A Deep Convolutional Neural Network Inspired by Auditory Perception for Underwater Acoustic Target Recognition
<p>The architecture of ADCNN.</p> "> Figure 2
<p>Spectrogram of recordings. (<b>a</b>) Cargo recording; (<b>b</b>) Passenger ship recording; (<b>c</b>) Tanker recording; (<b>d</b>) Environment noise recording.</p> "> Figure 3
<p>Visualization of the output of each filter. (<b>a</b>) Testing sample of Cargo class; (<b>b</b>) Testing sample of Passenger ship class.</p> "> Figure 4
<p>Result of t-SNE feature visualization. (<b>a</b>–<b>e</b>) Feature groups of deep filter sub-networks; (<b>f</b>) Features of layer-1; (<b>g</b>) Features of layer-2.</p> "> Figure 5
<p>ROC curves of the proposed model and its competitors. (<b>a</b>) Cargo class is positive class; (<b>b</b>) Passenger ship class is positive class; (<b>c</b>) Tanker class is positive class; (<b>d</b>) Environment noise class is positive class.</p> ">
Abstract
:1. Introduction
- In the proposed model, a deep filter sub-network, which is composed of deep convolution filters, max-pooling and several full connected layers, is presented to simulate the deep acoustic information extraction structure of auditory system.
- Inspired by the frequency component perception neural mechanism, the complex frequency components of ship-radiated noise are decomposed and modeled by a bank of multi-scale deep filter sub-networks.
- Inspired by the plasticity neural mechanism, the parameters of the multi-scale deep filter sub-networks are learned from the raw time domain ship-radiated noise signals.
- The experimental results demonstrate that the proposed ADCNN model is effective for underwater acoustic target recognition. It can decompose, model and classify ship-radiated noise signal efficiently, and achieve better classification performance than the compared methods.
2. Auditory Perception Inspired Deep Convolutional Neural Network for UATR
2.1. The Neural Mechanisms of Auditory Perception
2.2. The Architecture of ADCNN for UATR
3. Detailed Implementation of ADCNN for UATR
3.1. Learned Deep Filter Sub-Network
3.2. Ship Radiated Noise Signal Decomposition with a Bank of Multi-Scale Deep Filter Sub-Networks
3.3. The Plasticity of ADCNN Model for Underwater Acoustic Target Recognition
4. Experimental Dataset
5. Experiments and Discussion
5.1. Experimental Setup
- The ship-radiated noise frequency decompose performance of the proposed model is observed by visualizing the outputs of filters in the deep filter sub-networks.
- The classification performance of the proposed model is evaluated by receivers operating characteristic (ROC) curve, area under ROC curves (AUC) value and classification accuracy, and is compared with several different methods.
5.2. Time Domain Ship Radiated Noise Decomposition
5.3. Feature Visualization by t-SNE
5.4. Classification Experiments
6. Conclusions
Author Contributions
Funding
Conflicts of Interest
References
- Meng, Q.; Yang, S. A wave structure based method for recognition of marine acoustic target signals. J. Acoust. Soc. Am. 2015, 137, 2242. [Google Scholar] [CrossRef]
- Das, A.; Kumar, A.; Bahl, R. Marine vessel classification based on passive sonar data: The cepstrum-based approach. IET Radar Sonar Navig. 2013, 7, 87–93. [Google Scholar] [CrossRef]
- Wei, X.; Li, G.; Wang, Z.Q. Underwater target recognition based on wavelet packet and principal component analysis. Comput. Simul. 2011, 28, 8–290. [Google Scholar] [CrossRef]
- Yang, H.; Shen, S. The Feature Selection of Pattern Recognition; Publishing House of Electronic Industry: Beijing, China, 2016. [Google Scholar]
- Yang, H.; Gan, A.; Chen, H.; Yue, P.; Tang, J.; Li, J. Underwater acoustic target recognition using SVM ensemble via weighted sample and feature selection. In Proceedings of the 13th International Bhurban Conference on Applied Sciences and Technology, Islamabad, Pakistan, 12–16 January 2016. [Google Scholar] [CrossRef]
- Filho, W.S.; Seixas, J.M.D.; Moura, N.N.D. Preprocessing passive sonar signals for neural classification. IET Radar Sonar Navig. 2011, 5, 605–612. [Google Scholar] [CrossRef]
- Cao, X.; Zhang, X.; Yu, Y.; Niu, L. Deep learning-based recognition of underwater target. In Proceedings of the IEEE International Conference on Digital Signal Processing, Beijing, China, 16–18 October 2016. [Google Scholar] [CrossRef]
- Yue, H.; Zhang, L.; Wang, D.; Wang, Y.; Lu, Z. The classification of underwater acoustic targets based on deep learning methods. Adv. Intell. Syst. Res. 2017, 134, 526–529. [Google Scholar]
- Yang, H.; Shen, S.; Yao, X.; Sheng, M.; Wang, C. Competitive deep-belief networks for underwater acoustic target recognition. Sensors 2018, 18, 952. [Google Scholar] [CrossRef] [PubMed]
- Shen, S.; Yang, H.; Li, J.; Xu, G.; Sheng, M. Auditory inspired convolutional neural networks for ship type classification with raw hydrophone data. Entropy 2018, 20, 990. [Google Scholar] [CrossRef]
- Gazzaniga, M.; Ivry, R.B.; Mangun, G.R. Cognitive Neuroscience The Biology of the Mind; W. W. Norton & Company: New York, NY, USA, 2018. [Google Scholar]
- Dallos, P.; Popper, A.N.; Fay, R.R. The Cochlea. In Springer Handbook of Auditory Research; Springer: Cham, Switzerland, 1996; Volume 65, pp. 291–312. [Google Scholar] [CrossRef]
- Brundin, L.; Flock, A.; Canlon, B. Sound-induced motility of isolated cochlear outer hair cells is frequency-specific. Nature 1989, 342, 814. [Google Scholar] [CrossRef] [PubMed]
- Schreiner, C.E.; Langner, G. Laminar fine structure of frequency organization in auditory midbrain. Nature 1997, 388, 383–386. [Google Scholar] [CrossRef] [PubMed]
- Schreiner, C.E.; Read, H.; Sutter, M. Modular organization of frequency integration in primary auditory cortex. Ann. Rev. Neurosci. 2000, 23, 501–529. [Google Scholar] [CrossRef] [PubMed]
- Strainer, J.C.; Ulmer, J.L.; Yetkin, F.Z.; Haughton, V.M.; Daniels, D.L.; Millen, S.J. Functional MR of the primary auditory cortex: An analysis of pure tone activation and tone discrimination. AJNR Am. J. Neuroradiol. 1997, 18, 601–610. [Google Scholar] [PubMed]
- Talavage, T.M.; Ledden, P.J.; Benson, R.R.; Rosen, B.R.; Melcher, J.R. Frequency-dependent responses exhibited by multiple regions in human auditory cortex. Hear. Res. 2000, 150, 225–244. [Google Scholar] [CrossRef]
- Kolb, B.; Gibb, R.; Robinson, T.E. Brain plasticity and behavior. Curr. Dir. Psychol. Sci. 2003, 12, 1–5. [Google Scholar] [CrossRef]
- Robertson, D.; Irvine, D.R.F. Plasticity of frequency organization in auditory cortex of guinea pigs with partial unilateral deafness. J. Comp. Neurol. 2010, 282, 456–471. [Google Scholar] [CrossRef] [PubMed]
- Weinberger, N.M. Learning-induced changes of auditory receptive fields. Curr. Opin. Neurobiol. 1993, 3, 570. [Google Scholar] [CrossRef]
- Weinberger, N.M. Experience-dependent response plasticity in the auditory cortex: Issues, characteristics, mechanisms, and functions. In Springer Handbook of Auditory Research; Springer: Cham, Switzerland, 2004. [Google Scholar] [CrossRef]
- Lecun, Y.; Bottou, L.; Bengio, Y.; Haffner, P. Gradient-based learning applied to document recognition. Proc. IEEE 1998, 86, 2278–2324. [Google Scholar] [CrossRef]
- Lecun, Y.; Bengio, Y.; Hinton, G.E. Deep learning. Nature 2015, 521, 436. [Google Scholar] [CrossRef] [PubMed]
- Arora, S.; Bhaskara, A.; Ge, R.; Ma, T. Provable bounds for learning some deep representations. In Proceedings of the 31st international Conference on Machine Learning, Beijing, China, 21–26 June 2014. [Google Scholar]
- Van der Maaten, L.; Hinton, G.E. Visualizing high-dimensional data using t-SNE. Vigiliae Christ. 2008, 9, 2579–2605. [Google Scholar]
- Zhang, L.; Wu, D.; Han, X.; Zhu, Z. Feature extraction of underwater target signal using mel frequency cepstrum coefficients based on acoustic vector sensor. J. Sens. 2016, 2016, 7864213. [Google Scholar] [CrossRef]
- Hu, G.; Wang, K.; Peng, Y.; Qiu, M.; Shi, J.; Liu, L. Deep learning methods for underwater target feature extraction and recognition. Comput. Intell. Neurosci. 2018, 2018, 1214301. [Google Scholar] [CrossRef] [PubMed]
The RMSProp algorithm |
---|
Input: |
For to T do |
Set and |
End for |
Data Set | Class | No. Ships | No. Acoustic Event | Total Time (Hour) | No. Samples |
---|---|---|---|---|---|
Training | Cargo | 13 | 6523 | 10.87 | 97,800 |
Passenger ship | 7 | 7326 | 12.21 | 109,900 | |
Tanker | 35 | 5921 | 9.87 | 88,800 | |
Environment noise | non | 10,497 | 17.49 | 157,400 | |
Test | Cargo | 9 | 1200 | 3.33 | 3000 |
Passenger ship | 10 | 1200 | 3.33 | 3000 | |
Tanker | 16 | 1200 | 3.33 | 3000 | |
Environment noise | non | 1200 | 3.33 | 3000 |
Input | Methods | Accuracy/% |
---|---|---|
MFCC [26] features | DNN model | 78.92 |
Frequency spectrum features | DNN model [8] | 81.27 |
Raw time domain signal | CNN model [27] | 77.01 |
Raw time domain signal | Proposed model | 81.96 |
Predicted Label | Cargo | Passenger Ship | Tanker | Environment Noise | Recall (%) | |
---|---|---|---|---|---|---|
Ture Label | ||||||
Cargo | 927 | 100 | 141 | 32 | 77.25 | |
Passenger ship | 39 | 1045 | 80 | 36 | 87.08 | |
Tanker | 216 | 98 | 832 | 54 | 69.33 | |
Environment noise | 4 | 52 | 14 | 1130 | 94.17 | |
Precision (%) | 78.16 | 80.69 | 77.98 | 90.26 | 81.96 |
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Yang, H.; Li, J.; Shen, S.; Xu, G. A Deep Convolutional Neural Network Inspired by Auditory Perception for Underwater Acoustic Target Recognition. Sensors 2019, 19, 1104. https://doi.org/10.3390/s19051104
Yang H, Li J, Shen S, Xu G. A Deep Convolutional Neural Network Inspired by Auditory Perception for Underwater Acoustic Target Recognition. Sensors. 2019; 19(5):1104. https://doi.org/10.3390/s19051104
Chicago/Turabian StyleYang, Honghui, Junhao Li, Sheng Shen, and Guanghui Xu. 2019. "A Deep Convolutional Neural Network Inspired by Auditory Perception for Underwater Acoustic Target Recognition" Sensors 19, no. 5: 1104. https://doi.org/10.3390/s19051104
APA StyleYang, H., Li, J., Shen, S., & Xu, G. (2019). A Deep Convolutional Neural Network Inspired by Auditory Perception for Underwater Acoustic Target Recognition. Sensors, 19(5), 1104. https://doi.org/10.3390/s19051104