An empirical mode decomposition based hidden Markov model approach for detection of Bryde's whale pulse calls

Olayinka O Ogundile; Ayinde M Usman; Daniel J J Versfeld

doi:10.1121/10.0000717

An empirical mode decomposition based hidden Markov model approach for detection of Bryde's whale pulse calls

J Acoust Soc Am. 2020 Feb;147(2):EL125. doi: 10.1121/10.0000717.

Authors

Olayinka O Ogundile¹, Ayinde M Usman¹, Daniel J J Versfeld¹

Affiliation

¹ Department of Electrical and Electronic Engineering, Stellenbosch University, Stellenbosch, South Africa.

PMID: 32113315
DOI: 10.1121/10.0000717

Abstract

This letter proposes an empirical mode decomposition (EMD) based hidden Markov model (HMM) approach for the detection of mysticetes' pulse calls such as the Bryde's whales. The HMM detection capabilities depend on the deployed feature extraction (FE) technique. The EMD is proposed as a performance efficient alternative to the popular Mel-scale frequency cepstral coefficient (MFCC) and linear predictive coefficient (LPC) FE techniques. The amplitude modulation-frequency modulation components derived from the EMD process are modified to form feature vectors for the HMM. Also, the ensemble EMD (EEMD) is adapted in a similar way as the EMD. These proposed EMD-HMM and EEMD-HMM approaches achieved better performance in comparison to the MFCC-HMM and LPC-HMM approaches.

Publication types

Research Support, Non-U.S. Gov't