Dual deep feature fusion and neuro-fuzzy classification with grad-CAM++ for robust underwater object detection

Document Type : Research Paper

Authors

1 School of Computer science & Engineering, VIT-AP University, Inavolu, Amaravati, Andhra Pradesh-522237

2 School of Computer science & Engineering, VIT-AP University, Inavolu, Amaravati, Andhra Pradesh-522237

Abstract

Detecting underwater objects is crucial for scientific inquiry, resource development, and marine conservation. Current  underwater object detection methods face two main challenges: (1) high parameter counts leading to excessive memory  usage, and (2) reduced detection accuracy. To address these limitations, this study introduces a computationally efficient  underwater object recognition framework that integrates advanced feature extraction with fuzzy-driven adaptive  classification. A novel dual-stream extractor combining Improved YOLOv9 for refined spatial localization and a Modified  TResNet for enhanced semantic feature extraction is developed to capture complementary deep features. These streams  are fused using Rényi entropy, ensuring that only the most informative and non-redundant representations are preserved. The fused features are then classified using a Neuro-Fuzzy Random Vector Functional Link (NF-RVFL) network, enabling  fast, robust, and interpretable decision-making with minimal parameter overhead. Grad-CAM++ is incorporated to  highlight critical attention regions, further improving framework transparency. Extensive experiments conducted on four  public underwater datasets URPC, DLMU 2024, DUO, and RUOD show that the proposed framework consistently  surpasses state-of-the-art detectors, achieving notable improvements in mAP across all benchmarks. Comprehensive  ablation analysis and statistical evaluations confirm the contribution of each proposed component. Furthermore,  deployment evaluation demonstrates that the proposed framework’s reduced parameter count and high efficiency make  it well suited for real-time underwater platforms with limited computational resources. 

Keywords

Main Subjects

References

[1] A. Ali, M. Vadivel, Efficient energy congestion control scheme for wireless sensor networks using adaptive neuro fuzzy inference system with black widow optimization, Iranian Journal of Fuzzy Systems, 20(6) (2023), 187-202. https://doi.org/10.22111/ijfs.2023.43990.7748
[2] L. Cao, L. Shen, M. Yu, et al., Prior-guided dual-reference contrastive learning for underwater object detection, IEEE Transactions on Circuits and Systems for Video Technology, 35(8) (2025), 8209-8224. https://doi.org/10. 1109/TCSVT.2025.3545694
[3] P. Chakurkar, D. Vora, An integrated deep learning and fuzzy logic system for road crack severity analysis, and pedestrian fall risk prediction, Iranian Journal of Fuzzy Systems, 22(6) (2025), 103-123. https://doi.org/10. 22111/ijfs.2025.52137.9268
[4] W. Chen, H. Huang, S. Peng, et al., YOLO-face: A real-time face detector, The Visual Computer, 37(4) (2021), 805-813. https://doi.org/10.1007/s00371-020-01831-7
[5] X. Chen, Q. Zeng, C. Wei, et al., A unified underwater object detection network with image adaptive enhancement and attention-driven feature fusion, Optics and Laser Technology, 199 (2026), 115028. https://doi.org/10.1016/ j.optlastec.2026.115028
[6] L. Dai, H. Liu, P. Song, M. Liu, A gated cross-domain collaborative network for underwater object detection, Pattern Recognition, 149 (2024), 110222. https://doi.org/10.1016/j.patcog.2023.110222
[7] M. Eftekhari, A. Mehrpooya, F. Saberi-Movahed, V. Torra, How fuzzy concepts contribute to machine learning, Cham, Switzerland: Springer, 2022. https://doi.org/10.1007/978-3-030-94066-9
[8] B. Farhadinia, U. Aickelin, H. A. Khorshidi, A parametric similarity measure for extended picture fuzzy sets and its application in pattern recognition, Iranian Journal of Fuzzy Systems, 19(6) (2022), 141-160. https://doi.org/10. 22111/ijfs.2022.7217
[9] L. Folkman, K. A. Pitt, B. Stantic, A data-centric framework for combating domain shift in underwater object detection with image enhancement, Applied Intelligence, 55(4) (2025), 1-26. https://doi.org/10.1007/ s10489-024-06224-0
[10] C. Fu, X. Fan, J. Xiao, et al., Learning heavily-degraded prior for underwater object detection, IEEE Transactions on Circuits and Systems for Video Technology, 33(11) (2023), 6887-6896. https://doi.org/10.1109/TCSVT.2023. 3271644
[11] J. Gao, Y. Zhang, X. Geng, et al., PE-Transformer: Path enhanced transformer for improving underwater object detection, Expert Systems with Applications, 246 (2024), 123253. https://doi.org/10.1016/j.eswa.2024.123253
[12] K. Guo, Z. Sun, Zhang, A lightweight YOLOv8 integrating FasterNet for real-time underwater object detection, Journal of Real-Time Image Processing, 21(2) (2024), 49. https://doi.org/10.1007/s11554-024-01431-x
[13] X. Ji, S. Chen, L. Y. Hao, et al., FBDPN: CNN-Transformer hybrid feature boosting and differential pyramid network for underwater object detection, Expert Systems with Applications, 256 (2024), 124978. https://doi. org/10.1016/j.eswa.2024.124978
[14] D. Kong, Y. Zhang, X. Zhao, et al., CFSPNet: Cross-domain feature synergy and perception for robust underwater object detection, Expert Systems with Applications, 320 (2026), 132174. https://doi.org/10.1016/j.eswa.2026. 132174
[15] C. Li, W. Liu, G. Gong, et al., SU-YOLO: Spiking neural network for efficient underwater object detection, Neurocomputing, 644 (2025), 130310. https://doi.org/10.1016/j.neucom.2025.130310
[16] Z. Liu, B. Wang, Y. Li, et al., UnitModule: A lightweight joint image enhancement module for underwater object detection, Pattern Recognition, 151 (2024), 110435. https://doi.org/10.1016/j.patcog.2024.110435
[17] P. Patro, K. Kumar, G. S. Kumar, A. K. Sahu, Intelligent data classification using optimized fuzzy neural network and improved cuckoo search optimization, Iranian Journal of Fuzzy Systems, 20(6) (2023), 155-169. https://doi. org/10.22111/ijfs.2023.44767.7887
[18] N. Rishi, A. Kumar, R. Golash, Fisheries management with deep learning-based fish species detection: A sustainable approach, In International Conference on Cognitive Computing and Cyber Physical Systems (pp. 359-369). Springer, Singapore, 1128 (2025). https://doi.org/10.1007/978-981-97-7371-8_28
[19] X. Shen, H. Wang, Y. Li, et al., Criss-cross global interaction-based selective attention in YOLO for underwater object detection, Multimedia Tools and Applications, 83(7) (2024), 20003-20032. https://doi.org/10.1007/ s11042-023-16311-y
[20] P. Song, P. Li, L. Dai, et al., Boosting R-CNN: Reweighting R-CNN samples by RPN’s error for underwater object detection, Neurocomputing, 530 (2023), 150-164. https://doi.org/10.1016/j.neucom.2023.01.088
[21] T. Tian, J. Cheng, D. Wu, Z. Li, Lightweight underwater object detection based on image enhancement and multi-attention, Multimedia Tools and Applications, 83(23) (2024), 63075-63093. https://doi.org/10.1007/ s11042-023-18008-8
[22] A. Upadhayay, S. Srivastav, S. K. Biswas, et al., Evolutionary neuro-fuzzy random vector functional link with genetic optimization, In 2025 IEEE International Conference on Fuzzy Systems (FUZZ), (2025). https://doi.org/ 10.1109/FUZZ62266.2025.11152032
[23] B. Wang, Z. Wang, W. Guo, Y. Wang, A dual-branch joint learning network for underwater object detection, Knowledge-Based Systems, 293 (2024), 111672. https://doi.org/10.1016/j.knosys.2024.111672
[24] J. Wu, J. Chen, Q. Lu, et al., U-ATSS: A lightweight and accurate one-stage underwater object detection network, Signal Processing: Image Communication, 126 (2024), 117137. https://doi.org/10.1016/j.image.2024.117137
[25] Y. Wu, W. Wang, C. Lin, et al., Towards multimodal underwater object detection: A bidirectional feature re composition network and visual-sonar dataset, Expert Systems with Applications, 316 (2026), 131710. https: //doi.org/10.1016/j.eswa.2026.131710
[26] F. Yu, F. Xiao, C. Li, et al., AO-UOD: A novel paradigm for underwater object detection using Acousto-Optic fusion, IEEE Journal of Oceanic Engineering, 50(2) (2025). https://doi.org/10.1109/JOE.2025.3529121
[27] G. Yuan, J. Song, J. Li, IF-USOD: Multimodal information fusion interactive feature enhancement architecture for underwater salient object detection, Information Fusion, 117 (2025), 102806. https://doi.org/10.1016/j. inffus.2024.102806
[28] D. Zhang, C. Yu, Z. Li, et al., A lightweight network enhanced by attention-guided cross-scale interaction for underwater object detection, Applied Soft Computing, 184(Part B) (2025), 113811. https://doi.org/10.1016/j. asoc.2025.113811
[29] J. Zhou, Z. He, D. Zhang, et al., Spatial residual for underwater object detection, IEEE Transactions on Pattern Analysis and Machine Intelligence, 47(6) (2025), 4996-5013. https://doi.org/10.1109/TPAMI.2025.3548652
Iranian Journal of Fuzzy Systems
Volume 23, Issue 6
November and December 2026
Pages 157-188

Files

Share

How to cite

Statistics