Journal of the Institute of Convergence Signal Processing (융합신호처리학회논문지)

The Korea Institute of Convergence Signal Processing (한국융합신호처리학회)
권호 표지

Whale Sound Reconstruction using MFCC and L2-norm Minimization

MFCC와 L2-norm 최소화를 이용한 고래소리의 재생

  • Received : 2018.12.04
  • Accepted : 2018.12.24
  • Published : 2018.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Underwater transient signals are complex, variable and nonlinear, resulting in a difficulty in accurate modeling with reference patterns. We analyze one type of underwater transient signals, in the form of whale sounds, using the MFCC(Mel-Frequency Cepstral Constant) and synthesize them from the MFCC and the weighted $L_2$-norm minimization techniques. The whales in this experiments are Humpback whales, Right whales, Blue whales, Gray whales, Minke whales. The 20th MFCC coefficients are extracted from the original signals using the MATLAB programming and reconstructed using the weighted $L_2$-norm minimization with the inverse MFCC. Finally, we could find the optimum weighted factor, 3~4 for reconstruction of whale sounds.

수중에서의 일시적인 신호는 복잡하고, 변화가 심하며, 비선형적이므로 신호의 패턴을 정확히 모델링하기 어렵다. 본 논문에서는 수중 신호 중 하나인 고래 소리를 선택하여 음성분석 기법에 많이 사용하는 Cepstral 분석에 의한 MFCC 추출법을 이용하여 분석하였고, MFCC와 $L_2$-norm 최소화 기법을 이용하여 고래소리를 재생하였다 실험 분석에 사용된 고래의 종류는 혹등고래(Humpback whale), 참고래(Right whale), 대왕고래(Blue whale), 귀신고래(Gray whale), 밍크고래(Minke whale) 등 5종으로서 과거 한반도 동해안에 출몰한 적이 있는 고래들이다. 원본 고래소리에서 MATLAB프로그래밍을 이용하여 20차 MFCC계수들을 추출한 후 이를 가중 $L_2$-norm 최소화를 이용한 MFCC역변환을 통해 재생한다. 최종적으로 가중치가 3~4의 값에서 고래소리 재생이 가장 적합함을 알 수 있었다.

Keywords

References

  1. Walter M. X. Zimmer, "Passive Acoustic Monitoring of Cetaceans", Cambridge University Press, 2011.
  2. S.J. Park, J.W. Hong, "A Study on the Improvement of Legal System for the Revitalization of Korea's Marine Tourism", J. of the Korean Society of Marine Environment & Safety, Vol. 18, No. 2, pp. 131-138, 20123. https://doi.org/10.7837/kosomes.2012.18.2.131
  3. T.G. Lim, K.S. Bae, C.S. Hwang, H.U. Lee, "Classification of Underwater Transient Signals using MFCC Feature Vector", J. of Korea Communication Association, Vol. 32, No. 8, pp. 675-679, 2007.
  4. T.G. Lim, I.H. Kim, T.H. Kim, K.S. Bae, "Frame Based Classification of Underwater Transient Signal using MFCC Feature Vector and Neural Network", The Proceeding of Korea Electronics Association 2008, Vol. 31, No. 1, pp. 883-884, 2008.
  5. J.G. Jung, J.H. Park,D.W. Kim, C.S. Hwang, "Feature Extraction and Classification of Underwater Transient Signal using MFCC and Wavelet Packet Based on Entropy", The Proceeding of Korea Univ.-Industry Tech Association, Pp 781-784, Spring of 2009.
  6. J.H. Kim, T.H. Bok, D.G. Paeng, J.H. Bae, C.H. Lee, S.G. Kim, "Classification of Transient Signal in Ocean Background Noise using Bayesian Classifier", J. of The Korean Society of Ocean Engineers, Vol. 26, No. 4, pp. 57-63, 2012.
  7. D. Cazu, R. Lefort, J. Bonnel, J. Krywyk, "Bi-class of Humpback Whale Sound Units Against Complex Background Noise With Deep Convolution Neural Network", Workshop Track-ICLR 2017, pp. 1-7, 2017.
  8. Sherin B.M., Dr. Supriya M.H., "WOA based Selection and Parameter Optimization of SVM Kernel Function for Underwater Target Classification", International J. of Advanced Research in Computer Science, Vol. 8, No. 3, pp. 223-226, 2017. https://doi.org/10.26483/ijarcs.v8i8.4622
  9. Gang Min, Xiongwei Zhang, Jibin Yang, Xia Zou, "Speech Reconstruction from Mel-frequency Cepstral Coefficients via L1-norm Minimization", MMSP'15, Oct. 2015, Xiamen, China
  10. Xavier Serra and Julius Smith, "Spectral Modeling Synthesis: A Sound Analysis/Synthesis System Based on a Deterministic plus Stochastic Decomposition" Computer Music Journal, vol.14, No 4, pp.12-24, 1990. https://doi.org/10.2307/3680788