Semi-supervised and weakly-supervised deep neural networks and dataset for fish detection in turbid underwater videos
Ecological Informatics,
Год журнала:
2023,
Номер
78, С. 102303 - 102303
Опубликована: Сен. 11, 2023
Fish
are
key
members
of
marine
ecosystems,
and
they
have
a
significant
share
in
the
healthy
human
diet.
Besides,
fish
abundance
is
an
excellent
indicator
water
quality,
as
adapted
to
various
levels
oxygen,
turbidity,
nutrients,
pH.
To
detect
underwater
videos,
Deep
Neural
Networks
(DNNs)
can
be
great
assistance.
However,
training
DNNs
highly
dependent
on
large,
labeled
datasets,
while
labeling
turbid
video
frames
laborious
time-consuming
task,
hindering
development
accurate
efficient
models
for
detection.
address
this
problem,
firstly,
we
collected
dataset
called
FishInTurbidWater,
which
consists
collection
footage
gathered
from
waters,
quickly
weakly
(i.e.,
giving
higher
priority
speed
over
accuracy)
them
4-times
fast-forwarding
software.
Next,
designed
implemented
semi-supervised
contrastive
learning
detection
model
that
self-supervised
using
unlabeled
data,
then
fine-tuned
with
small
fraction
(20%)
our
FishInTurbidWater
data.
At
next
step,
trained,
novel
weakly-supervised
ensemble
DNN
transfer
ImageNet.
The
results
show
leads
more
than
20
times
faster
turnaround
time
between
result
generation,
reasonably
high
accuracy
(89%).
same
time,
proposed
waters
(94%)
accuracy,
still
cutting
by
factor
four,
compared
fully-supervised
trained
carefully
datasets.
Our
code
publicly
available
at
hyperlink
FishInTurbidWater.
Язык: Английский
Spatio-temporal heterogeneity of ecological water level in Poyang Lake, China
Ecological Informatics,
Год журнала:
2024,
Номер
82, С. 102694 - 102694
Опубликована: Июнь 18, 2024
Anthropogenic
activities
and
climate
change
have
caused
physical
ecological
changes
in
lakes
aggravated
water
level
fluctuations,
which
are
essential
factors
to
consider
for
nutrient
import,
protection,
biodiversity
maintenance.
Maintaining
levels
within
a
reasonable
range
is
maintaining
lake
function
health,
because
ecosystem
stability
compromised
when
fluctuations
exceed
specific
thresholds.
Thus,
the
(EWL)
an
important
index
aquatic
habitats
biodiversity.
A
method
quantifying
EWL
of
based
on
hydrological
statistical
analysis
was
constructed
bridge
gaps
existing
studies,
considering
both
alteration
spatio-temporal
heterogeneity
fluctuations.
Taking
Poyang
Lake
as
example,
has
recently
attracted
increasing
global
attention
owing
its
alterations
subsequent
problems,
applicability
rationality
results
were
verified.
The
indicate
that
occurs
at
representative
stations,
jointly
affected
by
anthropogenic
this
region.
For
instance,
construction
operation
Three
Gorges
Project
Hukou
Xingzi
station,
drought
further
station.
calculated
showed
obvious
heterogeneity,
consistent
with
topographic,
geographical,
climatic
characteristics
basin.
And
study
verified
through
literature
reviews
satisfiability
characteristic
species
requirements.
proposed
calculation
simple
feasible
easy
data
acquisition,
strong
universality,
broad
application
prospects,
offering
scientific
basis
quantitative
reference
resource
management
protection.
Язык: Английский
AI-driven forecasting of river discharge: the case study of the Himalayan mountainous river
Earth Science Informatics,
Год журнала:
2025,
Номер
18(2)
Опубликована: Фев. 1, 2025
Язык: Английский
Maximum energy entropy: A novel signal preprocessing approach for data-driven monthly streamflow forecasting
Ecological Informatics,
Год журнала:
2023,
Номер
79, С. 102452 - 102452
Опубликована: Дек. 28, 2023
In
recent
years,
the
application
of
Data-Driven
Models
(DDMs)
in
ecological
studies
has
garnered
significant
attention
due
to
their
capacity
accurately
simulate
complex
hydrological
processes.
These
models
have
proven
invaluable
comprehending
and
predicting
natural
phenomena.
However,
achieve
improved
outcomes,
certain
additive
components
such
as
signal
analysis
(SAM)
input
variable
selections
(IVS)
are
necessary.
SAMs
unveil
hidden
characteristics
within
time
series
data,
while
IVS
prevents
utilization
inappropriate
data.
realm
research,
understanding
these
patterns
is
pivotal
for
grasping
implications
streamflow
dynamics
guiding
effective
management
decisions.
Addressing
need
more
precise
forecasting,
this
study
proposes
a
novel
SAM
called
"Maximum
Energy
Entropy
(MEE)"
forecast
monthly
Ajichai
basin,
located
northwestern
Iran.
A
comparative
was
conducted,
pitting
MEE
against
well-known
methods
Discreet
Wavelet
(DW)
Wavelet-Entropy
(DWE),
ultimately
demonstrating
superiority
MEE.
The
results
showcased
superior
performance
our
proposed
method,
with
an
NSE
value
0.72,
compared
DW
(NSE
0.68)
DWE
0.68).
Furthermore,
exhibited
greater
reliability,
boasting
lower
Standard
Deviation
0.13
(0.26)
(0.19).
equips
researchers
decision-makers
accurate
predictions,
facilitating
well-informed
water
resource
planning.
To
further
evaluate
MEE's
accuracy
using
various
DDMs,
we
integrated
Artificial
Neural
Network
(ANN)
Genetic
Programming
(GP).
Additionally,
GP
served
method
selecting
appropriate
variables.
Ultimately,
combination
ANN
forecasting
model
(MEE-GP-ANN)
yielded
most
favorable
results.
Язык: Английский
Comprehensive Ecological Functional Zoning: A Data-Driven Approach for Sustainable Land Use and Environmental Management—A Case Study in Shenzhen, China
Land,
Год журнала:
2024,
Номер
13(9), С. 1413 - 1413
Опубликована: Сен. 2, 2024
A
comprehensive
approach
to
ecological
functional
zoning
in
the
Shenzhen
region
of
China
is
presented
this
study.
Through
integration
advanced
geospatial
analysis
tools,
multiple
data
sources,
and
sophisticated
statistical
techniques,
different
functions
have
been
identified
categorized
based
on
a
set
indicators
spatial
techniques.
The
three-level
framework
established
study
offers
policymakers,
urban
planners,
environmental
managers
nuanced
understanding
region’s
characteristics,
highlights
areas
significance
that
warrant
special
attention
protection.
It
has
demonstrated
data-driven
effective
delineating
distinct
zones
within
area.
This
study’s
findings
carry
significant
implications
for
future
land
use
planning,
conservation
efforts,
sustainable
development
practices
region.
In
essence,
contributes
broader
discourse
planning
management
by
providing
systematic
urbanizing
regions.
Язык: Английский
Morphological Model for Erosion Prediction of India’s Largest Braided River Using MIKE 21C Model
Earth Science Systems and Society,
Год журнала:
2024,
Номер
4
Опубликована: Янв. 15, 2024
The
Brahmaputra
River
has
a
dynamic,
highly
braided
channel
pattern
with
frequent
river
bar
formation,
making
it
morphologically
very
especially
during
the
monsoon
season
high
discharge
and
sediment
load.
To
understand
how
changes
over
time,
this
study
focused
on
two
stretches:
Palasbari-Gumi
Dibrugarh.
Using
2D
morphological
models
(MIKE-21C),
aimed
to
predict
erosion
patterns,
plan
protective
measures,
assess
short-term
(1
year),
medium-term
(3
long-term
(5
year)
periods.
Model
runs
were
conducted
design
variables
across
these
reaches,
encompassing
different
hydrological
scenarios
development-planning
scenarios.
coarse
sand
fraction
yielded
mean
annual
load
predictions
of
257
Mt/year
for
2021
year
314
under
bankfull
conditions
in
reach.
In
Dibrugarh
reach,
corresponding
values
78
100
Mt/year.
Notably,
historical
records
indicate
an
400
River.
model
results
compared
measurements
from
Acoustic
Doppler
Current
Profilers
(ADCP),
showing
good
accuracy
flow
velocities,
flood
levels,
loads.
Discrepancies
peak
velocities
ADCP
remain
consistently
below
9%
majority
recorded
data
points.
predicted
levels
condition
exhibited
outstanding
accuracy,
reaching
nearly
91%
at
site
notable
95%
site.
This
presented
valuable
methodology
enhancing
strategic
planning
implementation
training
endeavours,
particularly
within
dynamic
channels
rivers
such
as
approach
leverages
predictive
2–3
years
timeframe,
contributing
improved
management.
Язык: Английский
Flood Risk Assessment Basing on Flood Flow Modeling in the Oued Martil Region, Western Part of Northern Morocco
Ecological Engineering & Environmental Technology,
Год журнала:
2024,
Номер
25(5), С. 243 - 255
Опубликована: Март 29, 2024
In
the
context
of
climate
change,
risk
flooding
is
becoming
an
increasingly
global
concern.In
addition,
natural
factors,
economic
development
and
urban
expansion
are
significant
contributors
that
have
generated
a
strong
demand
for
management
risks,
especially
in
domain
floods
inundations.This
research
aims
to
address
issue
flood
Oued
Martil
region,
specifically
within
cities
Tetouan
western
part
Northern
Morocco.In
this
regard,
study
focuses
on
evaluating
performance
hydrological
analysis
plain
modeling
flows
bed
overflow
area.The
results
show
urbanized
densely
populated
areas
(with
high
vulnerability)
match
with
zones
or
moderate
hazard.Conversely,
damage
lower
forests
situated
low
hazard.The
obtained
from
hydraulic
can
assist
decision-makers
selecting
types
interventions
floodplain
by
providing
comprehensive
understanding
Martil's
behavior
during
exceedance
peak
flow
rates
different
return
periods.
Язык: Английский
Integrated Influence of Changing LULC and Aridity on Runoff Curve Numbers
Prashant Prashant,
Surendra Kumar Mishra,
A. K. Lohani
и другие.
Research Square (Research Square),
Год журнала:
2024,
Номер
unknown
Опубликована: Май 29, 2024
Abstract
The
popular
Soil
Conservation
Service-Curve
Number
(SCS-CN)
method
is
widely
used
for
direct
surface
runoff
estimation
from
a
given
amount
of
rainfall
in
watershed.
present
urban
sprawl,
socioeconomic
anthropogenic
activities,
and
environmental
changes
all
have
affected
the
cosmic
extent
land
use-land
cover
(LULC)
complex
climate,
both
spatially
temporally,
which
directly
affect
parameter
curve
number
(CN)
and,
turn,
runoff.
Therefore,
study
propels
disparity
representative
CNs
SCS-CN
methodology,
usually
derived
NEH-4
tables
based
on
use
soil
type
(CN
LU−ST)
observed
rainfall(P)-runoff(Q)
events
(CN
P−Q).
annual
series
CN
P−Q
CN
LU−ST
(from
1980
to
2020)
showed
existence
trends
inconsistency
between
Ong
River
basin
(India).
alteration
analysis
utilized
supervised
machine
learning
algorithm
indicated
two
major
LULC
classes
as
contributing
factors
increasing
CNs.
Furthermore,
attributes
implications
shifting
dynamics
(~
70%)
climate
variations
30%)
Employing
Aridity
Index
(AI),
solving
annual/decadal
values
revealed
strong
evidence
with
fit
high
R
2
range
(0.72,
0.99)
aridity
influencing
Язык: Английский
Emergy-Based Evaluation of the Sustainability of Agricultural Ecosystem in Dazhou, China, from 2002 to 2022
Sustainability,
Год журнала:
2024,
Номер
16(21), С. 9156 - 9156
Опубликована: Окт. 22, 2024
Our
aim
is
to
analyze
the
emergy
evaluation
indicators
of
agricultural
ecosystem
in
Dazhou,
northeastern
Sichuan,
and
provide
practical
effective
recommendations
for
sustainable
development.
Using
analysis,
inputs
outputs
an
from
2002
2022
were
calculated.
Five
selected
evaluation:
yield
ratio
(EYR),
self-sufficiency
(ESR),
input
(EIR),
environmental
load
(ELR),
indices
(ESI).
The
total
showed
upward
trend
2017,
thus
industrial
auxiliary
decreased,
somewhat
curbing
its
continued
rise
2017
2022.
structure
inputs,
descending
order,
as
follows:
>
organic
renewable
resources
non-renewable
resources.
output
was
highest
2007,
reaching
2.31
×
1022
Sej,
lowest
2012,
at
1.83
Sej.
outputs,
livestock
planting
fishery
forestry.
fluctuated
down
3.12
2.51,
with
average
2.88,
below
provincial
3.07.
0.30
0.26,
0.27,
above
0.13.
up
2.91,
2.66,
1.86.
3.8
4.75,
4.40,
which
higher
than
1.68.
0.81
0.53,
0.67,
1.17.
efficiency
resource
utilization
Dazhou
has
economic
have
increased,
it
a
consumptive
production
process.
pressure
on
local
natural
environment
increasing,
capacity
development
remains
low
level
over
long
term.
Язык: Английский