YOLOv5s-CA: A Modified YOLOv5s Network with Coordinate Attention for Underwater Target Detection
Underwater target detection techniques have been extensively applied to underwater vehicles for marine surveillance, aquaculture, and rescue applications. However, due to complex underwater environments and insufficient training samples, the existing underwater target recognition algorithm accuracy...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-03-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/23/7/3367 |
_version_ | 1797607080546271232 |
---|---|
author | Ge Wen Shaobao Li Fucai Liu Xiaoyuan Luo Meng-Joo Er Mufti Mahmud Tao Wu |
author_facet | Ge Wen Shaobao Li Fucai Liu Xiaoyuan Luo Meng-Joo Er Mufti Mahmud Tao Wu |
author_sort | Ge Wen |
collection | DOAJ |
description | Underwater target detection techniques have been extensively applied to underwater vehicles for marine surveillance, aquaculture, and rescue applications. However, due to complex underwater environments and insufficient training samples, the existing underwater target recognition algorithm accuracy is still unsatisfactory. A long-term effort is essential to improving underwater target detection accuracy. To achieve this goal, in this work, we propose a modified YOLOv5s network, called YOLOv5s-CA network, by embedding a Coordinate Attention (CA) module and a Squeeze-and-Excitation (SE) module, aiming to concentrate more computing power on the target to improve detection accuracy. Based on the existing YOLOv5s network, the number of bottlenecks in the first C3 module was increased from one to three to improve the performance of shallow feature extraction. The CA module was embedded into the C3 modules to improve the attention power focused on the target. The SE layer was added to the output of the C3 modules to strengthen model attention. Experiments on the data of the 2019 China Underwater Robot Competition were conducted, and the results demonstrate that the mean Average Precision (mAP) of the modified YOLOv5s network was increased by 2.4%. |
first_indexed | 2024-03-11T05:26:15Z |
format | Article |
id | doaj.art-dd9e18b6ab244bb9bf038bcf6e92fc76 |
institution | Directory Open Access Journal |
issn | 1424-8220 |
language | English |
last_indexed | 2024-03-11T05:26:15Z |
publishDate | 2023-03-01 |
publisher | MDPI AG |
record_format | Article |
series | Sensors |
spelling | doaj.art-dd9e18b6ab244bb9bf038bcf6e92fc762023-11-17T17:31:28ZengMDPI AGSensors1424-82202023-03-01237336710.3390/s23073367YOLOv5s-CA: A Modified YOLOv5s Network with Coordinate Attention for Underwater Target DetectionGe Wen0Shaobao Li1Fucai Liu2Xiaoyuan Luo3Meng-Joo Er4Mufti Mahmud5Tao Wu6School of Electrical Engineering, Yanshan University, Qinhuangdao 066004, ChinaSchool of Electrical Engineering, Yanshan University, Qinhuangdao 066004, ChinaSchool of Electrical Engineering, Yanshan University, Qinhuangdao 066004, ChinaSchool of Electrical Engineering, Yanshan University, Qinhuangdao 066004, ChinaInstitute of Artificial Intelligence and Marine Robotics, College of Marine Electrical Engineering, Dalian Maritime University, Dalian 116026, ChinaDepartment of Computer Science, Computing and Informatics Research Centre, Medical Technologies Innovation Facility of Nottingham Trent University, Nottingham NG11 8NS, UKDepartment of Frontier & Innovation Research, Wuhan Second Ship Design & Research Institute, Wuhan 430205, ChinaUnderwater target detection techniques have been extensively applied to underwater vehicles for marine surveillance, aquaculture, and rescue applications. However, due to complex underwater environments and insufficient training samples, the existing underwater target recognition algorithm accuracy is still unsatisfactory. A long-term effort is essential to improving underwater target detection accuracy. To achieve this goal, in this work, we propose a modified YOLOv5s network, called YOLOv5s-CA network, by embedding a Coordinate Attention (CA) module and a Squeeze-and-Excitation (SE) module, aiming to concentrate more computing power on the target to improve detection accuracy. Based on the existing YOLOv5s network, the number of bottlenecks in the first C3 module was increased from one to three to improve the performance of shallow feature extraction. The CA module was embedded into the C3 modules to improve the attention power focused on the target. The SE layer was added to the output of the C3 modules to strengthen model attention. Experiments on the data of the 2019 China Underwater Robot Competition were conducted, and the results demonstrate that the mean Average Precision (mAP) of the modified YOLOv5s network was increased by 2.4%.https://www.mdpi.com/1424-8220/23/7/3367underwater target detectiondeep learningYOLO neural networkCoordinate Attention |
spellingShingle | Ge Wen Shaobao Li Fucai Liu Xiaoyuan Luo Meng-Joo Er Mufti Mahmud Tao Wu YOLOv5s-CA: A Modified YOLOv5s Network with Coordinate Attention for Underwater Target Detection Sensors underwater target detection deep learning YOLO neural network Coordinate Attention |
title | YOLOv5s-CA: A Modified YOLOv5s Network with Coordinate Attention for Underwater Target Detection |
title_full | YOLOv5s-CA: A Modified YOLOv5s Network with Coordinate Attention for Underwater Target Detection |
title_fullStr | YOLOv5s-CA: A Modified YOLOv5s Network with Coordinate Attention for Underwater Target Detection |
title_full_unstemmed | YOLOv5s-CA: A Modified YOLOv5s Network with Coordinate Attention for Underwater Target Detection |
title_short | YOLOv5s-CA: A Modified YOLOv5s Network with Coordinate Attention for Underwater Target Detection |
title_sort | yolov5s ca a modified yolov5s network with coordinate attention for underwater target detection |
topic | underwater target detection deep learning YOLO neural network Coordinate Attention |
url | https://www.mdpi.com/1424-8220/23/7/3367 |
work_keys_str_mv | AT gewen yolov5scaamodifiedyolov5snetworkwithcoordinateattentionforunderwatertargetdetection AT shaobaoli yolov5scaamodifiedyolov5snetworkwithcoordinateattentionforunderwatertargetdetection AT fucailiu yolov5scaamodifiedyolov5snetworkwithcoordinateattentionforunderwatertargetdetection AT xiaoyuanluo yolov5scaamodifiedyolov5snetworkwithcoordinateattentionforunderwatertargetdetection AT mengjooer yolov5scaamodifiedyolov5snetworkwithcoordinateattentionforunderwatertargetdetection AT muftimahmud yolov5scaamodifiedyolov5snetworkwithcoordinateattentionforunderwatertargetdetection AT taowu yolov5scaamodifiedyolov5snetworkwithcoordinateattentionforunderwatertargetdetection |