A REVIEW OF UNDERWATER ACOUSTIC TARGET RECOGNITION IN THE ERA OF ARTIFICIAL INTELLIGENCE
DOI:
https://doi.org/10.62985/j.huit_ojs.vol26.no2E.406Keywords:
Passive sonar, deep learning, feature extraction, adversarial robustness, edge computingAbstract
Underwater Acoustic Target Recognition (UATR) is a critical technology for maritime domain awareness, naval defense, and oceanographic monitoring. With the increasing shift of naval operations toward complex littoral environments, conventional passive sonar systems that rely heavily on human operators and classical signal processing methods are encountering substantial difficulties. These challenges stem primarily from low signal-to-noise ratios (SNR), multipath propagation effects, and highly non-stationary background interference.
This paper presents a comprehensive review of the evolution of UATR technologies, progressing from traditional physics-based approaches such as LOFAR and DEMON to advanced deep learning frameworks. We critically analyze the limitations of conventional machine learning techniques (e.g., SVM and HMM) and evaluate how modern neural network architectures - including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), attention mechanisms, Transformers, and Generative Adversarial Networks (GANs) - have significantly enhanced feature extraction and classification performance in underwater environments.
Furthermore, this review addresses key contemporary challenges, including data scarcity, model interpretability, adversarial vulnerability, and the deployment of efficient models on resource-constrained edge devices such as sonobuoys. By synthesizing recent advancements, we highlight existing research gaps and propose future directions to develop more robust, intelligent, and practical underwater acoustic surveillance systems.
References
[1] B. T. Giang, Underwater Acoustics Theory and Applications, Nha Trang: Naval Academy Publishing House, 2021.
[2] G. Jin, F. Liu, H. Wu, and Q. Song, “Deep learning-based framework for expansion, recognition and classification of underwater acoustic signal,” Journal of Experimental & Theoretical Artificial Intelligence, vol. 32, no. 2, pp. 205–218, 2020, doi: 10.1080/0952813X.2019.1647560.
[3] D. Liu, W. Shen, W. Cao, W. Hou, and B. Wang, “Design of Siamese network for underwater target recognition with small sample size,” Applied Sciences, vol. 12, no. 20, Art. no. 10659, 2022, doi: 10.3390/app122010659.
[4] X. Zeng and S. Wang, “Underwater sound classification based on Gammatone filter bank and Hilbert-Huang transform,” in Proc. IEEE Int. Conf. Signal Processing, Communications and Computing (ICSPCC), Guilin, China, 2014, pp. 707–710, doi: https://doi.org/10.1109/ICSPCC.2014.6986287.
[5] G. Hu, K. Wang, and L. Liu, “Underwater acoustic target recognition based on depthwise separable convolution neural networks,” Sensors, vol. 21, no. 4, Art. no. 1429, 2021, doi: https://doi.org/10.3390/s21041429.
[6] J. Liu, Y. He, Z. Liu, and Y. Xiong, “Underwater target recognition based on line spectrum and support vector machine,” in Proc. 2014 Int. Conf. Mechatronics, Control and Electronic Engineering (MCE), Shenyang, China, 2014, pp. 79–84, doi: https://doi.org/10.2991/mce-14.2014.17.
[7] S. Feng, S. Ma, X. Zhu, M. Yan, and H. Xu, “Artificial intelligence-based underwater acoustic targets recognition: A survey,” Remote Sensing, vol. 16, no. 17, Art. no. 3333, 2024, doi: https://doi.org/10.3390/rs16173333.
[8] J. Choi, Y. Choo, and K. Lee, “Acoustic classification of surface and underwater vessels in the ocean using supervised machine learning,” Sensors, vol. 19, no. 16, Art. no. 3492, 2019, doi: https://doi.org/10.3390/s19163492.
[9] N. Müller, J. Reermann, and T. Meisen, “Navigating the depths: A comprehensive survey of deep learning for passive underwater acoustic target recognition,” IEEE Access, vol. 12, pp. 154092–154118, 2024, doi: https://doi.org/10.1109/ACCESS.2024.3480788.
[10] Y. Liang, S. Li, C. Yan, M. Li, and C. Jiang, “Explaining the black-box model: A survey of local interpretation methods for deep neural networks,” Neurocomputing, vol. 419, pp. 168–182, 2021, doi: https://doi.org/10.1016/j.neucom.2020.08.011.
[11] D. Li, F. Liu, T. Shen, L. Chen, X. Yang, and D. Zhao, “Generalizable underwater acoustic target recognition using feature extraction module of neural network,” Applied Sciences, vol. 12, no. 21, Art. no. 10804, 2022, doi: https://doi.org/10.3390/app122110804.
[12] S. Feng, X. Zhu, S. Ma, and Q. Lan, “Adversarial attacks in underwater acoustic target recognition with deep learning models,” Remote Sensing, vol. 15, no. 22, Art. no. 5386, 2023, doi: https://doi.org/10.3390/rs15225386.
[13] M. Irfan, Z. Jiangbin, S. Ali, M. Iqbal, Z. Masood, and U. Hamid, “DeepShip: An underwater acoustic benchmark dataset and a separable convolution based autoencoder for classification,” Expert Systems with Applications, vol. 183, Art. no. 115270, 2021, doi: https://doi.org/10.1016/j.eswa.2021.115270.
