The Journal of the Acoustical Society of Korea (한국음향학회지)

The Acoustical Society of Korea (한국음향학회)
권호 표지
DOI QR코드

DOI QR Code

Abnormal sonar signal detection using recurrent neural network and vector quantization

순환신경망과 벡터 양자화를 이용한 비정상 소나 신호 탐지

  • 이기배 (제주대학교 해양시스템공학과) ;
  • 고건혁 (제주대학교 해양시스템공학과) ;
  • 이종현 (제주대학교 해양시스템공학과)
  • Received : 2023.07.03
  • Accepted : 2023.10.16
  • Published : 2023.11.30
Download PDF
⟨ Previous Next ⟩

Abstract

Passive sonar signals mainly contain both normal and abnormal signals. The abnormal signals mixed with normal signals are primarily detected using an AutoEncoder (AE) that learns only normal signals. However, existing AEs may perform inaccurate detection by reconstructing distorted normal signals from mixed signal. To address these limitations, we propose an abnormal signal detection model based on a Recurrent Neural Network (RNN) and vector quantization. The proposed model generates a codebook representing the learned latent vectors and detects abnormal signals more accurately through the proposed search process of code vectors. In experiments using publicly available underwater acoustic data, the AE and Variational AutoEncoder (VAE) using the proposed method showed at least a 2.4 % improvement in the detection performance and at least a 9.2 % improvement in the extraction performance for abnormal signals than the existing models.

수동소나 신호에는 정상신호와 비정상 신호가 같이 존재하는 경우가 대부분이다. 정상신호와 혼재된 비정상 신호는 주로 정상신호만을 학습하는 오토인코더를 이용하여 탐지된다. 하지만 기존의 오토인코더는 혼재된 신호로부터 왜곡된 정상신호를 복원하므로 부정확한 탐지를 수행할 수 있다. 이러한 한계를 개선하고자, 본 논문에서는 순환신경망과 벡터 양자화 기반의 비정상 신호 탐지 모델을 제안한다. 제안된 모델은 학습된 잠재벡터들을 대표하는 코드 북을 생성하고, 제안된 코드벡터의 탐색을 통해 보다 정확하게 비정상 신호를 탐지한다. 공개된 수중 음향 데이터를 이용한 실험에서 제안된 기법이 적용된 오토인코더와 변이형 오토인코더는 기존 모델에 비해 최소 2.4 % 향상된 탐지 성능과 최소 9.2 % 높은 비정상 신호 추출 성능을 보였다.

Keywords

Acknowledgement

이 논문은 2022학년도 제주대학교 교원성과지원 사업에 의하여 연구되었음.

References

  1. L. Ruff, J. R. Kauffmann, R. A. Vandermeulen, G. Monatavon, W. Samek, M. Kloft, T. G. Dietterich, and K. R. Muller, "A unifying review of deep and shallow anomaly detection," Proc. IEEE, 109, 756-795 (2021). https://doi.org/10.1109/JPROC.2021.3052449
  2. K. Lee, C. H. Lee, and J. Lee, "Semi-supervised anomaly detection algorithm using probabilistic labeling (SAD-PL)," IEEE Access, 9, 142972-142981 (2021). https://doi.org/10.1109/ACCESS.2021.3120710
  3. K. Lee, G. H. Ko, and C. H. Lee, "Encoding and decoding-based normalizing flows for image anomaly localization," Electron. Lett. 59, e212829 (2023).
  4. W. Hu, J. Gao, B. Li, O. Wu, J. Du, and S. Maybank, "Anomaly detection using local kernel density estimation and context-based regression," IEEE Trans. Knowl. Data Eng. 32, 218-233 (2020). https://doi.org/10.1109/TKDE.2018.2882404
  5. L. Huang, X. Nguyen, M. Garofalakis, M. I. Jordan, A. Joseph, and N. Taft, "In-network PCA and anomaly detection," Proc. NIPS, 617-624 (2006).
  6. B. Scholkopf, R. Willianson, A. Smola, J. S. Taylor, and J. Platt, "Support vector method for novelty detection," Proc. NIPS. 582-588 (1999).
  7. D. M. Tax and R. P. Duin, "Support vector data description," Machine learning, 59, 45-66 (2004). https://doi.org/10.1023/B:MACH.0000008084.60811.49
  8. C. H. Lee and K. Lee, "Semi-supervised anomaly detection algorithm based on KL divergence (SADKL)," Proc. ADIX, 118-123 (2023).
  9. L. Ruff, R. A. Vandermeulen, N. Gornits, L. Deecke, S. A. Siddiqui, A. Binder, E. Muller, and M. Kloft, "Deep one-class classification," Proc. ICML, 4393-4402 (2018).
  10. K. Lee, G. H. Ko, and C. H. Lee, "Online anomaly detection algorithm based on deep support vector data description using incremental centroid update," J. Acoust. Soc. Kr. 41, 199-209 (2022).
  11. M. Sakurada and T. Yairi, "Anomaly detection using autoencoders with nonlinear dimensionality reduction," Proc. MLSDA, 4-11 (2014).
  12. K. Lee and C. H. Lee, "Abnormal signal detection based on parallel autoencoders," J. Acoust. Soc. Kr. 40, 337-346 (2021).
  13. Z. Mnasri, S. Rovetta, and F. Masulli, "Anomalous sound event detection: A survey of machine learning based methods and applications," Multimed. Tools App. 81, 5537-5586 (2022). https://doi.org/10.1007/s11042-021-11817-9
  14. E. Marchi, F. Vesperini, S. Squartini, and B. Schuller, "Deep recurrent neural networks-based autoencoders for acoustic novelty detection," Comput. Intell. Neurosci. 2017, 1-14 (2017). https://doi.org/10.1155/2017/4694860
  15. K. Suefusa, T. Nishida, H. Purohit, R. Tanabe, T. Endo, and Y. Kawaguchi, "Anomalous sound detection based on interpolation deep neural network," Proc. ICASSP, 271-275 (2020).
  16. Y. Tagawa, R. Maskeliunas, and R. Damasevicius, "Acoustic anomaly detection of mechanical failures in noisy real-life factory environments," Electronics, 10, 2329 (2021).
  17. P. J. Pereira, G. Coelho, A. Ribeiro, L. M. Matos, E. C. Nunes, A. Ferreira, A. Pilastri, and P. Cortes, "Using deep autoencoders for in-vehicle audio anomaly detection," Procedia. Comput. Sci. 192, 298-307 (2021). https://doi.org/10.1016/j.procs.2021.08.031
  18. D. Dehaene, O. Frigo, S. Combrexelle, and P. Eline, "Iterative energy-based projection on a normal data manifold for anomaly localization," Proc. ICLR, 1-17 (2020).
  19. A. Oord, O. Vinyals, and K. Kavukcuoglu, "Neural discrete representation learning," Proc. NIPS, 1-17 (2017).
  20. W. Williams, S. Ringer, T. Ash, J. Hughes, D. MacLeod, and J. Dougherty, "Hierarchical quantized autoencoders," Proc. NIPS, 1-12 (2020).
  21. A New Mindset for the Army: Silent Running, https://www.c4isrnet.com/opinion/2019/12/06/a-new-mindset-for-the-army-silent-running/, (Last viewed September 1, 2023).
  22. S. Kim, S. K. Jung, D. Kang, M. Kim, and S. Cho, "Application of the artificial intelligence for automatic detection of shipping noise in shallow-water" (in Korean), J. Acoust. Soc. Kr. 39, 279-285 (2020).
  23. K. M. Park and D. Kim, "Preprocessing performance of convolutional neural networks according to characteristic of underwater targets" (in Korean), J. Acoust. Soc. Kr. 41, 629-636 (2022).
  24. Y. C. Jung, B. U. Kim, S. K. An, W. J. Seong, K. H. Lee, and J. Y. Hahn, "An algorithm for submarine passive sonar simulator" (in Korean), J. Acoust. Soc. Kr. 32, 472-483 (2013). https://doi.org/10.7776/ASK.2013.32.6.472
  25. L. E. Kinsler, A. R. Frey, A. B. Coppens, and J. V. Sanders, Fundamentals of Acoustics (John Wiley & Sons, New Jersey, 1999), Chap. 15.
  26. D. P. Kingma and M. Welling, "An introduction to variational autoencoders," Found. Trends Mach. Learn. 12, 1-85 (2019). https://doi.org/10.1561/2200000068
  27. D. S. Dominguez, S. T. Guijarro, A. C. Lopez, and A. P. Gimenez, "ShipEar: An underwater vessel noise database," Appl. Acoust. 113, 64-69 (2016). https://doi.org/10.1016/j.apacoust.2016.06.008
  28. K. J. V. Raposa, G. Scowcroft, J. H. Miller, D. R. Ketten, and A. N. Popper, "Discovery of sound in the sea: Resources for educators, students, the public, and policymakers," in Handbook of The Effects of Noise on Aquatic Life 2, edited by A. N. Popper and A. Hawkins (Springer, New York, 2016).
  29. Y. Koizumi, Y. Kawaguchi, K. Imoto, T. Nakamura, Y. Nikaido, R. Tanabe, H. Purohit, K. Suefusa, T. Endo, M. Yasuda, and N. Harada, "Description and discussion on DCASE2020 challenge task2: unsupervised anomalous sound detection for machine condition monitoring," Proc. DCASE, 81-85 (2020).
  30. H. Rezatofighi, N. Tsoi, J. Y. Gwak, A. Sadeghian, I. Reid, and S. Savarese, "Generalized intersection over union: A metric and a loss for bounding box regression," Proc. CVPR, 658-666 (2019).