Classifying Anuran Call Spectrograms with Correlation Filters

 
 
 
  • Abstract
  • Keywords
  • References
  • PDF
  • Abstract


    A method to classify anurans based on the spectrogram representations of their call vocalizations is presented. The spectrogram representations have distinctive patterns that may be used to differentiate between species. As such, they can be treated as input images to advanced correlation filters frequency employed in human biometrics applications. A type of correlation filter that has been successfully implemented for face and fingerprint biometrics is considered here. In order to obtain clear spectrograms with distinguishing features, careful selection of spectrogram parameters is performed. To demonstrate this approach, two species of anurans commonly found in Southeast Asia are classified showing that the accuracy rate is dependent on the number of call-prints used to construct the correlation filter templates.


  • References


      [1] Gibbs JP, Rouhani S, & Shams L (2017), Frog and toad habitat occupancy across a Polychlorinated Biphenyl (PCB) contamination gradient, Journal of Herpetology, 51(2), 209-214.

      [2] Huang CJ, Yang YJ, Yang DX & Chen YJ (2009), Frog classification using machine learning techniques, Expert Systems with Applications, 36(2), 3737-3743.

      [3] Gingras B. & Fitch, WT (2013), A three-parameter model for classifying anurans into four genera based on advertisement calls, The Journal of the Acoustical Society of America, 133(1), 547-559.

      [4] Acevedo MA, Corrada-Bravo CJ, Corrada-Bravo H, Villanueva-Rivera LJ & Aide TM (2009), Automated classification of bird and amphibian calls using machine learning: A comparison of methods. Ecological Informatics, 4(4), 206-214,

      [5] Noda JJ, Travieso CM & Sánchez-Rodríguez D (2016), Methodology for automatic bioacoustic classification of anurans based on feature fusion. Expert Systems with Applications, 50, 100-106.

      [6] Vaca-Castaño G. & Rodriguez D, Using syllabic Mel Cepstrum features and k-nearest neighbors to identify anurans and bird species, Proceedings of IEEE Workshop on Signal Processing Systems (SIPS), (2010), 466-471.

      [7] Colonna J., Peet T., Ferreira CA, Jorge AM, Gomes EF & Gama, J, Automatic classification of anuran sounds using convolutional neural networks, Proceedings of the Ninth International C* Conference on Computer Science & Software Engineering, (2016), 73-78.

      [8] Gingras B. & Fitch WT (2013), A three-parameter model for classifying anurans into four genera based on advertisement calls. The Journal of the Acoustical Society of America, 133(1), 547-559.

      [9] Chen WP, Chen SS, Lin CC, Chen YZ & Lin WC (2012), Automatic recognition of frog calls using a multi-stage average spectrum. Computers & Mathematics with Applications, 64(5),1270-1281.

      [10] Dennis J, Tran HD & Li H (2011), Spectrogram image feature for sound event classification in mismatched conditions. IEEE Signal Processing Letters, 18(2), 130-133.

      [11] Grigg G, Taylor, A, Mc Callum, H. & Watson G, Monitoring frog communities: An application of machine learning, Proceedings of Eighth Innovative Applications of Artificial Intelligence Conference, Portland Oregon, (1996), 1564-1569.

      [12] Xie J, Towsey M., Zhang J., Dong X. and Roe P, Application of image processing techniques for frog call classification. Proceedings of IEEE International Conference on Image Processing (ICIP), (2015), 4190-4194.

      [13] Wang Q, Alfalou A. & Brosseau C (2017), New perspectives in face correlation research: A tutorial. Advances in Optics and Photonics, 9(1), 1-78.


 

View

Download

Article ID: 22879
 
DOI: 10.14419/ijet.v7i4.16.22879




Copyright © 2012-2015 Science Publishing Corporation Inc. All rights reserved.