An unsupervised Hidden Markov Model-based system for the detection and classification of blue whale vocalizations off Chile
In this paper, we present an automatic method, without human supervision, for the detection and classification of blue whale vocalizations from passive acoustic monitoring (PAM) data using Hidden Markov Model technology implemented with a state-of-the-art machine learning platform, the Kaldi speech...
Published in: | Annals of Operations Research |
---|---|
Main Authors: | , , , , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Taylor & Francis
2020
|
Subjects: | |
Online Access: | https://doi.org/10.1080/09524622.2018.1563758 https://repositorio.uchile.cl/handle/2250/174698 |
id |
ftunivchile:oai:repositorio.uchile.cl:2250/174698 |
---|---|
record_format |
openpolar |
spelling |
ftunivchile:oai:repositorio.uchile.cl:2250/174698 2023-06-18T03:40:03+02:00 An unsupervised Hidden Markov Model-based system for the detection and classification of blue whale vocalizations off Chile Buchan, Susannah J. Mahú Sinclair, Rodrigo Wuth, Jorge Balcazar Cabrera, Naysa Gutiérrez, Laura Neira, Sergio Becerra Yoma, Néstor 2020 application/pdf https://doi.org/10.1080/09524622.2018.1563758 https://repositorio.uchile.cl/handle/2250/174698 en eng Taylor & Francis Bioacoustics 2020, Vol. 29, No. 2, 140–167 doi:10.1080/09524622.2018.1563758 https://repositorio.uchile.cl/handle/2250/174698 Attribution-NonCommercial-NoDerivs 3.0 Chile http://creativecommons.org/licenses/by-nc-nd/3.0/cl/ Bioacoustics Blue whale vocalizations Unsupervised detection and classification HMM Machine learning Artículo de revista 2020 ftunivchile https://doi.org/10.1080/09524622.2018.1563758 2023-06-03T23:52:05Z In this paper, we present an automatic method, without human supervision, for the detection and classification of blue whale vocalizations from passive acoustic monitoring (PAM) data using Hidden Markov Model technology implemented with a state-of-the-art machine learning platform, the Kaldi speech processing toolkit. 157.5 hours of PAM data were annotated for model training and testing, selected from a dataset collected from the Corcovado Gulf, Chilean Patagonia in 2016. The system obtained produced 85.3% accuracy for detection and classification of a range of different blue whale vocalizations. This system was then validated by comparing its unsupervised detection and classification results with the published results of southeast Pacific blue whale song phrase (‘SEP2’) via spectrogram cross-correlation, involving a dataset collected with a different hydrophone instrument. The proposed system led to a reduction in the root mean square error relative to published results as high as 80% when compared with comparable methods employed elsewhere. This is a significant step in advancing the monitoring of endangered whale populations in this region, which remains poorly covered in terms of PAM and general ocean observation. With further training, testing and validation, this system can be applied to other target signals and regions of the world ocean. Article in Journal/Newspaper Blue whale Universidad de Chile: Repositorio académico Pacific Patagonia Annals of Operations Research 286 1-2 119 146 |
institution |
Open Polar |
collection |
Universidad de Chile: Repositorio académico |
op_collection_id |
ftunivchile |
language |
English |
topic |
Blue whale vocalizations Unsupervised detection and classification HMM Machine learning |
spellingShingle |
Blue whale vocalizations Unsupervised detection and classification HMM Machine learning Buchan, Susannah J. Mahú Sinclair, Rodrigo Wuth, Jorge Balcazar Cabrera, Naysa Gutiérrez, Laura Neira, Sergio Becerra Yoma, Néstor An unsupervised Hidden Markov Model-based system for the detection and classification of blue whale vocalizations off Chile |
topic_facet |
Blue whale vocalizations Unsupervised detection and classification HMM Machine learning |
description |
In this paper, we present an automatic method, without human supervision, for the detection and classification of blue whale vocalizations from passive acoustic monitoring (PAM) data using Hidden Markov Model technology implemented with a state-of-the-art machine learning platform, the Kaldi speech processing toolkit. 157.5 hours of PAM data were annotated for model training and testing, selected from a dataset collected from the Corcovado Gulf, Chilean Patagonia in 2016. The system obtained produced 85.3% accuracy for detection and classification of a range of different blue whale vocalizations. This system was then validated by comparing its unsupervised detection and classification results with the published results of southeast Pacific blue whale song phrase (‘SEP2’) via spectrogram cross-correlation, involving a dataset collected with a different hydrophone instrument. The proposed system led to a reduction in the root mean square error relative to published results as high as 80% when compared with comparable methods employed elsewhere. This is a significant step in advancing the monitoring of endangered whale populations in this region, which remains poorly covered in terms of PAM and general ocean observation. With further training, testing and validation, this system can be applied to other target signals and regions of the world ocean. |
format |
Article in Journal/Newspaper |
author |
Buchan, Susannah J. Mahú Sinclair, Rodrigo Wuth, Jorge Balcazar Cabrera, Naysa Gutiérrez, Laura Neira, Sergio Becerra Yoma, Néstor |
author_facet |
Buchan, Susannah J. Mahú Sinclair, Rodrigo Wuth, Jorge Balcazar Cabrera, Naysa Gutiérrez, Laura Neira, Sergio Becerra Yoma, Néstor |
author_sort |
Buchan, Susannah J. |
title |
An unsupervised Hidden Markov Model-based system for the detection and classification of blue whale vocalizations off Chile |
title_short |
An unsupervised Hidden Markov Model-based system for the detection and classification of blue whale vocalizations off Chile |
title_full |
An unsupervised Hidden Markov Model-based system for the detection and classification of blue whale vocalizations off Chile |
title_fullStr |
An unsupervised Hidden Markov Model-based system for the detection and classification of blue whale vocalizations off Chile |
title_full_unstemmed |
An unsupervised Hidden Markov Model-based system for the detection and classification of blue whale vocalizations off Chile |
title_sort |
unsupervised hidden markov model-based system for the detection and classification of blue whale vocalizations off chile |
publisher |
Taylor & Francis |
publishDate |
2020 |
url |
https://doi.org/10.1080/09524622.2018.1563758 https://repositorio.uchile.cl/handle/2250/174698 |
geographic |
Pacific Patagonia |
geographic_facet |
Pacific Patagonia |
genre |
Blue whale |
genre_facet |
Blue whale |
op_source |
Bioacoustics |
op_relation |
Bioacoustics 2020, Vol. 29, No. 2, 140–167 doi:10.1080/09524622.2018.1563758 https://repositorio.uchile.cl/handle/2250/174698 |
op_rights |
Attribution-NonCommercial-NoDerivs 3.0 Chile http://creativecommons.org/licenses/by-nc-nd/3.0/cl/ |
op_doi |
https://doi.org/10.1080/09524622.2018.1563758 |
container_title |
Annals of Operations Research |
container_volume |
286 |
container_issue |
1-2 |
container_start_page |
119 |
op_container_end_page |
146 |
_version_ |
1769004849310990336 |