Generalizable Underwater Acoustic Target Recognition Using Feature Extraction Module of Neural Network
The underwater acoustic target signal is affected by factors such as the underwater environment and the ship’s working conditions, causing the generalization of the recognition model is essential. This study is devoted to improving the generalization of recognition models, proposing a feature extrac...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/12/21/10804 |
| _version_ | 1797469307128512512 |
|---|---|
| author | Daihui Li Feng Liu Tongsheng Shen Liang Chen Xiaodan Yang Dexin Zhao |
| author_facet | Daihui Li Feng Liu Tongsheng Shen Liang Chen Xiaodan Yang Dexin Zhao |
| author_sort | Daihui Li |
| collection | DOAJ |
| description | The underwater acoustic target signal is affected by factors such as the underwater environment and the ship’s working conditions, causing the generalization of the recognition model is essential. This study is devoted to improving the generalization of recognition models, proposing a feature extraction module based on neural network and time-frequency analysis, and validating the feasibility of the model-based transfer learning method. A network-based filter based on one-dimensional convolution is built according to the calculation mode of the finite impulse response filter. An attention-based model is constructed using the convolution network components and full-connection components. The attention-based network utilizes convolution components to perform the Fourier transform and feeds back the optimization gradient of a specific task to the network-based filter. The network-based filter is designed to filter the observed signal for adaptive perception, and the attention-based model is constructed to extract the time-frequency features of the signal. In addition, model-based transfer learning is utilized to further improve the model’s performance. Experiments show that the model can perceive the frequency domain features of underwater acoustic targets, and the proposed method demonstrates competitive performance in various classification tasks on real data, especially those requiring high generalizability. |
| first_indexed | 2024-03-09T19:18:35Z |
| format | Article |
| id | doaj.art-eb29efefaf334d90bad5a1053c660405 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-09T19:18:35Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-eb29efefaf334d90bad5a1053c6604052023-11-24T03:33:06ZengMDPI AGApplied Sciences2076-34172022-10-0112211080410.3390/app122110804Generalizable Underwater Acoustic Target Recognition Using Feature Extraction Module of Neural NetworkDaihui Li0Feng Liu1Tongsheng Shen2Liang Chen3Xiaodan Yang4Dexin Zhao5National Innovation Institute of Defense Technology, Chinese Academy of Military Science, Beijing 100091, ChinaNational Innovation Institute of Defense Technology, Chinese Academy of Military Science, Beijing 100091, ChinaNational Innovation Institute of Defense Technology, Chinese Academy of Military Science, Beijing 100091, ChinaNational Innovation Institute of Defense Technology, Chinese Academy of Military Science, Beijing 100091, ChinaCollege of Electronic Information Engineering, Beihang University, Beijing 100190, ChinaNational Innovation Institute of Defense Technology, Chinese Academy of Military Science, Beijing 100091, ChinaThe underwater acoustic target signal is affected by factors such as the underwater environment and the ship’s working conditions, causing the generalization of the recognition model is essential. This study is devoted to improving the generalization of recognition models, proposing a feature extraction module based on neural network and time-frequency analysis, and validating the feasibility of the model-based transfer learning method. A network-based filter based on one-dimensional convolution is built according to the calculation mode of the finite impulse response filter. An attention-based model is constructed using the convolution network components and full-connection components. The attention-based network utilizes convolution components to perform the Fourier transform and feeds back the optimization gradient of a specific task to the network-based filter. The network-based filter is designed to filter the observed signal for adaptive perception, and the attention-based model is constructed to extract the time-frequency features of the signal. In addition, model-based transfer learning is utilized to further improve the model’s performance. Experiments show that the model can perceive the frequency domain features of underwater acoustic targets, and the proposed method demonstrates competitive performance in various classification tasks on real data, especially those requiring high generalizability.https://www.mdpi.com/2076-3417/12/21/10804underwater acoustics target recognitiondeep learningtime-frequency analysisfeature extractiondata analysis |
| spellingShingle | Daihui Li Feng Liu Tongsheng Shen Liang Chen Xiaodan Yang Dexin Zhao Generalizable Underwater Acoustic Target Recognition Using Feature Extraction Module of Neural Network Applied Sciences underwater acoustics target recognition deep learning time-frequency analysis feature extraction data analysis |
| title | Generalizable Underwater Acoustic Target Recognition Using Feature Extraction Module of Neural Network |
| title_full | Generalizable Underwater Acoustic Target Recognition Using Feature Extraction Module of Neural Network |
| title_fullStr | Generalizable Underwater Acoustic Target Recognition Using Feature Extraction Module of Neural Network |
| title_full_unstemmed | Generalizable Underwater Acoustic Target Recognition Using Feature Extraction Module of Neural Network |
| title_short | Generalizable Underwater Acoustic Target Recognition Using Feature Extraction Module of Neural Network |
| title_sort | generalizable underwater acoustic target recognition using feature extraction module of neural network |
| topic | underwater acoustics target recognition deep learning time-frequency analysis feature extraction data analysis |
| url | https://www.mdpi.com/2076-3417/12/21/10804 |
| work_keys_str_mv | AT daihuili generalizableunderwateracoustictargetrecognitionusingfeatureextractionmoduleofneuralnetwork AT fengliu generalizableunderwateracoustictargetrecognitionusingfeatureextractionmoduleofneuralnetwork AT tongshengshen generalizableunderwateracoustictargetrecognitionusingfeatureextractionmoduleofneuralnetwork AT liangchen generalizableunderwateracoustictargetrecognitionusingfeatureextractionmoduleofneuralnetwork AT xiaodanyang generalizableunderwateracoustictargetrecognitionusingfeatureextractionmoduleofneuralnetwork AT dexinzhao generalizableunderwateracoustictargetrecognitionusingfeatureextractionmoduleofneuralnetwork |