Passive sonar automated target classifier for shallow waters using end-to-end learnable deep convolutional LSTMs
Open accessAbstract
Automated target recognition systems are increasingly employed in sonar systems to reduce manning and associated challenges. Although passive acoustic target recognition is an exceptionally challenging endeavor especially in shallow water scenarios, it is being used by naval forces of the world by virtue of its inherent advantages compared to the alternatives. In order to address these challenges as well as to exploit the latent and subtle features in the signal stream from the hydrophones, an end-to-end differentiable architecture is proposed in this paper. Here the key strategy is to rely on the data, instead of relying on the prior knowledge about the data. The raw acoustic signals from the hydrophones are directly fed to a pre-initialized 1-dimensional convolutional layer followed by a cascade of 2-dimensional convolutional spectro-temporal feature learners. Various auditory scales are used for pre-initializing, so as to emphasize the frequencies of interest. In order to better capture the temporal relations, a Bidirectional-LSTM layer with a trainable attention module is employed. The best configuration of the proposed classifier system yields an accuracy of 95.2% on a large acoustic dataset, collected from the shallows of the Indian ocean.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Novel methods successfully employ a combination of classic signal processing methods with machine-learning architectures to compensate for signal complexity by selecting the most distinctive features for classification.
“Novel methods are successfully employing a combination of classic signal processing methods (such as STFT) with ML architectures [125] as a strategy to compensate the signal complexity by selecting its most distinctive features for the classification procedure.”
From A Survey of Underwater Acoustic Data Classification Methods Using Deep Learning for Shoreline Surveillance (2022) · cites this paper as [125]
The convolutions applied in the system of [125] produced behaviour equivalent to a Short-time Fourier transform, with adaptive parameters learned from the original raw data.
“The analysis of the learned features showed that the convolutions applied were able to produce a behaviour equivalent to a Short-time Fourier transform, with adaptive parameters learned from the original raw data.”
From A Survey of Underwater Acoustic Data Classification Methods Using Deep Learning for Shoreline Surveillance (2022) · cites this paper as [125]
A system using convolutional filters as feature extractors, two-dimensional convolutional layers, and a Bidirectional-LSTM with a trainable attention module achieved high-accuracy results when tested on a shallow-water dataset from the Indian Ocean.
“This system was tested on a dataset obtained in shallow waters in the Indian Ocean, achieving high-accuracy results.”
From A Survey of Underwater Acoustic Data Classification Methods Using Deep Learning for Shoreline Surveillance (2022) · cites this paper as [125]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Ship Type Classification by Convolutional Neural Networks with Auditory-Like Mechanisms
Shen, S., Yang, H., Yao, X., Li, J., Xu, G., & Sheng, M. (2020) · Sensors · doi:10.3390/s20010253 · Cited by 57 · see citing works
Improved Auditory Inspired Convolutional Neural Networks for Ship Type Classification
Shen, S., Yang, H., & Li, J. (2019) · OCEANS 2019 - Marseille · doi:10.1109/oceanse.2019.8867312 · Cited by 19 · see citing works
Flagged first, then by citations
Research works printed but not linked: 30
- N.G. Pace, F. Jensen, Impact of Littoral Environmental Variability on Acoustic Predictions and Sonar Performance, vol. 1, Springer Science & Business Media, 2002.
- P. Abbot, I. Dyer, Sonar performance predictions incorporating environmental variability, in: Impact of Littoral Environmental Variability of Acoustic Predictions and Sonar Performance, Springer, 2002, pp. 611-618..
- A. Das, Shallow ambient noise variability due to distant shipping noise and tide, Appl. Acoust. 72 (9) (2011) 660-664, https://doi.org/10.1016/j. apacoust.2011.03.003. · DOI
- S. Ramji, G. Latha, V. Rajendran, S. Ramakrishnan, Wind dependence of ambient noise in shallow water of Bay of Bengal, Appl. Acoust. 69 (12) (2008) 1294-1298, https://doi.org/10.1016/j.apacoust.2007.09.001. · DOI
- H.B. Ali, Oceanographic variability in shallow-water acoustics and the dual role of the sea bottom, IEEE J. Oceanic Eng. 18 (1) (1993) 31-41, https://doi.org/ 10.1109/48.211498. · DOI
- E. Kozaczka, G. Grelowska, Shipping low frequency noise and its propagation in shallow water, Acta Phys. Polon. A 119 (6A) (2011) 1009-1012. doi:10.12693/APhysPolA.119.1009.. · DOI
- Y. Bengio, A. Courville, P. Vincent, Representation learning: a review and new perspectives, IEEE Trans. Pattern Anal. Mach. Intell. 35 (8) (2013) 1798-1828.
- H. Gimse, Classification of marine vessels using sonar data and a neural network, Master's thesis, NTNU (2017).
- Z. Wei, Y. Ju, M. Song, A method of underwater acoustic signal classification based on deep neural network, in: 2018 5th International Conference on Information Science and Control Engineering (ICISCE), IEEE, 2018, pp. 46-50.
- J. Ren, Z. Huang, C. Li, X. Guo, J. Xu, Feature analysis of passive underwater targets recognition based on deep neural network, in: OCEANS 2019-Marseille, IEEE, 2019, pp. 1-5..
- T. N. Sainath, B. Kingsbury, A. R. Mohamed, B. Ramabhadran, Learning filter banks within a deep neural network framework, in: 2013 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2013 -Proceedings, IEEE, 2013, pp. 297-302. doi:10.1109/ASRU.2013.6707746.. · DOI
- Z. Tüske, P. Golik, R. Schlüter, H. Ney, Acoustic modeling with deep neural networks using raw time signal for LVCSR, in: Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH, No. September, 2014, pp. 890-894..
- N. Zeghidour, N. Usunier, I. Kokkinos, T. Schaiz, G. Synnaeve, E. Dupoux, Learning filterbanks from raw speech for phone recognition, in: ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 2018-April, 2018, pp. 5509-5513. arXiv: arXiv:1711.01161v2, doi:10.1109/ICASSP.2018.8462015.. · DOI
- E. Cakir, E.C. Ozan, T. Virtanen, Filterbank learning for deep neural network based polyphonic sound event detection, in: Proceedings of the International Joint Conference on Neural Networks, vol. 2016-October, IEEE, 2016, pp. 3399-3406. doi:10.1109/IJCNN.2016.7727634.. · DOI
- J. Lee, T. Kim, J. Park, J. Nam, Raw waveform-based audio classification using sample-level cnn architectures, arXiv preprint arXiv:1712.00866.. · arXiv
- T.F. Quatieri, Discrete-time Speech Signal Processing: Principles and Practice, Pearson Education India, 2006.
- Y. Hoshen, R.J. Weiss, K.W. Wilson, Speech acoustic modeling from raw multichannel waveforms, in: 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2015, pp. 4624-4628.
- M. Ravanelli, Y. Bengio, Speaker recognition from raw waveform with sincnet, in: 2018 IEEE Spoken Language Technology Workshop (SLT), IEEE, 2018, pp. 1021-1028..
- P. Agrawal, S. Ganapathy, Unsupervised raw waveform representation learning for asr, Proc. Interspeech 2019 (2019) 3451-3455.
- E. Loweimi, P. Bell, S. Renals, On learning interpretable cnns with parametric modulated kernel-based filters, in: Proc. Interspeech, 2019, pp. 3480-3484..
- H. Khan, B. Yener, Learning filter widths of spectral decompositions with wavelets, in: Advances in Neural Information Processing Systems, 2018, pp. 4601-4612..
- S. Tucker, G.J. Brown, Classification of transient sonar sounds using perceptually motivated features, IEEE J. Oceanic Eng. 30 (3) (2005) 588-600.
- X.H. Zhang, C.Y. Kang, Z.J. Xia, Recognition of radiated noises of ships using auditory features and support vector machines, Advances in Neural Networks -Isnn 2005, Pt 2, Proceedings 3497 (60472108), 2005, pp. 387-392..
- H. Traunmüller, On vowels: Spectral features, related aspects of production and sociophonetic dimensions, Ph.D. thesis, Stockholms universitet, Stockholm (1983)..
- R.D. Patterson, K. Robinson, J. Holdsworth, D. McKeown, C. Zhang, M. Allerhand, Complex sounds and auditory images, in: Auditory Physiology and Perception, Elsevier, 1992, pp. 429-446.
- A.L. Maas, A.Y. Hannun, A.Y. Ng, Rectifier nonlinearities improve neural network acoustic models, in: International Conference on Machine Learning, vol. 30, 2013..
- S. Ioffe, C. Szegedy, Batch normalization: Accelerating deep network training by reducing internal covariate shift, in: International Conference on Machine Learning, 2015, pp. 448-456..
- D. P. Kingma, J. Ba, Adam: a method for stochastic optimization, arXiv preprint arXiv:1412.6980.. · arXiv
- C. Li, Z. Huang, J. Xu, Y. Yan, Underwater target classification using deep learning, in: OCEANS 2018 MTS/IEEE Charleston, 2018, pp. 1-5. doi:10.1109/ OCEANS.2018.8604906.. · DOI
- Y. Yu, X. Cao, X. Zhang, Underwater target classification using deep neural network, in: 2018 OCEANS -MTS/IEEE Kobe Techno-Oceans (OTO), 2018, pp. 1-5. doi:10.1109/OCEANSKOBE.2018.8559111.. · DOI
Cited by
Sorted by citations
Domingos, L. C. F., Santos, P. E., Skelton, P. S. M., Brinkworth, R. S. A., & Sammut, K. (2022) · Sensors · doi:10.3390/s22062181 · Cited by 93 · see citing works
“Novel methods are successfully employing a combination of classic signal processing methods (such as STFT) with ML architectures [125] as a strategy to compensate the signal complexity by selecting its most distinctive features for the classification procedure.”
BackgroundMethodInfluential
- Authors
- Kamal, S., Satheesh Chandran, C., & Supriya, M. H.
- Affiliations
- Cochin University of Science and Technology (IN)
- Corresponding author
- Kamal, Suraj
- ORCID
- Kamal, Suraj
- Venue
- Engineering Science and Technology, an International Journal
- Publisher
- Elsevier BV
- Published
- 2021-08
- Volume
- 24
- Issue
- 4
- Pages
- 860-871
- ISSN
- 2215-0986 (ISSN-L)
- OpenAlex
- W3135247824
- Cited by
- 33
- References
- 2
- Citations by year
- 2026: 5 · 2025: 8 · 2024: 10 · 2023: 6 · 2022: 4
- FWCI
- 2.41
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Engineering Science and Technology an International Journal (publishedVersion, cc-by-nc-nd)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Cochin University of Science and Technology
- Topics
- Physical Sciences › Earth and Planetary Sciences › Oceanography › Underwater Acoustics Research
- Physical Sciences › Computer Science › Signal Processing › Music and Audio Processing
- Physical Sciences › Engineering › Ocean Engineering › Underwater Vehicles and Communication Systems
- Keywords
- passive sonar, automatic target recognition, shallow water acoustics, convolutional LSTM, bidirectional LSTM, attention mechanism, end-to-end learning, underwater target classification
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
