Model for Underwater Acoustic Target Recognition with Attention Mechanism Based on Residual Concatenate
Underwater acoustic target recognition remains a formidable challenge in underwater acoustic signal processing. Current target recognition approaches within underwater acoustic frameworks predominantly rely on acoustic image target recognition models. However, this method grapples with two primary s...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-12-01
|
| Series: | Journal of Marine Science and Engineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2077-1312/12/1/24 |
| _version_ | 1827371765206089728 |
|---|---|
| author | Zhe Chen Guohao Xie Mingsong Chen Hongbing Qiu |
| author_facet | Zhe Chen Guohao Xie Mingsong Chen Hongbing Qiu |
| author_sort | Zhe Chen |
| collection | DOAJ |
| description | Underwater acoustic target recognition remains a formidable challenge in underwater acoustic signal processing. Current target recognition approaches within underwater acoustic frameworks predominantly rely on acoustic image target recognition models. However, this method grapples with two primary setbacks; the pronounced frequency similarity within acoustic images often leads to the loss of critical target data during the feature extraction phase, and the inherent data imbalance within the underwater acoustic target dataset predisposes models to overfitting. In response to these challenges, this research introduces an underwater acoustic target recognition model named Attention Mechanism Residual Concatenate Network (ARescat). This model integrates residual concatenate networks combined with Squeeze-Excitation (SE) attention mechanisms. The entire process culminates with joint supervision employing Focal Loss for precise feature classification. In our study, we conducted recognition experiments using the ShipsEar database and compared the performance of the ARescat model with the classic ResNet18 model under identical feature extraction conditions. The findings reveal that the ARescat model, with a similar quantity of model parameters as ResNet18, achieves a 2.8% higher recognition accuracy, reaching an impressive 95.8%. This enhancement is particularly notable when comparing various models and feature extraction methods, underscoring the ARescat model’s superior proficiency in underwater acoustic target recognition. |
| first_indexed | 2024-03-08T10:46:17Z |
| format | Article |
| id | doaj.art-690f0038a3c342088aa6d449157bebf8 |
| institution | Directory Open Access Journal |
| issn | 2077-1312 |
| language | English |
| last_indexed | 2024-03-08T10:46:17Z |
| publishDate | 2023-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Marine Science and Engineering |
| spelling | doaj.art-690f0038a3c342088aa6d449157bebf82024-01-26T17:13:18ZengMDPI AGJournal of Marine Science and Engineering2077-13122023-12-011212410.3390/jmse12010024Model for Underwater Acoustic Target Recognition with Attention Mechanism Based on Residual ConcatenateZhe Chen0Guohao Xie1Mingsong Chen2Hongbing Qiu3School of Information and Communication, Guilin University of Electronic Technology, Guilin 541004, ChinaSchool of Ocean Engineering, Guilin University of Electronic Technology, Beihai 536000, ChinaSchool of Information and Communication, Guilin University of Electronic Technology, Guilin 541004, ChinaSchool of Information and Communication, Guilin University of Electronic Technology, Guilin 541004, ChinaUnderwater acoustic target recognition remains a formidable challenge in underwater acoustic signal processing. Current target recognition approaches within underwater acoustic frameworks predominantly rely on acoustic image target recognition models. However, this method grapples with two primary setbacks; the pronounced frequency similarity within acoustic images often leads to the loss of critical target data during the feature extraction phase, and the inherent data imbalance within the underwater acoustic target dataset predisposes models to overfitting. In response to these challenges, this research introduces an underwater acoustic target recognition model named Attention Mechanism Residual Concatenate Network (ARescat). This model integrates residual concatenate networks combined with Squeeze-Excitation (SE) attention mechanisms. The entire process culminates with joint supervision employing Focal Loss for precise feature classification. In our study, we conducted recognition experiments using the ShipsEar database and compared the performance of the ARescat model with the classic ResNet18 model under identical feature extraction conditions. The findings reveal that the ARescat model, with a similar quantity of model parameters as ResNet18, achieves a 2.8% higher recognition accuracy, reaching an impressive 95.8%. This enhancement is particularly notable when comparing various models and feature extraction methods, underscoring the ARescat model’s superior proficiency in underwater acoustic target recognition.https://www.mdpi.com/2077-1312/12/1/24SE attention mechanismresidual network (ResNet)underwater acoustic target recognitionfeature extraction |
| spellingShingle | Zhe Chen Guohao Xie Mingsong Chen Hongbing Qiu Model for Underwater Acoustic Target Recognition with Attention Mechanism Based on Residual Concatenate Journal of Marine Science and Engineering SE attention mechanism residual network (ResNet) underwater acoustic target recognition feature extraction |
| title | Model for Underwater Acoustic Target Recognition with Attention Mechanism Based on Residual Concatenate |
| title_full | Model for Underwater Acoustic Target Recognition with Attention Mechanism Based on Residual Concatenate |
| title_fullStr | Model for Underwater Acoustic Target Recognition with Attention Mechanism Based on Residual Concatenate |
| title_full_unstemmed | Model for Underwater Acoustic Target Recognition with Attention Mechanism Based on Residual Concatenate |
| title_short | Model for Underwater Acoustic Target Recognition with Attention Mechanism Based on Residual Concatenate |
| title_sort | model for underwater acoustic target recognition with attention mechanism based on residual concatenate |
| topic | SE attention mechanism residual network (ResNet) underwater acoustic target recognition feature extraction |
| url | https://www.mdpi.com/2077-1312/12/1/24 |
| work_keys_str_mv | AT zhechen modelforunderwateracoustictargetrecognitionwithattentionmechanismbasedonresidualconcatenate AT guohaoxie modelforunderwateracoustictargetrecognitionwithattentionmechanismbasedonresidualconcatenate AT mingsongchen modelforunderwateracoustictargetrecognitionwithattentionmechanismbasedonresidualconcatenate AT hongbingqiu modelforunderwateracoustictargetrecognitionwithattentionmechanismbasedonresidualconcatenate |