Dilated convolution.

<div><p>This study aims to explore a data-driven cultural background fusion method to improve the accuracy of environmental art image classification. A novel Dual Kernel Squeeze and Excitation Network (DKSE-Net) model is proposed for the complex cultural background and diverse visual rep...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Chenchen Liu (503953) (author)
مؤلفون آخرون: Haoyue Guo (3991031) (author)
منشور في: 2025
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1852021967354331136
author Chenchen Liu (503953)
author2 Haoyue Guo (3991031)
author2_role author
author_facet Chenchen Liu (503953)
Haoyue Guo (3991031)
author_role author
dc.creator.none.fl_str_mv Chenchen Liu (503953)
Haoyue Guo (3991031)
dc.date.none.fl_str_mv 2025-03-20T17:24:56Z
dc.identifier.none.fl_str_mv 10.1371/journal.pone.0313946.g003
dc.relation.none.fl_str_mv https://figshare.com/articles/figure/Dilated_convolution_/28634219
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Biotechnology
Ecology
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
receptive fields using
experimental results indicate
diverse visual representation
deep learning technology
7 %, 3
complex cultural background
selective kernel network
model &# 8217
complex cultural backgrounds
global average pooling
environmental art images
dual kernel squeeze
feature capture ability
layer convolution process
second layer convolution
higher classification accuracy
net model achieves
cultural backgrounds
initial layer
processing images
generalization ability
excitation network
classification accuracy
art models
feature maps
feature extraction
model combines
pointwise convolution
dilated convolution
depthwise convolution
xlink ">
thus enhancing
technical support
study aims
senet ).
prevent overfitting
performance provides
local features
l2 regularization
important reference
existing state
comprehensively extract
comparative models
broad potential
batch normalization
adaptive adjustment
dc.title.none.fl_str_mv Dilated convolution.
dc.type.none.fl_str_mv Image
Figure
info:eu-repo/semantics/publishedVersion
image
description <div><p>This study aims to explore a data-driven cultural background fusion method to improve the accuracy of environmental art image classification. A novel Dual Kernel Squeeze and Excitation Network (DKSE-Net) model is proposed for the complex cultural background and diverse visual representation in environmental art images. This model combines the advantages of adaptive adjustment of receptive fields using the Selective Kernel Network (SKNet) and the characteristics of enhancing channel features using the Squeeze and Excitation Network (SENet). Constructing a DKSE module can comprehensively extract the global and local features of the image. The DKSE module adopts various techniques such as dilated convolution, L2 regularization, Dropout, etc. in the multi-layer convolution process. Firstly, dilated convolution is introduced into the initial layer of the model to enhance the original art image’s feature capture ability. Secondly, the pointwise convolution is constrained by L2 regularization, thus enhancing the accuracy and stability of the convolution. Finally, the Dropout technology randomly discards the feature maps before and after global average pooling to prevent overfitting and improve the model’s generalization ability. On this basis, the Rectified Linear Unit activation function and depthwise convolution are introduced after the second layer convolution, and batch normalization is performed to improve the efficiency and robustness of feature extraction. The experimental results indicate that the proposed DKSE-Net model significantly outperforms traditional Convolutional Neural Networks (CNNs) and other existing state-of-the-art models in the task of environmental art image classification. Specifically, the DKSE-Net model achieves a classification accuracy of 92.7%, 3.5 percentage points higher than the comparative models. Moreover, when processing images with complex cultural backgrounds, DKSE-Net can effectively integrate different cultural features, achieving a higher classification accuracy and stability. This enhancement in performance provides an important reference for image classification research based on the fusion of cultural backgrounds and demonstrates the broad potential of deep learning technology in the environmental art field.</p></div>
eu_rights_str_mv openAccess
id Manara_18ff482d2dce167a23ec2fa601b6fd7a
identifier_str_mv 10.1371/journal.pone.0313946.g003
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/28634219
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Dilated convolution.Chenchen Liu (503953)Haoyue Guo (3991031)BiotechnologyEcologySpace ScienceBiological Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedreceptive fields usingexperimental results indicatediverse visual representationdeep learning technology7 %, 3complex cultural backgroundselective kernel networkmodel &# 8217complex cultural backgroundsglobal average poolingenvironmental art imagesdual kernel squeezefeature capture abilitylayer convolution processsecond layer convolutionhigher classification accuracynet model achievescultural backgroundsinitial layerprocessing imagesgeneralization abilityexcitation networkclassification accuracyart modelsfeature mapsfeature extractionmodel combinespointwise convolutiondilated convolutiondepthwise convolutionxlink ">thus enhancingtechnical supportstudy aimssenet ).prevent overfittingperformance provideslocal featuresl2 regularizationimportant referenceexisting statecomprehensively extractcomparative modelsbroad potentialbatch normalizationadaptive adjustment<div><p>This study aims to explore a data-driven cultural background fusion method to improve the accuracy of environmental art image classification. A novel Dual Kernel Squeeze and Excitation Network (DKSE-Net) model is proposed for the complex cultural background and diverse visual representation in environmental art images. This model combines the advantages of adaptive adjustment of receptive fields using the Selective Kernel Network (SKNet) and the characteristics of enhancing channel features using the Squeeze and Excitation Network (SENet). Constructing a DKSE module can comprehensively extract the global and local features of the image. The DKSE module adopts various techniques such as dilated convolution, L2 regularization, Dropout, etc. in the multi-layer convolution process. Firstly, dilated convolution is introduced into the initial layer of the model to enhance the original art image’s feature capture ability. Secondly, the pointwise convolution is constrained by L2 regularization, thus enhancing the accuracy and stability of the convolution. Finally, the Dropout technology randomly discards the feature maps before and after global average pooling to prevent overfitting and improve the model’s generalization ability. On this basis, the Rectified Linear Unit activation function and depthwise convolution are introduced after the second layer convolution, and batch normalization is performed to improve the efficiency and robustness of feature extraction. The experimental results indicate that the proposed DKSE-Net model significantly outperforms traditional Convolutional Neural Networks (CNNs) and other existing state-of-the-art models in the task of environmental art image classification. Specifically, the DKSE-Net model achieves a classification accuracy of 92.7%, 3.5 percentage points higher than the comparative models. Moreover, when processing images with complex cultural backgrounds, DKSE-Net can effectively integrate different cultural features, achieving a higher classification accuracy and stability. This enhancement in performance provides an important reference for image classification research based on the fusion of cultural backgrounds and demonstrates the broad potential of deep learning technology in the environmental art field.</p></div>2025-03-20T17:24:56ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0313946.g003https://figshare.com/articles/figure/Dilated_convolution_/28634219CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/286342192025-03-20T17:24:56Z
spellingShingle Dilated convolution.
Chenchen Liu (503953)
Biotechnology
Ecology
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
receptive fields using
experimental results indicate
diverse visual representation
deep learning technology
7 %, 3
complex cultural background
selective kernel network
model &# 8217
complex cultural backgrounds
global average pooling
environmental art images
dual kernel squeeze
feature capture ability
layer convolution process
second layer convolution
higher classification accuracy
net model achieves
cultural backgrounds
initial layer
processing images
generalization ability
excitation network
classification accuracy
art models
feature maps
feature extraction
model combines
pointwise convolution
dilated convolution
depthwise convolution
xlink ">
thus enhancing
technical support
study aims
senet ).
prevent overfitting
performance provides
local features
l2 regularization
important reference
existing state
comprehensively extract
comparative models
broad potential
batch normalization
adaptive adjustment
status_str publishedVersion
title Dilated convolution.
title_full Dilated convolution.
title_fullStr Dilated convolution.
title_full_unstemmed Dilated convolution.
title_short Dilated convolution.
title_sort Dilated convolution.
topic Biotechnology
Ecology
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
receptive fields using
experimental results indicate
diverse visual representation
deep learning technology
7 %, 3
complex cultural background
selective kernel network
model &# 8217
complex cultural backgrounds
global average pooling
environmental art images
dual kernel squeeze
feature capture ability
layer convolution process
second layer convolution
higher classification accuracy
net model achieves
cultural backgrounds
initial layer
processing images
generalization ability
excitation network
classification accuracy
art models
feature maps
feature extraction
model combines
pointwise convolution
dilated convolution
depthwise convolution
xlink ">
thus enhancing
technical support
study aims
senet ).
prevent overfitting
performance provides
local features
l2 regularization
important reference
existing state
comprehensively extract
comparative models
broad potential
batch normalization
adaptive adjustment