A Fast and Efficient Ship Detector Via HDR-CoT Module

Jane Litouw; Marsel Marhaen Wungow; Dayen Manoppo; Ni Made S. Mustikayani; Hoang Van Thanh; Muhamad Dwisnanto Putro

doi:10.30871/jaic.v10i5.13947

Authors

Jane Litouw Department of Electrical Engineering, Faculty of Engineering, Sam Ratulangi University
Marsel Marhaen Wungow Department of Electrical Engineering, Faculty of Engineering, Sam Ratulangi University
Dayen Manoppo Department of Electrical Engineering, Faculty of Engineering, Sam Ratulangi University
Ni Made S. Mustikayani Department of Electrical Engineering, Faculty of Engineering, Sam Ratulangi University
Hoang Van Thanh Faculty of Information Technology, Quang Binh University, Vietnam
Muhamad Dwisnanto Putro Sam Ratulangi University

DOI:

https://doi.org/10.30871/jaic.v10i5.13947

Keywords:

Ship Detection, Modified YOLO-FAST, Efficient Model, Deep Learning, Contextual Transformer

Abstract

Reliable vessel surveillance requires robust computer vision for accurate ship detection. Although YOLO-FAST is computationally efficient, its lightweight architecture has limited capability to extract discriminative features from small objects. To address this limitation, this study proposes an enhanced ship detection framework by integrating a novel Hadamard-Dilated Residual Contextual Transformer (HDR-CoT) module into the YOLO-FAST architecture. The proposed HDR-CoT module integrates Hadamard multiplication with parallel dilated depthwise convolutions. In addition, residual connections enhance feature representation while maintaining low computational overhead. Experimental results demonstrate the effectiveness of the proposed model. It achieves an mAP50 of 0.952 and an mAP50:95 of 0.689 while requiring only 1.73 million parameters and 6.3 GFLOPs. The proposed network achieves a favorable balance between detection performance and efficiency, improving mAP50:95 by 2.99% while reducing the parameters by 24.93%. It also achieves competitive performance compared with other lightweight architectures. CPU-based evaluation further shows that the model operates at 20.54 FPS. This configuration achieves an optimal trade-off between near-real-time inference speed and detection reliability compared with YOLO-FAST for resource-constrained maritime applications.

Downloads

Download data is not yet available.

References

[1] L. Qian, Y. Zheng, L. Li, Y. Ma, C. Zhou, and D. Zhang, “A New Method of Inland Water Ship Trajectory Prediction Based on Long Short-Term Memory Network Optimized by Genetic Algorithm,” Applied Sciences (Switzerland), vol. 12, no. 8, Apr. 2022, doi: 10.3390/app12084073.

[2] Y. Mo, C. Li, W. Ren, and W. Wang, “Spotlighting on Objects: Prior Knowledge-Driven Maritime Image Dehazing and Object Detection Framework,” IEEE Journal of Oceanic Engineering, vol. 50, no. 3, pp. 1978–1992, 2025, doi: 10.1109/JOE.2025.3545289.

[3] A. Miri Rekavandi, L. Xu, F. Boussaid, A.-K. Seghouane, S. Hoefs, and M. Bennamoun, “A Guide to Image- and Video-Based Small Object Detection Using Deep Learning: Case Study of Maritime Surveillance,” IEEE Transactions on Intelligent Transportation Systems, vol. 26, no. 3, pp. 2851–2879, 2025, doi: 10.1109/TITS.2025.3530678.

[4] V. Sanikommu, S. P. Marripudi, H. R. Yekkanti, R. Divi, R. Chandrakanth, and P. Mahindra, “Edge computing for detection of ship and ship port from remote sensing images using YOLO,” Front. Artif. Intell., vol. 8, 2025, doi: 10.3389/frai.2025.1508664.

[5] M. Zhu, D. Han, B. Han, and X. Huang, “YOLO-HPSD: A high-precision ship target detection model based on YOLOv10,” PLoS One, vol. 20, no. 5 May, May 2025, doi: 10.1371/journal.pone.0321863.

[6] L. Shen, T. Gao, and Q. Yin, “YOLO-LPSS: A Lightweight and Precise Detection Model for Small Sea Ships,” J. Mar. Sci. Eng., vol. 13, no. 5, May 2025, doi: 10.3390/jmse13050925.

[7] Y. Li and S. Wang, “EGM-YOLOv8: A Lightweight Ship Detection Model with Efficient Feature Fusion and Attention Mechanisms,” J. Mar. Sci. Eng., vol. 13, no. 4, Apr. 2025, doi: 10.3390/jmse13040757.

[8] L. Min, F. Dou, Y. Zhang, D. Shao, L. Li, and B. Wang, “CM-YOLO: Context Modulated Representation Learning for Ship Detection,” IEEE Transactions on Geoscience and Remote Sensing, vol. 63, pp. 1–14, 2025, doi: 10.1109/TGRS.2025.3538848.

[9] C. Qu et al., “DSONet: A Lightweight Framework for the Detection of Multiscale Dense Ship Occlusion,” IEEE Internet Things J., vol. 12, no. 20, pp. 42638–42651, 2025, doi: 10.1109/JIOT.2025.3593829.

[10] W. Messaoud, R. Trabelsi, A. Cabani, and F. Abdelkefi, “Adaptive Head Pruning for Attention Mechanism in the Maritime Domain,” IEEE Transactions on Artificial Intelligence, vol. 6, no. 11, pp. 2966–2976, 2025, doi: 10.1109/TAI.2025.3558724.

[11] W. Huang et al., “DAShip: A Large-Scale Annotated Dataset for Ship Detection Using Distributed Acoustic Sensing Technique,” IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., vol. 18, pp. 4093–4107, 2025, doi: 10.1109/JSTARS.2024.3525082.

[12] Y. Wang, S. Zhang, J. Xu, Z. Cheng, and G. Du, “YOLO-StarLS: A Ship Detection Algorithm Based on Wavelet Transform and Multi-Scale Feature Extraction for Complex Environments,” Symmetry (Basel)., vol. 17, no. 7, Jul. 2025, doi: 10.3390/sym17071116.

[13] W. Wu, X. Fan, Z. Hu, and Y. Zhao, “CGDU-DETR: An End-to-End Detection Model for Ship Detection in Day–Night Transition Environments,” J. Mar. Sci. Eng., vol. 13, no. 6, 2025, doi: 10.3390/jmse13061155.

[14] Y. Liu et al., “YOLOv7oSAR: A Lightweight High-Precision Ship Detection Model for SAR Images Based on the YOLOv7 Algorithm,” Remote Sens. (Basel)., vol. 16, no. 5, Mar. 2024, doi: 10.3390/rs16050913.

[15] R. D. Kwon, G. R. Nam, J. Tak, J. Shin, H. Cha, and S. W. Lee, “Enhancing Ship Classification in Optical Satellite Imagery: Integrating Convolutional Block Attention Module with ResNet for Improved Performance,” Aug. 2024, [Online]. Available: http://arxiv.org/abs/2404.02135

[16] Z. Jiang, L. Su, and Y. Sun, “YOLOv7-Ship: A Lightweight Algorithm for Ship Object Detection in Complex Marine Environments,” J. Mar. Sci. Eng., vol. 12, no. 1, Jan. 2024, doi: 10.3390/jmse12010190.

[17] B. Erum Ayesha, T. Satyanarayana Murthy, P. Ramesh Babu, and R. Kuchipudi, “Ship Detection in Remote Sensing Imagery for Arbitrarily Oriented Object Detection.”

[18] N. Toprak and Y. Yalman, “Ship Detection from Optical Satellite Images Using Convolutional Neural Networks,” Turkish Journal of Engineering, vol. 9, no. 2, pp. 342–353, Jun. 2025, doi: 10.31127/tuje.1529660.

[19] S. Kang, Z. Hu, L. Liu, K. Zhang, and Z. Cao, “Object Detection YOLO Algorithms and Their Industrial Applications: Overview and Comparative Analysis,” Mar. 01, 2025, Multidisciplinary Digital Publishing Institute (MDPI). doi: 10.3390/electronics14061104.

[20] Q. Cao et al., “LH-YOLO: A Lightweight and High-Precision SAR Ship Detection Model Based on the Improved YOLOv8n,” Remote Sens. (Basel)., vol. 16, no. 22, Nov. 2024, doi: 10.3390/rs16224340.

[21] W. Shi, W. Zheng, and Z. Xu, “Ship-Yolo: A Deep Learning Approach for Ship Detection in Remote Sensing Images,” J. Mar. Sci. Eng., vol. 13, no. 4, Apr. 2025, doi: 10.3390/jmse13040737.

[22] J. Song, J. Xie, Q. Wang, and T. Shen, “An improved YOLO-based method with lightweight C3 modules for object detection in resource-constrained environments,” J. Supercomput., vol. 81, no. 5, p. 702, 2025, doi: 10.1007/s11227-025-07187-w.

[23] Z. Song, X. Zhang, and P. Tan, “YOLO-Fast: a lightweight object detection model for edge devices,” J. Supercomput., vol. 81, no. 5, p. 724, 2025, doi: 10.1007/s11227-025-07172-3.

[24] J. Chen et al., “Run, Don’t Walk: Chasing Higher FLOPS for Faster Neural Networks,” May 2023, [Online]. Available: http://arxiv.org/abs/2303.03667

[25] Y. Li, T. Yao, Y. Pan, and T. Mei, “Contextual Transformer Networks for Visual Recognition,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 45, no. 2, pp. 1489–1500, 2023, doi: 10.1109/TPAMI.2022.3164083.

[26] J. Chen et al., “Run, Don’t Walk: Chasing Higher FLOPS for Faster Neural Networks,” May 2023, [Online]. Available: http://arxiv.org/abs/2303.03667

[27] P. Zhao and S. Jia, “SPFFNet: Strip Perception and Feature Fusion Spatial Pyramid Pooling for Fabric Defect Detection,” in 2025 IEEE International Conference on Big Data (BigData), 2025, pp. 2701–2706. doi: 10.1109/BigData66926.2025.11402084.

[28] J. Hu, L. Shen, and G. Sun, “Squeeze-and-Excitation Networks,” in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 7132–7141. doi: 10.1109/CVPR.2018.00745.

[29] Z. Shao, W. Wu, Z. Wang, W. Du, and C. Li, “SeaShips: A Large-Scale Precisely Annotated Dataset for Ship Detection,” IEEE Trans. Multimedia, vol. 20, no. 10, pp. 2593–2604, Oct. 2018, doi: 10.1109/TMM.2018.2865686.

[30] X. Sun, H. Wu, G. Yu, and N. Zheng, “Research on the Automatic Detection of Ship Targets Based on an Improved YOLO v5 Algorithm and Model Optimization,” Mathematics, vol. 12, no. 11, 2024, doi: 10.3390/math12111714.

[31] K. Lan, X. Jiang, X. Ding, H. Lin, and S. Chan, “High-Efficiency and High-Precision Ship Detection Algorithm Based on Improved YOLOv8n,” Mathematics, vol. 12, no. 7, Apr. 2024, doi: 10.3390/math12071072.

[32] A. Wang et al., “YOLOv10: Real-Time End-to-End Object Detection,” in Advances in Neural Information Processing Systems, A. Globerson, L. Mackey, D. Belgrave, A. Fan, U. Paquet, J. Tomczak, and C. Zhang, Eds., Curran Associates, Inc., 2024, pp. 107984–108011. doi: 10.52202/079017-3429.

[33] J. Jing and C. Li, “Identification of lightweight rail defects based on YOLOv11 improvement,” in Fifth International Conference on Optical Imaging and Image Processing (ICOIP 2025), X. Hao and L. He, Eds., SPIE, 2025, p. 1368823. doi: 10.1117/12.3075801.

[34] X. Chen, N. Jiang, Z. Yu, W. Qian, and T. Huang, “Citrus leaf disease detection based on improved YOLO11 with C3K2,” in International Conference on Computer Graphics, Artificial Intelligence, and Data Processing (ICCAID 2024), X. Xu and A. bin Mohd Zain, Eds., SPIE, 2025, p. 135602Z. doi: 10.1117/12.3061501.

[35] Y. Tian, Q. Ye, and D. Doermann, “YOLOv12: Attention-Centric Real-Time Object Detectors Latency (ms) MS COCO mAP (%)”, doi: 10.0.