FA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background

Infrared target detection is a popular applied field in object detection as well as a challenge. This paper proposes the focus and attention mechanism-based YOLO (FA-YOLO), which is an improved method to detect the infrared occluded vehicles in the complex background of remote sensing images. Firstl...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Shuangjiang Du, Baofu Zhang, Pin Zhang, Peng Xiang, Hong Xue
Formato:	article
Lenguaje:	EN
Publicado:	Hindawi-Wiley 2021
Materias:	Technology T Telecommunication TK5101-6720
Acceso en línea:	https://doaj.org/article/7122b0d4c9294739a9061d50032458ba
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:7122b0d4c9294739a9061d50032458ba
record_format	dspace
spelling	oai:doaj.org-article:7122b0d4c9294739a9061d50032458ba2021-11-29T00:55:57ZFA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background1530-867710.1155/2021/1896029https://doaj.org/article/7122b0d4c9294739a9061d50032458ba2021-01-01T00:00:00Zhttp://dx.doi.org/10.1155/2021/1896029https://doaj.org/toc/1530-8677Infrared target detection is a popular applied field in object detection as well as a challenge. This paper proposes the focus and attention mechanism-based YOLO (FA-YOLO), which is an improved method to detect the infrared occluded vehicles in the complex background of remote sensing images. Firstly, we use GAN to create infrared images from the visible datasets to make sufficient datasets for training as well as using transfer learning. Then, to mitigate the impact of the useless and complex background information, we propose the negative sample focusing mechanism to focus on the confusing negative sample training to depress the false positives and increase the detection precision. Finally, to enhance the features of the infrared small targets, we add the dilated convolutional block attention module (dilated CBAM) to the CSPdarknet53 in the YOLOv4 backbone. To verify the superiority of our model, we carefully select 318 infrared occluded vehicle images from the VIVID-infrared dataset for testing. The detection accuracy-mAP improves from 79.24% to 92.95%, and the F1 score improves from 77.92% to 88.13%, which demonstrates a significant improvement in infrared small occluded vehicle detection.Shuangjiang DuBaofu ZhangPin ZhangPeng XiangHong XueHindawi-WileyarticleTechnologyTTelecommunicationTK5101-6720ENWireless Communications and Mobile Computing, Vol 2021 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Technology T Telecommunication TK5101-6720
spellingShingle	Technology T Telecommunication TK5101-6720 Shuangjiang Du Baofu Zhang Pin Zhang Peng Xiang Hong Xue FA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background
description	Infrared target detection is a popular applied field in object detection as well as a challenge. This paper proposes the focus and attention mechanism-based YOLO (FA-YOLO), which is an improved method to detect the infrared occluded vehicles in the complex background of remote sensing images. Firstly, we use GAN to create infrared images from the visible datasets to make sufficient datasets for training as well as using transfer learning. Then, to mitigate the impact of the useless and complex background information, we propose the negative sample focusing mechanism to focus on the confusing negative sample training to depress the false positives and increase the detection precision. Finally, to enhance the features of the infrared small targets, we add the dilated convolutional block attention module (dilated CBAM) to the CSPdarknet53 in the YOLOv4 backbone. To verify the superiority of our model, we carefully select 318 infrared occluded vehicle images from the VIVID-infrared dataset for testing. The detection accuracy-mAP improves from 79.24% to 92.95%, and the F1 score improves from 77.92% to 88.13%, which demonstrates a significant improvement in infrared small occluded vehicle detection.
format	article
author	Shuangjiang Du Baofu Zhang Pin Zhang Peng Xiang Hong Xue
author_facet	Shuangjiang Du Baofu Zhang Pin Zhang Peng Xiang Hong Xue
author_sort	Shuangjiang Du
title	FA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background
title_short	FA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background
title_full	FA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background
title_fullStr	FA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background
title_full_unstemmed	FA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background
title_sort	fa-yolo: an improved yolo model for infrared occlusion object detection under confusing background
publisher	Hindawi-Wiley
publishDate	2021
url	https://doaj.org/article/7122b0d4c9294739a9061d50032458ba
work_keys_str_mv	AT shuangjiangdu fayoloanimprovedyolomodelforinfraredocclusionobjectdetectionunderconfusingbackground AT baofuzhang fayoloanimprovedyolomodelforinfraredocclusionobjectdetectionunderconfusingbackground AT pinzhang fayoloanimprovedyolomodelforinfraredocclusionobjectdetectionunderconfusingbackground AT pengxiang fayoloanimprovedyolomodelforinfraredocclusionobjectdetectionunderconfusingbackground AT hongxue fayoloanimprovedyolomodelforinfraredocclusionobjectdetectionunderconfusingbackground
_version_	1718407714325921792

FA-YOLO: An Improved YOLO Model for Infrared Occlusion Object Detection under Confusing Background

Ejemplares similares