Detection of Bat Acoustics Signals Using Voice Activity Detection Techniques with Random Forests Classification

  • Adrian T. RuizEmail author
  • Julian Equihua
  • Santiago Martínez
  • Everardo Robredo
  • Günther Palm
  • Friedhelm Schwenker
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11401)


Bats are indicators for ecosystem health, and therefore the determination of bat activity and species abundance provides essential information for biodiversity research and conservation monitoring. In this study, we propose a computational method for the detection of bat echolocation calls. This method uses feature engineering and consists of a statistical model-based Voice Activity Detector combined with a Random Forests classifier (VAD+RF). Using an open-access library (, we trained and tested the performance of our method and compare it to other existing detection methods. These methods include a detector based on deep neural networks along with other commercial detection systems. To visualize the detector performance over the full range of possible class distributions and misclassification costs, we calculated the Cost Curves and \(F_1\)-measure Curves. Results show that the detecting power of VAD+RF is comparable to methods based on deep learning. Based on the results we give recommendations to improve the future designs of the bat call detector.


Bat echolocation Animal sound detection Random forests Voice activity detection Convolutional neural networks ROC curves 


  1. 1.
    Wildlife Acoustics: Kaleidoscope (2012).
  2. 2.
    Breiman, L.: Random forests. Mach. Learn. 45(1), 5–32 (2001)CrossRefGoogle Scholar
  3. 3.
    Brookes, M.: VOICEBOX: a speech processing toolbox for matlab (2006).
  4. 4.
    Drummond, C., Holte, R.C.: Cost curves: an improved method for visualizing classifier performance. Mach. Learn. 65(1), 95–130 (2006)CrossRefGoogle Scholar
  5. 5.
    Ephraim, Y., Malah, D.: Speech enhancement using a minimum mean-square error log-spectral amplitude estimator. IEEE Trans. Acoust. 33(2), 443–445 (1985)CrossRefGoogle Scholar
  6. 6.
    Fenton, M.B., Bell, G.P.: Recognition of species of insectivorous bats by their echolocation calls. J. Mammal. 62(2), 233–243 (1981)CrossRefGoogle Scholar
  7. 7.
    Jones, G., Jacobs, D.S., Kunz, T.H., Wilig, M.R., Racey, P.A.: Carpe noctem: the importance of bats as bioindicators. Endanger. Species Res. 8(1–2), 93–115 (2009)CrossRefGoogle Scholar
  8. 8.
    Mac Aodha, O., et al.: Bat detective–deep learning tools for bat acoustic signal detection. PLoS Comput. Biol. 14(3), 1–19 (2018)CrossRefGoogle Scholar
  9. 9.
    MATLAB: Version 8.5 (R2015a). The MathWorks Inc., Natick (2015)Google Scholar
  10. 10.
    Ruiz, A.T., Jung, K., Tschapka, M., Schwenker, F., Palm, G.: Automated identification method for detection and classification of neotropical bats. In: 8th International Conference of Pattern Recognition Systems (ICPRS 2017), pp. 1–6, July 2017Google Scholar
  11. 11.
    Skowronski, M.D., Fenton, M.B.: Model-based automated detection of echolocation calls using the link detector. J. Acoust. Soc. Am. 124(1), 328–36 (2008)CrossRefGoogle Scholar
  12. 12.
    Skowronski, M.D., Harris, J.G.: Acoustic detection and classification of microchiroptera using machine learning: lessons learned from automatic speech recognition. J. Acoust. Soc. Am. 119(3), 1817–1833 (2006)CrossRefGoogle Scholar
  13. 13.
    Sohn, J., Kim, N.S., Sung, W.: A statistical model-based voice activity detection. IEEE Signal Process. Lett. 6(1), 1–3 (1999)CrossRefGoogle Scholar
  14. 14.
    Soleymani, R., Granger, E., Fumera, G.: F-measure curves for visualizing classifier performance with imbalanced data. In: Pancioni, L., Schwenker, F., Trentin, E. (eds.) ANNPR 2018. LNCS (LNAI), vol. 11081, pp. 165–177. Springer, Cham (2018). Scholar
  15. 15.
    Szewczak, J.: Sonobat v. 3 (2010)Google Scholar
  16. 16.
    Binary Acoustic Technology: Scan’r v.1.7.7 (2014)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Adrian T. Ruiz
    • 1
    • 2
    Email author
  • Julian Equihua
    • 2
  • Santiago Martínez
    • 2
  • Everardo Robredo
    • 2
  • Günther Palm
    • 1
  • Friedhelm Schwenker
    • 1
  1. 1.Institute of Neural Information ProcessingUlm UniversityUlmGermany
  2. 2.National Commission for the Knowledge and Use of Biodiversity (CONABIO)Mexico, D.F.Mexico

Personalised recommendations