
evaluation
small-scale
SVIs
was
obtained
using
perceived
scale-11
(PRS-11)
questionnaire.
On
this
basis,
benefit
models
were
established,
including
explanatory
predictive
models.
used
Pearson’s
correlation
linear
regression
analysis
identify
key
affecting
benefits,
XGBoost
1.7.3
algorithm
predict
scores
scale.
accessibility
results
from
sDNA
then
overlayed
form
a
comprehensive
assessment
matrix
restoration
benefits
dimensions
“areas
with
potential”.
In
way,
three
types
spatial
(LRB-HA,
HRB-LA,
LRB-LA)
sequential
orders
temporal
(short-term,
medium-term,
long-term)
combined
propose
dual
control
accessibility.
provides
methodological
guidelines
empirical
data
regeneration
promotes
efficiency.
addition,
it
can
offer
positive
references
neighborhood-scale
urban
design
sustainable
development.
Engineering Construction & Architectural Management,
Год журнала:
2025,
Номер
unknown
Опубликована: Фев. 6, 2025
Purpose
Understanding
the
restorative
potential
of
built
environments
is
essential
for
promoting
mental
well-being.
However,
existing
studies
often
rely
on
static
image-based
methods,
which
are
inherently
limited
in
capturing
temporal
and
spatial
dynamics
environmental
perception.
These
methods
frequently
introduce
biases,
such
as
selective
framing
abrupt
transitions,
failing
to
reflect
natural
viewing
behavior.
Addressing
these
limitations,
this
study
investigates
qualities
campus
using
dynamic
VR
stimuli
eye-tracking
technology.
By
providing
continuous
information,
offer
a
more
immersive
ecologically
valid
approach
understanding
how
specific
features
contribute
psychological
restoration.
Design/methodology/approach
This
technology
stimuli.
Campus
were
filmed
through
walking
sequences
paired
with
PRS
audio
prompts.
About
40
university
students
participated
experiment,
data
processed
computer
vision-based
semantic
segmentation
concept
relative
areas
interest,
followed
by
correlation
analysis
quality
scores.
Findings
The
results
revealed
that
elements
“sky,”
“tree,”
“waterscape”
“landscape
corridor”
significantly
positively
correlated
being-away
fascination
dimensions,
indicating
their
role
attention
supporting
recovery.
Conversely,
architectural
like
“architectural
“building
facade”
negatively
extent
dimension,
while
open
space”
correlated,
enhancing
perception
exploration.
Originality/value
findings
underscore
importance
spaces
also
revealing
complex
influence
features.
provides
valuable
insights
optimizing
design
support
students’
health
