Cargando…

Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images

This study aimed to address the problems of low detection accuracy and inaccurate positioning of small-object detection in remote sensing images. An improved architecture based on the Swin Transformer and YOLOv5 is proposed. First, Complete-IOU (CIOU) was introduced to improve the K-means clustering...

Descripción completa

Detalles Bibliográficos
Autores principales: Cao, Xuan, Zhang, Yanwei, Lang, Song, Gong, Yan
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10098803/
https://www.ncbi.nlm.nih.gov/pubmed/37050694
http://dx.doi.org/10.3390/s23073634
_version_ 1785024902124797952
author Cao, Xuan
Zhang, Yanwei
Lang, Song
Gong, Yan
author_facet Cao, Xuan
Zhang, Yanwei
Lang, Song
Gong, Yan
author_sort Cao, Xuan
collection PubMed
description This study aimed to address the problems of low detection accuracy and inaccurate positioning of small-object detection in remote sensing images. An improved architecture based on the Swin Transformer and YOLOv5 is proposed. First, Complete-IOU (CIOU) was introduced to improve the K-means clustering algorithm, and then an anchor of appropriate size for the dataset was generated. Second, a modified CSPDarknet53 structure combined with Swin Transformer was proposed to retain sufficient global context information and extract more differentiated features through multi-head self-attention. Regarding the path-aggregation neck, a simple and efficient weighted bidirectional feature pyramid network was proposed for effective cross-scale feature fusion. In addition, extra prediction head and new feature fusion layers were added for small objects. Finally, Coordinate Attention (CA) was introduced to the YOLOv5 network to improve the accuracy of small-object features in remote sensing images. Moreover, the effectiveness of the proposed method was demonstrated by several kinds of experiments on the DOTA (Dataset for Object detection in Aerial images). The mean average precision on the DOTA dataset reached 74.7%. Compared with YOLOv5, the proposed method improved the mean average precision (mAP) by 8.9%, which can achieve a higher accuracy of small-object detection in remote sensing images.
format Online
Article
Text
id pubmed-10098803
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-100988032023-04-14 Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images Cao, Xuan Zhang, Yanwei Lang, Song Gong, Yan Sensors (Basel) Article This study aimed to address the problems of low detection accuracy and inaccurate positioning of small-object detection in remote sensing images. An improved architecture based on the Swin Transformer and YOLOv5 is proposed. First, Complete-IOU (CIOU) was introduced to improve the K-means clustering algorithm, and then an anchor of appropriate size for the dataset was generated. Second, a modified CSPDarknet53 structure combined with Swin Transformer was proposed to retain sufficient global context information and extract more differentiated features through multi-head self-attention. Regarding the path-aggregation neck, a simple and efficient weighted bidirectional feature pyramid network was proposed for effective cross-scale feature fusion. In addition, extra prediction head and new feature fusion layers were added for small objects. Finally, Coordinate Attention (CA) was introduced to the YOLOv5 network to improve the accuracy of small-object features in remote sensing images. Moreover, the effectiveness of the proposed method was demonstrated by several kinds of experiments on the DOTA (Dataset for Object detection in Aerial images). The mean average precision on the DOTA dataset reached 74.7%. Compared with YOLOv5, the proposed method improved the mean average precision (mAP) by 8.9%, which can achieve a higher accuracy of small-object detection in remote sensing images. MDPI 2023-03-31 /pmc/articles/PMC10098803/ /pubmed/37050694 http://dx.doi.org/10.3390/s23073634 Text en © 2023 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Cao, Xuan
Zhang, Yanwei
Lang, Song
Gong, Yan
Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images
title Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images
title_full Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images
title_fullStr Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images
title_full_unstemmed Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images
title_short Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images
title_sort swin-transformer-based yolov5 for small-object detection in remote sensing images
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10098803/
https://www.ncbi.nlm.nih.gov/pubmed/37050694
http://dx.doi.org/10.3390/s23073634
work_keys_str_mv AT caoxuan swintransformerbasedyolov5forsmallobjectdetectioninremotesensingimages
AT zhangyanwei swintransformerbasedyolov5forsmallobjectdetectioninremotesensingimages
AT langsong swintransformerbasedyolov5forsmallobjectdetectioninremotesensingimages
AT gongyan swintransformerbasedyolov5forsmallobjectdetectioninremotesensingimages