Measurement Science and Technology,
Год журнала:
2024,
Номер
36(1), С. 015411 - 015411
Опубликована: Окт. 14, 2024
Abstract
Facial
expression
generation
technology
has
achieved
notable
progress
in
computer
vision
and
artificial
intelligence.
However,
challenges
persist
regarding
background
consistency,
clarity,
detailed
representation.
Additionally,
the
instability
of
generative
adversarial
networks
(GANs)
during
training
affects
both
image
quality
diversity.
While
diffusion
models
have
demonstrated
potential
advantages
over
GANs,
research
on
controllable
remains
limited.
To
address
these
challenges,
this
paper
proposes
a
highly
natural
facial
method
based
denoising
implicit
(DDIM)
with
embedded
vein
features.
This
approach
avoids
by
employing
gradual
to
generate
specific
expressions,
thereby
enhancing
diversity
authenticity
images.
Vein
features
are
introduced
within
generated
images
protect
intellectual
property
(IP)
algorithm-generated
digital
resources.
Firstly,
text
guide
words
combined
as
conditional
inputs
improve
Secondly,
classification
coding
network
is
generation,
thus
accuracy
consistency
produced
expressions.
Furthermore,
feature
fusion
multi-directional
local
dynamic
operator
integrates
DDIM
frequency-domain
watermarking
achieve
IP
protection.
Experimental
results
demonstrate
effectiveness
across
several
public
datasets,
including
FFHQ,
CelebA,
FV-USM,
SDUMLA-HMT.
Notably,
CelebA
dataset,
average
recognition
rate
increased
11.41%,
100.00%
for
happy
The
exhibit
high
degree
video
conversion
tests
reveal
smooth
effect.
These
confirm
that
not
only
advances
but
also
significantly
enhances
steganographic
protection
2022 IEEE 7th International conference for Convergence in Technology (I2CT),
Год журнала:
2023,
Номер
unknown, С. 1 - 4
Опубликована: Апрель 7, 2023
In
contrast
to
traditional
computing,
soft
computing
can
be
used
adjust
prediction
models
and
locate
answers
difficult,
real-world
issues.
Soft
is
more
forgiving
of
ambiguity
than
classical
computing.
For
the
gold
standard
human
brain.
includes
innovations
like
machine
learning,
artificial
neural
networks,
genetic
algorithms,
fuzzy
logic,
expert
systems.
First
proposed
in
1980s,
now
a
central
focus
academic
inquiry
across
disciplines.
widely
commercial,
industrial,
household
settings
due
its
low
cost
high
performance.
finding
uses
today.
This
article
compares
contrasts
benefits
with
those
contemporary
The
IT
sector
is
fast
embracing
cloud
computing,
a
nextgeneration
technology.
Even
though
this
technology
offers
many
benefits,
there
are
also
some
downsides.
most
challenging
aspect
of
security.
We
described
three-tier
architecture
for
private
cloud-based
encrypted
communication
where
developed
brand-new
custom
approach
and
combined
steganography
to
provide
an
additional
layer
protection
data
transmission
across
the
cloud.
suggested
technique
effectively
serves
as
road
map
building
Cloud
security
architecture.
Additionally,
key
may
be
distributed
users
using
efficient
management
strategies,
ensuring
that
only
those
who
authorized
have
access
data.
The
term
"robot"
refers
to
an
electromechanical
device
that,
as
a
result
of
its
incorporation
electronic
and
computer
programming,
may
carry
out
tasks
either
independently
or
in
conjunction
with
human
operator
[1].
Robots
can
be
designed
perform
functions
any
order
that
the
programmer
specifies.
have
found
uses
broad
variety
disciplines,
including
those
connected
military,
healthcare,
industry,
among
number
other
fields.
programmed
mobile
stationary
manner,
choice
which
mode
use
is
often
dictated
by
are
intended
carried
robots.
It
extremely
essential
for
robot
able
traverse
environment
capable
efficiently
completing
tasks,
avoiding
obstacles,
participating
activities.
This
capability
navigation,
dependent
on
sensors
supply
environmental
data
feedback
signals,
operator-independent
autonomous
if
"intelligence"
built
into
code.
Sensors
required
provide
signals.
return
learning
opportunity
further
used
affordable
energy,
agriculture,
environmentally
sound
technologies,
etc.
Semnas Ristek (Seminar Nasional Riset dan Inovasi Teknologi),
Год журнала:
2024,
Номер
8(01)
Опубликована: Янв. 31, 2024
Studi
ini
mengevaluasi
efektivitas
metode
watermarking
dalam
menyembunyikan
informasi
rahasia
pada
citra
digital
menggunakan
Discrete
Cosine
Transform
(DCT)
dan
Singular
Value
Decomposition
(SVD).
Pendekatan
penting
untuk
menjaga
keamanan
hak
cipta
era
digital.
Penggunaan
DCT
memungkinkan
penyematan
watermark
tanpa
mengorbankan
kualitas
visual
citra.
Hasil
evaluasi
Mean
Squared
Error
(MSE)
menunjukkan
bahwa
Lena.bmp
mencapai
nilai
MSE
terendah
Level
1
dengan
0.075,
sementara
Peppers.png
memiliki
0.0083,
Baboon.jpg
0.0097.
Pada
sisi
lain,
hasil
Peak
Signal-to-Noise
Ratio
(PSNR)
PSNR
tertinggi
tercatat
ketiga
48.17
dB.
Temuan
yang
diterapkan
SVD
berhasil
menyematkan
tingkat
preservasi
tinggi.
Bulletin of Electrical Engineering and Informatics,
Год журнала:
2024,
Номер
13(4), С. 2537 - 2546
Опубликована: Июнь 1, 2024
This
study
introduces
an
advanced
method
for
evaluating
non-blind
watermarking
quality,
leveraging
both
visible
and
invisible
techniques
grounded
in
principles
of
discrete
cosine
transform
(DCT)
modified
singular
value
decomposition
(M-SVD).
The
primary
focus
is
to
refine
the
assessment
process
watermarked
images
by
integrating
M-SVD,
known
its
efficacy
measuring
image
quality
performance.
Results
from
M-SVD
implementation
exhibit
a
striking
resemblance
original
images.
mean
squared
error
(MSE)
values
range
0.0003
0.0168,
while
peak
signal-to-noise
ratio
(PSNR)
vary
between
42.52
dB
82.72
dB.
These
outcomes
underscore
potential
DCT
bolstering
processes,
especially
contexts.
Matrik Jurnal Manajemen Teknik Informatika dan Rekayasa Komputer,
Год журнала:
2024,
Номер
23(3), С. 509 - 518
Опубликована: Июнь 14, 2024
Image
watermarking
is
a
sophisticated
method
often
used
to
assert
ownership
and
ensure
the
integrity
of
digital
images.
This
research
aimed
propose
evaluate
an
advanced
technique
that
utilizes
combination
singular
value
decomposition
methodology
discrete
cosine
transformation
embed
Dian
Nuswantoro
University
symbol
as
proof
into
Specific
goals
included
optimizing
embedding
process
high
fidelity
embedded
watermark
evaluating
fuzziness
maintain
visual
quality
watermarked
image.
The
methods
in
this
were
transformation,
which
are
implemented
because
their
complementary
strengths.
Singular
offers
robustness
stability,
while
provides
efficient
frequency
domain
thereby
increasing
overall
effectiveness
process.
results
study
showed
efficacy
Lena
image
gray
scale
having
mean
square
error
0.0001,
peak
signal-to-noise
ratio
89.13
decibels
(dB),
universal
index
0.9945,
similarity
structural
0.999.
These
findings
confirmed
proposed
approach
maintains
providing
resistance.
In
conclusion,
contributed
new
designed
verify
institutional
images,
specifically
benefiting
University.
It
significant
potential
for
wider
application
rights
management.
Research Square (Research Square),
Год журнала:
2023,
Номер
unknown
Опубликована: Июнь 9, 2023
Abstract
Two-factor
authentication
(2FA)
method
provides
an
additional
layer
of
security
to
the
user
accounts
and
systems
beyond
a
single
factor
like
simple
password.
Now
days
biometric
based
is
widely
adopted
as
it
reduces
impersonation
fraud
account
takeover
attacks.
Biometric
data
are
relevant
with
user’s
personal
information
can
potentially
be
exploited
by
attacker
in
future
compromise
user.
Hence
protection
also
vital
part
along
secure
protected
data.
In
this
proposed
work,
2FA
mechanism
implemented
using
Spread
Spectrum
Watermarking
method.
Instead
storing
fingerprint
database,
embedded
physical
token/security
token
invisible
watermarking
image.
user's
unique
ID
(UID)
stored
database
watermark
or
token,
which
serves
second
factor.
Once
both
factors
other
embedded,
becomes
smart
card.
To
access
data,
compared
on
card
for
identification.
further
validate
identity,
UID
decrypted
matched
against
database.
The
watermarking-based
model's
enhanced
outcomes
demonstrated
through
detailed
simulation
analysis.
Comparative
analysis
results
confirms
superiority
model
over
traditional
biometric-based
systems.
2022 International Conference on Computational Intelligence and Sustainable Engineering Solutions (CISES),
Год журнала:
2023,
Номер
unknown, С. 993 - 998
Опубликована: Апрель 28, 2023
Presently,
we
are
living
in
a
digital
world
where
can
interact
and
speak
with
anybody
on
the
planet
through
social
networking
sites
such
as
Twitter
Facebook.
is
fastest-growing
service
among
many
others.
Because
of
its
prominence,
it
deals
slew
spammers
that
transmit
falsified
misleading
material.
Researchers
had
tough
time
obtaining
un-spammed
legitimate
information
from
due
to
this
spamming.
In
research,
provided
unique
way
design/create
system
distinguishes
between
spammer
authentic
account
tweets
real-time.
For
dataset,
used
API.
Four
machine
learning
classifiers
Decision
Tree
(DT),
Support
Vector
Machine
(SVM),
Naïve
Bayesian
(NB),
Logistic
Regression
(LR)
were
implemented.
We
compared
performance
these
ML
terms
accuracy,
stability,
scalability
found
Nave
classifier
best
results,
i.e.,
96.83
%
accuracy
all.
