DLFE-YOLO: an enhanced framework for underwater object detection based on YOLOv8

Abstract

Underwater object detection remains challenging because of low visibility, complex lighting, and varied target scales, which limit the effectiveness of conventional deep learning methods. To overcome these limitations, this study proposes DLFE-You Only Look Once (YOLO), a novel YOLOv8-based framework that enhances feature representation and detection robustness in underwater environments. Its key innovations include its improved attention mechanisms and adaptive feature modeling to better capture irregular and small targets, as well as a loss function that emphasizes difficult-to-detect objects. Extensive experiments on two public benchmarks (URPC2020 and URPC2018) demonstrate that the proposed framework consistently outperforms state-of-the-art detectors, achieving mAP@0.5 values of 84.4% and 77.9%, respectively. Transfer learning experiments further indicate that DLFE-YOLO maintains strong adaptability across datasets, effectively leveraging knowledge from larger datasets to improve detection in smaller or less diverse underwater domains. These results highlight the model’s ability to address the shortcomings of existing methods. It improves both detection accuracy and generalization and provides a practical solution for intelligent underwater perception systems.

Keywords

Underwater object detection / ResNet / YOLO / Loss function

References

[1] Alam M, Khan IR, Siddiqui F, Alam MA (2024) Artificial intelligence as key enabler for safeguarding the marine resources. In: De D et al (eds) Artificial intelligence and edge computing for sustainable ocean health. Springer, Cham, pp 409–451. https://doi.org/10.1007/978-3-031-64642-3_18

[2] Ali ML, Zhang Z. The YOLO framework: a comprehensive review of evolution, applications, and benchmarks in object detection. Computers, 2024, 13(12 336

[3] Bajpai V, Sharma A, Subudhi BN, Veerakumar T, Jakhetiya V (2021) Underwater U-Net: deep learning with U-Net for visual underwater moving object detection. In: OCEANS 2021: San Diego–Porto. IEEE, pp 1–4. https://doi.org/10.23919/OCEANS44145.2021.9705761

[4] Cai SB, Zhang XK, Mo YC. A lightweight underwater detector enhanced by attention mechanism, GSConv and WIoU on YOLOv8. Sci Rep, 2024, 14(1 25797

[5] Cui YN, Knoll A. Dual-domain strip attention for image restoration. Neural Netw, 2024, 171: 429-439

[6] Deluxni N, Sudhakaran P, Kitmo NMF. A review on image enhancement and restoration techniques for underwater optical imaging applications. IEEE Access, 2023, 11: 111715-111737

[7] Du LX, Zhang RY, Wang XT. Overview of two-stage object detection algorithms. J Phys Conf Ser, 2020, 1544(1 012033

[8] Elmezain M, Saad Saoud L, Sultan A, Heshmat M, Seneviratne L, Hussain I. Advancing underwater vision: a survey of deep learning models for underwater object recognition and tracking. IEEE Access, 2025, 13: 17830-17867

[9] Er MJ, Chen J, Zhang YN, Gao WX. Research challenges, recent advances, and popular datasets in deep learning-based underwater marine object detection: a review. Sensors, 2023, 23(4 1990

[10] Fang C, Yang X. Lightweight yolov8 for wheat head detection. IEEE Access, 2024, 12: 66214-66222

[11] Fu CP, Liu RS, Fan X, Chen PY, Fu H, Yuan WQ, et al.. Rethinking general underwater object detection: datasets, challenges, and solutions. Neurocomputing, 2023, 517: 243-256

[12] Fu JS, Tian Y. Improved YOLOv7 underwater object detection based on attention mechanism. Eng Lett, 2024, 32(7): 1377-1385

[13] Gaines SD, Lester SE, Grorud-Colvert K, Costello C, Pollnac R. Evolving science of marine reserves: new developments and emerging research frontiers. Proc Natl Acad Sci USA, 2010, 107(43): 18251-18255

[14] He KM, Zhang XY, Ren SQ, Sun J (2016) Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, pp 770–778. https://doi.org/10.1109/CVPR.2016.90

[15] Hua X, Cui XP, Xu XH, Qiu SH, Liang YJ, Bao XQ, et al.. Underwater object detection algorithm based on feature enhancement and progressive dynamic aggregation strategy. Pattern Recogn, 2023, 139 109511

[16] Jia JQ, Fu M, Liu XF, Zheng B. Underwater object detection based on improved EfficientDet. Remote Sens, 2022, 14(18 4487

[17] Li H, Wu AY, Jiang ZY, Liu F, Luo MY (2024) Improving object detection in YOLOv8n with the C2f-f module and multi-scale fusion reconstruction. In: 2024 IEEE 6th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC). IEEE, pp 374–379. https://doi.org/10.1109/IMCEC59810.2024.10575292

[18] Li Q, Li JH, Li T, Li ZY, Zhang P. Spectral-spatial depth-based framework for hyperspectral underwater target detection. IEEE Trans Geosci Remote Sens, 2023, 61: 1-15

[19] Li ZX, He QH, Yang WY. E-FPN: an enhanced feature pyramid network for UAV scenarios detection. Vis Comput, 2025, 41(1): 675-693

[20] Liang XT, Song PH (2022) Excavating RoI attention for underwater object detection. In: 2022 IEEE International Conference on Image Processing (ICIP). IEEE, pp 2651–2655. https://doi.org/10.1109/ICIP46576.2022.9897515

[21] Liu FH, Fei YH, Su M, Zang R, Sun XF, Wang GC (2024a) A novel lightweight model for underwater small object detection. In: 2024 China Automation Congress (CAC). IEEE, pp 2226–2231. https://doi.org/10.1109/CAC63892.2024.10865529

[22] Liu HK, Luo XN, Li F (2024b) YOLOv5 garbage target detection based on lightweight convolution in marine environment. In: 2024 International Symposium on Digital Home (ISDH). IEEE, pp 31–36. https://doi.org/10.1109/ISDH64927.2024.00012

[23] Liu JJ, Liu ZT, Wang YP, Li F (2024c) VVNet: underwater object detection network based on vision Transformer and vision retnet for underwater robot picking. In: 2024 International Symposium on Digital Home (ISDH). IEEE, pp 19–24. https://doi.org/10.1109/ISDH64927.2024.00010

[24] Liu K, Peng L, Tang SR. Underwater object detection using TC-YOLO with attention mechanisms. Sensors, 2023, 23(5 2567

[25] Liu XP, Zhao KK, Liu C, Chen L. Bi2f-yolo: a novel framework for underwater object detection based on YOLOv7. Intell Mar Technol Syst, 2025, 3 9

[26] Luo Y, Wu AP, Fu QQ. MAS-YOLOv11: an improved underwater object detection algorithm based on YOLOv11. Sensors (Basel), 2025, 25(11 3433

[27] Redmon J, Divvala S, Girshick R, Farhadi A (2016) You only look once: unified, real-time object detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, pp 779–788. https://doi.org/10.1109/CVPR.2016.91

[28] Sahoo SK, Choudhury BB, Dhal PR. Exploring the role of robotics in maritime technology: innovations, challenges, and future prospects. Spectr Mech Eng Oper Res, 2024, 1(1): 159-176

[29] Shen X, Wang HB, Cui TX, Guo ZC, Fu XP. Multiple information perception-based attention in YOLO for underwater object detection. Vis Comput, 2024, 40(3): 1415-1438

[30] Varghese R, Sambath M (2023) A comprehensive review on two-stage object detection algorithms. In: 2023 International Conference on Quantum Technologies, Communications, Computing, Hardware and Embedded Systems Security (iQ-CCHESS). IEEE, pp 1–7. https://doi.org/10.1109/iQ-CCHESS56596.2023.10391506

[31] Verma S, Sharma DM. Role of digitalization in a sustainable blue economy. Int J Teach Learn High Educ, 2025, 4(3): 30-40

[32] Wang AJ, Jiao HN, Chen ZC, Yang J. An improved YOLO-based waste detection model and its integration to robotic gripping systems. Comput Mater Contin, 2025, 84(3): 5773-5790

[33] Wang H, Xiao NF. Underwater object detection method based on improved faster RCNN. Appl Sci, 2023, 13(4 2746

[34] Wang XY, Xue G, Huang ST, Liu YJ. Underwater object detection algorithm based on adding channel and spatial fusion attention mechanism. J Mar Sci Eng, 2023, 11(6 1116

[35] Wu YH, Duan YH, Wei YG, An D, Liu JC. Application of intelligent and unmanned equipment in aquaculture: a review. Comput Electron Agric, 2022, 199 107201

[36] Wynn RB, Huvenne VA, Le Bas TP, Murton BJ, Connelly DP, Bett BJ, et al.. Autonomous underwater vehicles (AUVs): their past, present and future contributions to the advancement of marine geoscience. Mar Geol, 2014, 352: 451-468

[37] Yaseen M (2024) What is YOLOv8: an in-depth exploration of the internal features of the next-generation object detector. Preprint at arXiv:2408.15857.

[38] Yu JH, Jiang YN, Wang ZY, Cao ZM, Huang T (2016) UnitBox: an advanced object detection network. In: Proceedings of the 24th ACM International Conference on Multimedia. ACM, pp 516–520. https://doi.org/10.1145/2964284.2967274

[39] Zhang CPF, Zhang GY, Li H, Liu H, Tan J, Xue XJ. Underwater target detection algorithm based on improved YOLOv4 with SemiDSConv and FIoU loss function. Front Mar Sci, 2023, 10 1153416

[40] Zhang JQ, Lei J, Xie WY, Fang ZM, Li YS, Du Q. SuperYOLO: super resolution assisted object detection in multimodal remote sensing imagery. IEEE Trans Geosci Remote Sens, 2023, 61: 1-15

[41] Zhang X, Song YZ, Song TT, Yang DG, Ye YC, Zhou J, et al.. LDConv: linear deformable convolution for improving convolutional neural networks. Image Vis Comput, 2024, 149 105190

[42] Zhang YF, Ren WQ, Zhang Z, Jia Z, Wang L, Tan TN. Focal and efficient iou loss for accurate bounding box regression. Neurocomputing, 2022, 506: 146-157

[43] Zhang ZH, Chen P, Zhang SQ, Huang HQ, Pan YL, Pan DL. A review of machine learning applications in ocean color remote sensing. Remote Sens, 2025, 17(10 1776

[44] Zhao D, Yang B, Dou YK, Guo XJ (2022) Underwater fish detection in sonar image based on an improved faster RCNN. In: 2022 9th International Forum on Electrical Engineering and Automation (IFEEA). IEEE, pp 358–363. https://doi.org/10.1109/IFEEA57288.2022.10038226

[45] Zhao L, Ren X, Fu LL, Yun Q, Yang JR. UWS-YOLO: advancing underwater sonar object detection via transfer learning and orthogonal-snake convolution mechanisms. J Mar Sci Eng, 2025, 13(10 1847

[46] Zheng ZH, Wang P, Liu W, Li JZ, Ye RG, Ren DW. Distance-iou loss: faster and better learning for bounding box regression. Proceedings of the AAAI Conference on Artificial Intelligence, 2020

[47] Zhou H, Kong MW, Yuan HX, Pan YY, Wang XR, Chen R, et al.. Real-time underwater object detection technology for complex underwater environments based on deep learning. Ecol Inform, 2024, 82 102680

[48] Zhou JJ, Xu DM, Min XY, Wu D (2024b) An improved underwater target detection algorithm based on YOLOX. In: OCEANS 2024–Singapore. IEEE, pp 1–7. https://doi.org/10.1109/OCEANS51537.2024.10682303

[49] Zhou XY, Chen WM, Wei XH. Improved field obstacle detection algorithm based on YOLOv8. Agriculture (Basel), 2024, 14(12 2263

[50] Zhu HG, Luo XN, Zhai ZY, Li J, Li F, Zhu XS (2024) UNet-based YOLO for underwater object detection. In: 2024 IEEE First International Conference on Data Intelligence and Innovative Application (DIIA). IEEE, pp 1–7. https://doi.org/10.1109/DIIA62678.2024.10871910

[51] Zou ZX, Chen KY, Shi ZW, Guo YH, Ye JP. Object detection in 20 years: a survey. Proc IEEE, 2023, 111(3): 257-276

RIGHTS & PERMISSIONS

The Author(s)