LSKA Network Structure.

<div><p>At present, the low-altitude economy is booming, and the application of drones has shown explosive growth, injecting new vitality into economic development. UAVs will face complex environmental perception and security risks when operating in low airspace. Accurate target detectio...

Full description

Saved in:
Bibliographic Details
Main Author: Xinwei Wang (352488) (author)
Other Authors: Yue Hu (201714) (author), Qing Liang (654718) (author), Yujie He (3068052) (author), Li Zhou (54356) (author)
Published: 2025
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1852017392518955008
author Xinwei Wang (352488)
author2 Yue Hu (201714)
Qing Liang (654718)
Yujie He (3068052)
Li Zhou (54356)
author2_role author
author
author
author
author_facet Xinwei Wang (352488)
Yue Hu (201714)
Qing Liang (654718)
Yujie He (3068052)
Li Zhou (54356)
author_role author
dc.creator.none.fl_str_mv Xinwei Wang (352488)
Yue Hu (201714)
Qing Liang (654718)
Yujie He (3068052)
Li Zhou (54356)
dc.date.none.fl_str_mv 2025-08-21T18:01:57Z
dc.identifier.none.fl_str_mv 10.1371/journal.pone.0327732.g006
dc.relation.none.fl_str_mv https://figshare.com/articles/figure/LSKA_Network_Structure_/29963191
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
weight file size
uav autonomous perception
shown explosive growth
insufficient feature capture
injecting new vitality
extract critical features
enrich feature fusion
efficient feature extraction
enabling convolution operations
9 %, 4
sppf module incorporates
scale feature processing
mechanism captures long
mean average precision
improve detection accuracy
algorithm introduces akconv
model &# 8217
large convolution kernels
improved algorithm achieves
33 %, respectively
low altitude economy
convolution kernels
altitude economy
detection precision
original algorithm
scale searches
paper achieves
lska mechanism
improved yolov8s
c2f module
akconv allows
accurate large
detection recall
5 %,
xlink ">
thereby achieving
study focuses
security risks
sampling shapes
range dependencies
precisely adapt
perform rapid
parameter count
orderly operation
model lightweighting
low efficiency
low airspace
key support
experimental results
economic development
dc.title.none.fl_str_mv LSKA Network Structure.
dc.type.none.fl_str_mv Image
Figure
info:eu-repo/semantics/publishedVersion
image
description <div><p>At present, the low-altitude economy is booming, and the application of drones has shown explosive growth, injecting new vitality into economic development. UAVs will face complex environmental perception and security risks when operating in low airspace. Accurate target detection technology has become a key support to ensure the orderly operation of UAVs. This paper studies UAV target detection algorithm based on deep learning, in order to improve detection accuracy and speed, and meet the needs of UAV autonomous perception under the background of low altitude economy. This study focuses on the limitations of the YOLOv8s target detection algorithm, including its low efficiency in multi-scale feature processing and insufficient small target detection capability, which hinder its ability to perform rapid and accurate large-scale searches for drones. An improved target detection algorithm is proposed to address these issues. The algorithm introduces AKConv into the C2F module. AKConv allows for convolution kernels with arbitrary numbers and sampling shapes, enabling convolution operations to more precisely adapt to targets at different locations, thereby achieving more efficient feature extraction. To further enhance the model’s ability to extract critical features of small targets, the SPPF module incorporates the LSKA mechanism. This mechanism captures long-range dependencies and adaptivity more effectively while addressing computational complexity issues associated with large convolution kernels. Finally, the Bi-FPN feature pyramid network structure is introduced at the 18th layer of the model to accelerate and enrich feature fusion in the neck. Combined with the SCDown structure, a novel Bi-SCDown-FPN feature pyramid network structure is proposed, making it more suitable for detecting targets with insufficient feature capture in complex environments. Experimental results on the VisDrone2019 UAV dataset show that the improved algorithm achieves a 5.9%, 4.5%, and 6.1% increase in detection precision, detection recall, and mean average precision, respectively, compared to the original algorithm. Moreover, the parameter count and weight file size are reduced by 13.41% and 13.33%, respectively. Compared to other mainstream target detection algorithms, the proposed method demonstrates certain advantages. In summary, the target detection algorithm proposed in this paper achieves a dual improvement in model lightweighting and detection accuracy.</p></div>
eu_rights_str_mv openAccess
id Manara_bd3c533a5ce47a5259c9b2efd47d36d3
identifier_str_mv 10.1371/journal.pone.0327732.g006
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/29963191
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling LSKA Network Structure.Xinwei Wang (352488)Yue Hu (201714)Qing Liang (654718)Yujie He (3068052)Li Zhou (54356)Space ScienceBiological Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedweight file sizeuav autonomous perceptionshown explosive growthinsufficient feature captureinjecting new vitalityextract critical featuresenrich feature fusionefficient feature extractionenabling convolution operations9 %, 4sppf module incorporatesscale feature processingmechanism captures longmean average precisionimprove detection accuracyalgorithm introduces akconvmodel &# 8217large convolution kernelsimproved algorithm achieves33 %, respectivelylow altitude economyconvolution kernelsaltitude economydetection precisionoriginal algorithmscale searchespaper achieveslska mechanismimproved yolov8sc2f moduleakconv allowsaccurate largedetection recall5 %,xlink ">thereby achievingstudy focusessecurity riskssampling shapesrange dependenciesprecisely adaptperform rapidparameter countorderly operationmodel lightweightinglow efficiencylow airspacekey supportexperimental resultseconomic development<div><p>At present, the low-altitude economy is booming, and the application of drones has shown explosive growth, injecting new vitality into economic development. UAVs will face complex environmental perception and security risks when operating in low airspace. Accurate target detection technology has become a key support to ensure the orderly operation of UAVs. This paper studies UAV target detection algorithm based on deep learning, in order to improve detection accuracy and speed, and meet the needs of UAV autonomous perception under the background of low altitude economy. This study focuses on the limitations of the YOLOv8s target detection algorithm, including its low efficiency in multi-scale feature processing and insufficient small target detection capability, which hinder its ability to perform rapid and accurate large-scale searches for drones. An improved target detection algorithm is proposed to address these issues. The algorithm introduces AKConv into the C2F module. AKConv allows for convolution kernels with arbitrary numbers and sampling shapes, enabling convolution operations to more precisely adapt to targets at different locations, thereby achieving more efficient feature extraction. To further enhance the model’s ability to extract critical features of small targets, the SPPF module incorporates the LSKA mechanism. This mechanism captures long-range dependencies and adaptivity more effectively while addressing computational complexity issues associated with large convolution kernels. Finally, the Bi-FPN feature pyramid network structure is introduced at the 18th layer of the model to accelerate and enrich feature fusion in the neck. Combined with the SCDown structure, a novel Bi-SCDown-FPN feature pyramid network structure is proposed, making it more suitable for detecting targets with insufficient feature capture in complex environments. Experimental results on the VisDrone2019 UAV dataset show that the improved algorithm achieves a 5.9%, 4.5%, and 6.1% increase in detection precision, detection recall, and mean average precision, respectively, compared to the original algorithm. Moreover, the parameter count and weight file size are reduced by 13.41% and 13.33%, respectively. Compared to other mainstream target detection algorithms, the proposed method demonstrates certain advantages. In summary, the target detection algorithm proposed in this paper achieves a dual improvement in model lightweighting and detection accuracy.</p></div>2025-08-21T18:01:57ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0327732.g006https://figshare.com/articles/figure/LSKA_Network_Structure_/29963191CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/299631912025-08-21T18:01:57Z
spellingShingle LSKA Network Structure.
Xinwei Wang (352488)
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
weight file size
uav autonomous perception
shown explosive growth
insufficient feature capture
injecting new vitality
extract critical features
enrich feature fusion
efficient feature extraction
enabling convolution operations
9 %, 4
sppf module incorporates
scale feature processing
mechanism captures long
mean average precision
improve detection accuracy
algorithm introduces akconv
model &# 8217
large convolution kernels
improved algorithm achieves
33 %, respectively
low altitude economy
convolution kernels
altitude economy
detection precision
original algorithm
scale searches
paper achieves
lska mechanism
improved yolov8s
c2f module
akconv allows
accurate large
detection recall
5 %,
xlink ">
thereby achieving
study focuses
security risks
sampling shapes
range dependencies
precisely adapt
perform rapid
parameter count
orderly operation
model lightweighting
low efficiency
low airspace
key support
experimental results
economic development
status_str publishedVersion
title LSKA Network Structure.
title_full LSKA Network Structure.
title_fullStr LSKA Network Structure.
title_full_unstemmed LSKA Network Structure.
title_short LSKA Network Structure.
title_sort LSKA Network Structure.
topic Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
weight file size
uav autonomous perception
shown explosive growth
insufficient feature capture
injecting new vitality
extract critical features
enrich feature fusion
efficient feature extraction
enabling convolution operations
9 %, 4
sppf module incorporates
scale feature processing
mechanism captures long
mean average precision
improve detection accuracy
algorithm introduces akconv
model &# 8217
large convolution kernels
improved algorithm achieves
33 %, respectively
low altitude economy
convolution kernels
altitude economy
detection precision
original algorithm
scale searches
paper achieves
lska mechanism
improved yolov8s
c2f module
akconv allows
accurate large
detection recall
5 %,
xlink ">
thereby achieving
study focuses
security risks
sampling shapes
range dependencies
precisely adapt
perform rapid
parameter count
orderly operation
model lightweighting
low efficiency
low airspace
key support
experimental results
economic development