BSE-YOLO: An Enhanced Lightweight Multi-Scale Underwater Object Detection Model
Underwater images often exhibit characteristics such as low contrast, blurred and small targets, object clustering, and considerable variations in object morphology. Traditional detection methods tend to be susceptible to omission and false positives under these circumstances. Furthermore, owing to...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/13/3890 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850116285388554240 |
|---|---|
| author | Yuhang Wang Hua Ye Xin Shu |
| author_facet | Yuhang Wang Hua Ye Xin Shu |
| author_sort | Yuhang Wang |
| collection | DOAJ |
| description | Underwater images often exhibit characteristics such as low contrast, blurred and small targets, object clustering, and considerable variations in object morphology. Traditional detection methods tend to be susceptible to omission and false positives under these circumstances. Furthermore, owing to the constrained memory and limited computing power of underwater robots, there is a significant demand for lightweight models in underwater object detection tasks. Therefore, we propose an enhanced lightweight YOLOv10n-based model, BSE-YOLO. Firstly, we replace the original neck with an improved Bidirectional Feature Pyramid Network (Bi-FPN) to reduce parameters. Secondly, we propose a Multi-Scale Attention Synergy Module (MASM) to enhance the model’s perception of difficult features and make it focus on the important regions. Finally, we integrate Efficient Multi-Scale Attention (EMA) into the backbone and neck to improve feature extraction and fusion. The experiment results demonstrate that the proposed BSE-YOLO reaches 83.7% <i>mAP</i>@0.5 on URPC2020 and 83.9% <i>mAP</i>@0.5 on DUO, with the parameters reducing 2.47 M. Compared to the baseline model YOLOv10n, our BSE-YOLO improves <i>mAP</i>@0.5 by 2.2% and 3.0%, respectively, while reducing the number of parameters by approximately 0.2 M. The BSE-YOLO achieves a good balance between accuracy and lightweight, providing an effective solution for underwater object detection. |
| format | Article |
| id | doaj-art-0e537e6cc3974ef288e69a11245dfb4f |
| institution | OA Journals |
| issn | 1424-8220 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-0e537e6cc3974ef288e69a11245dfb4f2025-08-20T02:36:22ZengMDPI AGSensors1424-82202025-06-012513389010.3390/s25133890BSE-YOLO: An Enhanced Lightweight Multi-Scale Underwater Object Detection ModelYuhang Wang0Hua Ye1Xin Shu2School of Computer Science, Jiangsu University of Science and Technology, Zhenjiang 212100, ChinaSchool of Computer Science, Jiangsu University of Science and Technology, Zhenjiang 212100, ChinaSchool of Computer Science, Jiangsu University of Science and Technology, Zhenjiang 212100, ChinaUnderwater images often exhibit characteristics such as low contrast, blurred and small targets, object clustering, and considerable variations in object morphology. Traditional detection methods tend to be susceptible to omission and false positives under these circumstances. Furthermore, owing to the constrained memory and limited computing power of underwater robots, there is a significant demand for lightweight models in underwater object detection tasks. Therefore, we propose an enhanced lightweight YOLOv10n-based model, BSE-YOLO. Firstly, we replace the original neck with an improved Bidirectional Feature Pyramid Network (Bi-FPN) to reduce parameters. Secondly, we propose a Multi-Scale Attention Synergy Module (MASM) to enhance the model’s perception of difficult features and make it focus on the important regions. Finally, we integrate Efficient Multi-Scale Attention (EMA) into the backbone and neck to improve feature extraction and fusion. The experiment results demonstrate that the proposed BSE-YOLO reaches 83.7% <i>mAP</i>@0.5 on URPC2020 and 83.9% <i>mAP</i>@0.5 on DUO, with the parameters reducing 2.47 M. Compared to the baseline model YOLOv10n, our BSE-YOLO improves <i>mAP</i>@0.5 by 2.2% and 3.0%, respectively, while reducing the number of parameters by approximately 0.2 M. The BSE-YOLO achieves a good balance between accuracy and lightweight, providing an effective solution for underwater object detection.https://www.mdpi.com/1424-8220/25/13/3890underwater object detectionYOLOv10nbidirectional feature pyramid networkmulti-scale attention synergylightweight network |
| spellingShingle | Yuhang Wang Hua Ye Xin Shu BSE-YOLO: An Enhanced Lightweight Multi-Scale Underwater Object Detection Model Sensors underwater object detection YOLOv10n bidirectional feature pyramid network multi-scale attention synergy lightweight network |
| title | BSE-YOLO: An Enhanced Lightweight Multi-Scale Underwater Object Detection Model |
| title_full | BSE-YOLO: An Enhanced Lightweight Multi-Scale Underwater Object Detection Model |
| title_fullStr | BSE-YOLO: An Enhanced Lightweight Multi-Scale Underwater Object Detection Model |
| title_full_unstemmed | BSE-YOLO: An Enhanced Lightweight Multi-Scale Underwater Object Detection Model |
| title_short | BSE-YOLO: An Enhanced Lightweight Multi-Scale Underwater Object Detection Model |
| title_sort | bse yolo an enhanced lightweight multi scale underwater object detection model |
| topic | underwater object detection YOLOv10n bidirectional feature pyramid network multi-scale attention synergy lightweight network |
| url | https://www.mdpi.com/1424-8220/25/13/3890 |
| work_keys_str_mv | AT yuhangwang bseyoloanenhancedlightweightmultiscaleunderwaterobjectdetectionmodel AT huaye bseyoloanenhancedlightweightmultiscaleunderwaterobjectdetectionmodel AT xinshu bseyoloanenhancedlightweightmultiscaleunderwaterobjectdetectionmodel |