Machine
Learning
(ML)
techniques,
specifically
Support
Vector
(SVM)
and
Extreme
Gradient
Boosting
(XGBoost),
were
employed
to
achieve
precise
intuitive
real-time
eye
tracking
mouse
control
through
computer
vision.
However,
XGBoost
may
suffer
from
overfitting
when
dealing
with
a
large
number
of
features
compared
the
training
data
size,
or
noisy
imbalanced
data.
To
address
this
issue,
paper
introduces
Recurrent
Neural
Network
(RNN),
for
in
Human
Computer
Interaction
(HCI).
Gaze
Direction
Estimation
(GDE)
is
initially
estimate
gaze
direction,
utilizing
pupil
positions
camera
calibration
parameters.
The
estimated
direction
then
used
as
input
RNN
eye-tracking
HCI.
experimental
results
shows
that
GDE-RNN
has
26.03%
8.24%
superior
accuracy,
24.93%
86.76%
better
precision,
27.36%
8.75%
high
recall
comparison
SVM
control.
Applied Sciences,
Год журнала:
2024,
Номер
14(2), С. 640 - 640
Опубликована: Янв. 11, 2024
In
the
area
of
aviation
safety,
importance
human
factors
is
indisputable.
This
research
endeavors
to
assess
in
predicting
fatalities
during
mishaps.
Utilizing
reports
from
Aviation
Safety
Network
Database,
encompassing
1105
accidents
and
incidents
spanning
2007
2016,
neural
networks
were
trained
forecast
probability
fatalities.
Our
findings
underscore
that
involved,
by
themselves,
can
yield
strong
predictions.
As
a
term
comparison,
other
variables
(type
occurrence,
flight
phase,
aircraft
fate)
used
as
predictors,
with
poorer
results;
combining
these
factors,
prediction
only
marginally
better,
if
at
all,
than
based
on
alone.
So,
although
supplementary
benefit
predictive
results
derived
their
contribution
remains
minimal.
Consequently,
this
study
illuminates
paramount
influencing
fatalities,
guiding
stakeholders
immediate
interventions
investments
which
are
most
warranted
prevent
them.
Chinese Journal of Aeronautics,
Год журнала:
2024,
Номер
37(8), С. 277 - 292
Опубликована: Апрель 12, 2024
With
cutting-edge
technologies
and
considering
airline
human-resource-saving,
a
single
pilot
in
commercial
jets
could
be
technically
feasible.
Investigating
changes
captains'
natural
behaviours
are
initially
required
to
comprehend
the
specific
safe
human
performance
envelope
for
safeguarding
single-pilot
flight,
particularly
high-risk
situations.
This
paper
investigates
how
transforms
fixing
emergencies
when
operating
from
Dual-Pilot
Operations
(DPO)
Single-Pilot
(SPO)
through
physiological-based
approach.
Twenty
pilots
flew
an
emergency-included
flight
with/without
first
officers'
assistance.
The
neural
activities
scanning
were
recorded
using
32-channel
Electroencephalogram
(EEG)
glasses-based
eye
tracker,
with
observation
post-experiment
questionnaires
evaluate
operations
pilots'
perception.
Flying
alone,
there
was
significantly
increased
cortical
activity
θ
β
waves
over
frontal,
parietal,
temporal
lobes
during
more
complicated
emergencies,
focused
less
on
primary
display
while
spending
time
other
interfaces.
physiological
fluctuating
patterns
associated
risky
SPO
highlighted
by
cross-correlating
multimodal
data.
experimental-based
noteworthy
insights
may
wish
inform
measures
lessen
persistent
fluctuation,
assisting
airlines
creating
SPO-oriented
intelligent
systems
give
captains
adequate
support
assuring
safer
air
transportation.
