Concilium,
Год журнала:
2024,
Номер
24(3), С. 229 - 248
Опубликована: Фев. 22, 2024
This
research
aimed
to
analyze
the
importance
of
artificial
intelligence
and
sustainability
in
higher
education
according
literature
field
present
relationships
this
context
with
United
Nations
Sustainable
Development
Goals
(SDGs).
The
adopted
strategies
included
bibliometric
analysis
using
VOSviewer
software
review,
considering
Web
Science
scientific
database.
resulted
clustering
four
groups.
blue
cluster
highlighted
emergence
interest
studies
on
AI
following
Covid-19
pandemic.
green
emphasized
more
efficient
teaching
methods
adapted
demands
education,
as
well
need
empower
teachers
use
developing
students'
skills
competencies,
emphasizing
sustainability.
yellow
indicated
presence
based
triad
sustainable
innovation,
aiming
prepare
students
for
future
challenges.
red
impact
focusing
student
learning,
efficiency,
performance.
Finally,
identified
main
technologies
their
relationship
SDGs.
reflections
presented
here
can
contribute
expanding
discussions
between
education.
From
a
practical
standpoint,
it
serve
foundation
university
managers
promote
integration
into
processes,
Forests,
Год журнала:
2024,
Номер
15(8), С. 1374 - 1374
Опубликована: Авг. 6, 2024
Trees’
structural
defects
are
responsible
for
the
reduction
in
forest
product
quality
and
accident
of
tree
collapse
under
extreme
environmental
conditions.
Although
manual
view
inspection
assessing
health
condition
is
reliable,
it
inefficient
discriminating,
locating,
quantifying
with
various
features
(i.e.,
crack
hole).
There
a
general
need
investigation
efficient
ways
to
assess
these
enhance
sustainability
trees.
In
this
study,
deep
learning
algorithms
lightweight
You
Only
Look
Once
(YOLO)
encoder-decoder
network
named
DeepLabv3+
combined
unmanned
aerial
vehicle
(UAV)
observations
evaluate
trees’
defects.
Experimentally,
we
found
that
state-of-the-art
detector
YOLOv7-tiny
offers
real-time
50–60
fps)
long-range
sensing
5
m)
but
has
limited
capacity
acquire
patterns
at
millimeter
scale.
To
address
limitation,
further
utilized
cascaded
different
architectures
ResNet18,
ResNet50,
Xception,
MobileNetv2
obtain
actual
morphology
through
close-range
pixel-wise
image
semantic
segmentation.
Moreover,
proposed
hybrid
scheme
YOLOv7-tiny_DeepLabv3+_UAV
assesses
tree’s
defect
size
an
averaged
accuracy
92.62%
(±6%).
Concilium,
Год журнала:
2024,
Номер
24(3), С. 229 - 248
Опубликована: Фев. 22, 2024
This
research
aimed
to
analyze
the
importance
of
artificial
intelligence
and
sustainability
in
higher
education
according
literature
field
present
relationships
this
context
with
United
Nations
Sustainable
Development
Goals
(SDGs).
The
adopted
strategies
included
bibliometric
analysis
using
VOSviewer
software
review,
considering
Web
Science
scientific
database.
resulted
clustering
four
groups.
blue
cluster
highlighted
emergence
interest
studies
on
AI
following
Covid-19
pandemic.
green
emphasized
more
efficient
teaching
methods
adapted
demands
education,
as
well
need
empower
teachers
use
developing
students'
skills
competencies,
emphasizing
sustainability.
yellow
indicated
presence
based
triad
sustainable
innovation,
aiming
prepare
students
for
future
challenges.
red
impact
focusing
student
learning,
efficiency,
performance.
Finally,
identified
main
technologies
their
relationship
SDGs.
reflections
presented
here
can
contribute
expanding
discussions
between
education.
From
a
practical
standpoint,
it
serve
foundation
university
managers
promote
integration
into
processes,