Main architecture of Swin Transformer.
<div><p>The traditional method of corn quality detection relies heavily on the subjective judgment of inspectors and suffers from a high error rate. To address these issues, this study employs the Swin Transformer as an enhanced base model, integrating machine vision and deep learning te...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | , , , |
| منشور في: |
2025
|
| الموضوعات: | |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| _version_ | 1852023305837477888 |
|---|---|
| author | Ning Zhang (23771) |
| author2 | Yuanqi Chen (4526056) Enxu Zhang (20613821) Ziyang Liu (242055) Jie Yue (1597459) |
| author2_role | author author author author |
| author_facet | Ning Zhang (23771) Yuanqi Chen (4526056) Enxu Zhang (20613821) Ziyang Liu (242055) Jie Yue (1597459) |
| author_role | author |
| dc.creator.none.fl_str_mv | Ning Zhang (23771) Yuanqi Chen (4526056) Enxu Zhang (20613821) Ziyang Liu (242055) Jie Yue (1597459) |
| dc.date.none.fl_str_mv | 2025-01-24T18:26:28Z |
| dc.identifier.none.fl_str_mv | 10.1371/journal.pone.0312363.g005 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/figure/Main_architecture_of_Swin_Transformer_/28271930 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Biotechnology Science Policy Space Science Biological Sciences not elsewhere classified Chemical Sciences not elsewhere classified Information Systems not elsewhere classified integrating machine vision enhanced base model experimental results demonstrate deep learning techniques recognition accuracy rate high error rate extracted features undergo corn quality detection corn quality assessment 152 valid images broken corn experimental samples deep features xlink "> traditional method technical approach swint high swin transformer subsequently fused subjective judgment novel perspective maize images key metrics fused features f1 score convolutional module attention layer 89 %. |
| dc.title.none.fl_str_mv | Main architecture of Swin Transformer. |
| dc.type.none.fl_str_mv | Image Figure info:eu-repo/semantics/publishedVersion image |
| description | <div><p>The traditional method of corn quality detection relies heavily on the subjective judgment of inspectors and suffers from a high error rate. To address these issues, this study employs the Swin Transformer as an enhanced base model, integrating machine vision and deep learning techniques for corn quality assessment. Initially, images of high-quality, moldy, and broken corn were collected. After preprocessing, a total of 20,152 valid images were obtained for the experimental samples. The network then extracts both shallow and deep features from these maize images, which are subsequently fused. Concurrently, the extracted features undergo further processing through a specially designed convolutional block. The fused features, combined with those processed by the convolutional module, are fed into an attention layer. This attention layer assigns weights to the features, facilitating accurate final classification. Experimental results demonstrate that the MC-Swin Transformer model proposed in this paper significantly outperforms traditional convolutional neural network models in key metrics such as accuracy, precision, recall, and F1 score, achieving a recognition accuracy rate of 99.89%. Thus, the network effectively and efficiently classifies different corn qualities. This study not only offers a novel perspective and technical approach to corn quality detection but also holds significant implications for the advancement of smart agriculture.</p></div> |
| eu_rights_str_mv | openAccess |
| id | Manara_7dfe8d22c10b3f0bd792c2f3087c8d7c |
| identifier_str_mv | 10.1371/journal.pone.0312363.g005 |
| network_acronym_str | Manara |
| network_name_str | ManaraRepo |
| oai_identifier_str | oai:figshare.com:article/28271930 |
| publishDate | 2025 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | Main architecture of Swin Transformer.Ning Zhang (23771)Yuanqi Chen (4526056)Enxu Zhang (20613821)Ziyang Liu (242055)Jie Yue (1597459)BiotechnologyScience PolicySpace ScienceBiological Sciences not elsewhere classifiedChemical Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedintegrating machine visionenhanced base modelexperimental results demonstratedeep learning techniquesrecognition accuracy ratehigh error rateextracted features undergocorn quality detectioncorn quality assessment152 valid imagesbroken cornexperimental samplesdeep featuresxlink ">traditional methodtechnical approachswint highswin transformersubsequently fusedsubjective judgmentnovel perspectivemaize imageskey metricsfused featuresf1 scoreconvolutional moduleattention layer89 %.<div><p>The traditional method of corn quality detection relies heavily on the subjective judgment of inspectors and suffers from a high error rate. To address these issues, this study employs the Swin Transformer as an enhanced base model, integrating machine vision and deep learning techniques for corn quality assessment. Initially, images of high-quality, moldy, and broken corn were collected. After preprocessing, a total of 20,152 valid images were obtained for the experimental samples. The network then extracts both shallow and deep features from these maize images, which are subsequently fused. Concurrently, the extracted features undergo further processing through a specially designed convolutional block. The fused features, combined with those processed by the convolutional module, are fed into an attention layer. This attention layer assigns weights to the features, facilitating accurate final classification. Experimental results demonstrate that the MC-Swin Transformer model proposed in this paper significantly outperforms traditional convolutional neural network models in key metrics such as accuracy, precision, recall, and F1 score, achieving a recognition accuracy rate of 99.89%. Thus, the network effectively and efficiently classifies different corn qualities. This study not only offers a novel perspective and technical approach to corn quality detection but also holds significant implications for the advancement of smart agriculture.</p></div>2025-01-24T18:26:28ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0312363.g005https://figshare.com/articles/figure/Main_architecture_of_Swin_Transformer_/28271930CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/282719302025-01-24T18:26:28Z |
| spellingShingle | Main architecture of Swin Transformer. Ning Zhang (23771) Biotechnology Science Policy Space Science Biological Sciences not elsewhere classified Chemical Sciences not elsewhere classified Information Systems not elsewhere classified integrating machine vision enhanced base model experimental results demonstrate deep learning techniques recognition accuracy rate high error rate extracted features undergo corn quality detection corn quality assessment 152 valid images broken corn experimental samples deep features xlink "> traditional method technical approach swint high swin transformer subsequently fused subjective judgment novel perspective maize images key metrics fused features f1 score convolutional module attention layer 89 %. |
| status_str | publishedVersion |
| title | Main architecture of Swin Transformer. |
| title_full | Main architecture of Swin Transformer. |
| title_fullStr | Main architecture of Swin Transformer. |
| title_full_unstemmed | Main architecture of Swin Transformer. |
| title_short | Main architecture of Swin Transformer. |
| title_sort | Main architecture of Swin Transformer. |
| topic | Biotechnology Science Policy Space Science Biological Sciences not elsewhere classified Chemical Sciences not elsewhere classified Information Systems not elsewhere classified integrating machine vision enhanced base model experimental results demonstrate deep learning techniques recognition accuracy rate high error rate extracted features undergo corn quality detection corn quality assessment 152 valid images broken corn experimental samples deep features xlink "> traditional method technical approach swint high swin transformer subsequently fused subjective judgment novel perspective maize images key metrics fused features f1 score convolutional module attention layer 89 %. |