MSS-Net: a lightweight network incorporating shifted large kernel and multi-path attention for ship detection in remote sensing images

For the challenges with nearshore small ship detection in remote sensing images (RSIs) under complex background, a lightweight network called “multi-path attention and shifted large kernel network for ship detection” (briefly called “MSS-Net”) in RSIs is proposed. Firstly, shifted large kernel with...

Full description

Saved in:
Bibliographic Details
Main Authors: Guoqing Zhou, Xiangting Wang, Sheng Liu, Yuefeng Wang, Ertao Gao, Jiangying Wu, Yanling Lu, Linbo Yu, Weiyi Wang, Kun Li
Format: Article
Language:English
Published: Elsevier 2025-09-01
Series:International Journal of Applied Earth Observations and Geoinformation
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1569843225004522
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849222560071286784
author Guoqing Zhou
Xiangting Wang
Sheng Liu
Yuefeng Wang
Ertao Gao
Jiangying Wu
Yanling Lu
Linbo Yu
Weiyi Wang
Kun Li
author_facet Guoqing Zhou
Xiangting Wang
Sheng Liu
Yuefeng Wang
Ertao Gao
Jiangying Wu
Yanling Lu
Linbo Yu
Weiyi Wang
Kun Li
author_sort Guoqing Zhou
collection DOAJ
description For the challenges with nearshore small ship detection in remote sensing images (RSIs) under complex background, a lightweight network called “multi-path attention and shifted large kernel network for ship detection” (briefly called “MSS-Net”) in RSIs is proposed. Firstly, shifted large kernel with feature enhancement module (SLKE) is developed to enlarge receptive field by decomposing large kernels and shift operation while performing dynamic channel attention. Secondly, multi-path attention (MPA) is designed to effectively retain the co-calibration of spatial-channel information of ships. Thirdly, shared convolutional detection head (SCDH) is built to unify multi-scale features, reducing parameter redundancy. The proposed MSS-Net is validated through three public datasets, TGRS-HRRSD, MASATI and LEVIR. Using YOLOv8 as a baseline model for comparison analysis. The results demonstrate that the mAP50 reaches 97.5%, 78.8%, and 93.2% with the three datasets, respectively. The mAP50 with the proposed MSS-Net is higher 2.9% than YOLOX, 4.4% than RetinaNet in popular one-stage ship detection models, and 4.8% than Faster R-CNN; 2.7% than Cascade R-CNN in two-stage ship detection models. Moreover, the parameters in the MSS-Net reduces 26.7% relative to the baseline model, achieving a lightweight design. Besides, ablation experiments are conducted with the TGRS-HRRSD dataset. The results demonstrates that the SLKE increases mAP50 by 1.1%, the MPA increases mAP50 by 1.7%, while the SCDH reduces parameters by 35%. These results demonstrate that the MSS-Net achieves notable advances for lightweight ship detection.
format Article
id doaj-art-489b344766ed4402ac7fdb3fcc07e7da
institution Kabale University
issn 1569-8432
language English
publishDate 2025-09-01
publisher Elsevier
record_format Article
series International Journal of Applied Earth Observations and Geoinformation
spelling doaj-art-489b344766ed4402ac7fdb3fcc07e7da2025-08-26T04:14:10ZengElsevierInternational Journal of Applied Earth Observations and Geoinformation1569-84322025-09-0114310480510.1016/j.jag.2025.104805MSS-Net: a lightweight network incorporating shifted large kernel and multi-path attention for ship detection in remote sensing imagesGuoqing Zhou0Xiangting Wang1Sheng Liu2Yuefeng Wang3Ertao Gao4Jiangying Wu5Yanling Lu6Linbo Yu7Weiyi Wang8Kun Li9Corresponding author.; Guilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaGuilin University of Technology, Guilin 541004, ChinaFor the challenges with nearshore small ship detection in remote sensing images (RSIs) under complex background, a lightweight network called “multi-path attention and shifted large kernel network for ship detection” (briefly called “MSS-Net”) in RSIs is proposed. Firstly, shifted large kernel with feature enhancement module (SLKE) is developed to enlarge receptive field by decomposing large kernels and shift operation while performing dynamic channel attention. Secondly, multi-path attention (MPA) is designed to effectively retain the co-calibration of spatial-channel information of ships. Thirdly, shared convolutional detection head (SCDH) is built to unify multi-scale features, reducing parameter redundancy. The proposed MSS-Net is validated through three public datasets, TGRS-HRRSD, MASATI and LEVIR. Using YOLOv8 as a baseline model for comparison analysis. The results demonstrate that the mAP50 reaches 97.5%, 78.8%, and 93.2% with the three datasets, respectively. The mAP50 with the proposed MSS-Net is higher 2.9% than YOLOX, 4.4% than RetinaNet in popular one-stage ship detection models, and 4.8% than Faster R-CNN; 2.7% than Cascade R-CNN in two-stage ship detection models. Moreover, the parameters in the MSS-Net reduces 26.7% relative to the baseline model, achieving a lightweight design. Besides, ablation experiments are conducted with the TGRS-HRRSD dataset. The results demonstrates that the SLKE increases mAP50 by 1.1%, the MPA increases mAP50 by 1.7%, while the SCDH reduces parameters by 35%. These results demonstrate that the MSS-Net achieves notable advances for lightweight ship detection.http://www.sciencedirect.com/science/article/pii/S1569843225004522Remote SensingDeep learningShip detectionLarge kernelAttention mechanismLightweight network
spellingShingle Guoqing Zhou
Xiangting Wang
Sheng Liu
Yuefeng Wang
Ertao Gao
Jiangying Wu
Yanling Lu
Linbo Yu
Weiyi Wang
Kun Li
MSS-Net: a lightweight network incorporating shifted large kernel and multi-path attention for ship detection in remote sensing images
International Journal of Applied Earth Observations and Geoinformation
Remote Sensing
Deep learning
Ship detection
Large kernel
Attention mechanism
Lightweight network
title MSS-Net: a lightweight network incorporating shifted large kernel and multi-path attention for ship detection in remote sensing images
title_full MSS-Net: a lightweight network incorporating shifted large kernel and multi-path attention for ship detection in remote sensing images
title_fullStr MSS-Net: a lightweight network incorporating shifted large kernel and multi-path attention for ship detection in remote sensing images
title_full_unstemmed MSS-Net: a lightweight network incorporating shifted large kernel and multi-path attention for ship detection in remote sensing images
title_short MSS-Net: a lightweight network incorporating shifted large kernel and multi-path attention for ship detection in remote sensing images
title_sort mss net a lightweight network incorporating shifted large kernel and multi path attention for ship detection in remote sensing images
topic Remote Sensing
Deep learning
Ship detection
Large kernel
Attention mechanism
Lightweight network
url http://www.sciencedirect.com/science/article/pii/S1569843225004522
work_keys_str_mv AT guoqingzhou mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT xiangtingwang mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT shengliu mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT yuefengwang mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT ertaogao mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT jiangyingwu mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT yanlinglu mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT linboyu mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT weiyiwang mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages
AT kunli mssnetalightweightnetworkincorporatingshiftedlargekernelandmultipathattentionforshipdetectioninremotesensingimages