Similar articles

Abstract: Due to the complex and changeable marine environment, the active sonar target recognition problem has always been difficult in the field of underwater acoustics. Deep learning-based fusion recognition technology provides an effective way to solve this problem, but relying on simple concatenation strategies to fuse multi-domain features can cause information redundancy, and it is not easy to effectively mine correlation information between domains. Therefore, this paper proposes an attention mechanism-based multi-domain feature fusion approach for active sonar target recognition. By preprocessing active sonar echo signals and constructing a multi-domain feature extraction and fusion network, this method uses a one-dimensional convolutional neural network with long short-term memory (1DCNN-LSTM) and a two-dimensional convolutional neural network (2DCNN) with channel attention introduced to extract deep features from different domains. Subsequently, combining feature concatenation and constructing multi-domain cross-attention, intra- and cross-domain feature fusion is performed, which can effectively eliminate redundant information and promote inter-domain information interaction, while maximizing the retention of target features. Experimental results show that compared with single-domain methods, the network using an attention mechanism for multi-domain feature fusion strengthens cross-domain information interaction and significantly improves feature representation capability. Compared with other methods, the proposed method has obvious advantages in performance and maintains stable generalization ability in scenarios with low signal-clutter ratios.

一种基于注意力机制的主动声呐目标多域特征融合识别方法

[1]Arrabito GR, Cooke BE, McFadden SM, 2005. Recommendations for enhancing the role of the auditory modality for processing sonar data. Appl Acoust, 66(8):986-1005.

[2]Choo Y, Lee K, Hong W, et al., 2024. Active underwater target detection using a shallow neural network with spectrogram-based temporal variation features. IEEE J Ocean Eng, 49(1):279-293.

[3]Domingos LCF, Santos PE, Skelton PSM, et al., 2022. A survey of underwater acoustic data classification methods using deep learning for shoreline surveillance. Sensors, 22(6):2181.

[4]Dosovitskiy A, Beyer L, Kolesnikov A, et al., 2021. An image is worth 16x16 words: Transformers for image recognition at scale.

[5]Fang SL, Du SP, Luo XW, et al., 2019. Feature analysis and recognition technology of underwater acoustic targets. Bull Chin Acad Sci, 34(3):297-305(in Chinese).

[6]Gao F, Jin XP, Zhou XW, et al., 2025. MSFMamba: multiscale feature fusion state space model for multisource remote sensing image classification. IEEE Trans Geosci Remote Sens, 63:5504116.

[7]Han XC, Ren CX, Wang LM, et al., 2022. Underwater acoustic target recognition method based on a joint neural network. PLoS ONE, 17(4):e0266425.

[8]He KM, Zhang XY, Ren SQ, et al., 2016. Deep residual learning for image recognition. Proc IEEE Conf on Computer Vision and Pattern Recognition, p.770-778.

[9]He L, Liu SY, An R, et al., 2023. An end-to-end framework based on vision-language fusion for remote sensing cross-modal text-image retrieval. Mathematics, 11(10):2279.

[10]Hong F, Liu CW, Guo L, et al., 2021. Underwater acoustic target recognition with a residual network and the optimized feature extraction method. Appl Sci, 11(4):1442.

[11]Hu G, Wang KJ, Peng Y, et al., 2018. Deep learning methods for underwater target feature extraction and recognition. Comput Intell Neurosci, 2018:1214301.

[12]Hu G, Wang KJ, Liu LL, 2021. Underwater acoustic target recognition based on depthwise separable convolution neural networks. Sensors, 21(4):1429.

[13]Hu J, Shen L, Sun G, 2018. Squeeze-and-excitation networks. Proc IEEE/CVF Conf on Computer Vision and Pattern Recognition, p.7132-7141.

[14]Huang HN, Li Y, 2019. Underwater acoustic detection: current status and future trends. Bull Chin Acad Sci, 34(3):264-271(in Chinese).

[15]Kamal S, Chandran CS, Supriya MH, 2021. Passive sonar automated target classifier for shallow waters using end-to-end learnable deep convolutional LSTMs. Eng Sci Technol Int J, 24(4):860-871.

[16]Khan A, Fouda MM, Do DT, et al., 2024. Underwater target detection using deep learning: methodologies, challenges, applications, and future evolution. IEEE Access, 12:12618-12635.

[17]Lee S, Seo I, Seok J, et al., 2020. Active sonar target classification with power-normalized cepstral coefficients and convolutional neural network. Appl Sci, 10(23):8450.

[18]Li HC, Lu YT, Zhu HD, 2024. Multi-modal sentiment analysis based on image and text fusion based on cross-attention mechanism. Electronics, 13(11):2069.

[19]Liu DL, Zhao XC, Cao WJ, et al., 2020. Design and performance evaluation of a deep neural network for spectrum recognition of underwater targets. Comput Intell Neurosci, 2020:8848507.

[20]Pan XY, Sun J, Feng TH, et al., 2025. Underwater target recognition based on adaptive multi-feature fusion network. Multim Tools Appl, 84(10):7297-7317.

[21]Shadlou Jahromi M, Bagheri V, Rostami H, et al., 2019. Feature extraction in fractional Fourier domain for classification of passive sonar signals. J Signal Process Syst, 91(5):511-520.

[22]Shin FB, Kil DH, Wayland RF, 1997. Active impulsive echo discrimination in shallow water by mapping target physics-derived features to classifiers. IEEE J Ocean Eng, 22(1):66-80.

[23]Simonyan K, Zisserman A, 2015. Very deep convolutional networks for large-scale image recognition.

[24]Vaswani A, Shazeer N, Parmar N, et al., 2017. Attention is all you need. Proc 31^st Int Conf on Neural Information Processing Systems, p.6000-6010.

[25]Wang QC, Du SP, Zhang W, et al., 2024. Active sonar target recognition method based on multi-domain transformations and attention-based fusion network. IET Radar Sonar Navig, 18(10):1814-1828.

[26]Yang HH, Gan AQ, Chen HL, et al., 2016. Underwater acoustic target recognition using SVM ensemble via weighted sample and feature selection. 13^th Int Bhurban Conf on Applied Sciences and Technology, p.522-527.

[27]Young VW, Hines PC, 2007. Perception-based automatic classification of impulsive-source active sonar echoes. J Acoust Soc Am, 122(3):1502-1517.

[28]Zhang W, Wu YQ, Wang DZ, et al., 2018. Underwater target feature extraction and classification based on Gammatone filter and machine learning. Int Conf on Wavelet Analysis and Pattern Recognition, p.42-47.