Designing a BirdNET classifier for high wind detection in passive acoustic recordings to support wildlife monitoring
The Journal of the Acoustical Society of America · 01.06.2025 · 157 (6) · 4502-4512
DOI: 10.1121/10.0036887 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Journal of the Acoustical Society of America, online erschienen am 20.06.2025. Verfasst von Danielle T Fradet und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biological Sciences, University of New Hampshire, Durham, New Hampshire 03824, USA. Hauptthemen laut MEDLINE: Wind, Vocalization, Animal, Acoustics, Spheniscidae. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 65 Literaturangaben. Gefördert von: NSF, National Science Foundation.
Abstract (Originalsprache)
Passive acoustic monitoring (PAM) is a powerful tool for ecological research, but recordings can be compromised by background noise such as wind. Addressing wind noise (e.g., clipping and masking) in bioacoustic data remains a challenge, especially as climate change is predicted to increase wind speeds, particularly near the poles. Adélie penguins (Pygoscelis adeliae), key indicators of the Antarctic ecosystem, are well-suited for PAM, where large-scale monitoring could assess climate-driven population changes—if wind noise is managed effectively. In this study, the convolutional neural network, BirdNET, inversely identifies unwanted sounds in Adélie penguin colony recordings. Multiple custom models were developed in which the background nontarget noise was Adélie vocalizations, and wind conditions (low, medium, and high) were the target classes. The best-performing model achieved an F-score of 0.43 and accuracy of 0.53. The high wind class within this model had a precision of 0.76 and recall of 0.94. A six-step workflow is presented for creating custom BirdNET models, evaluating their performance and determining an optimal confidence threshold prior to model application on an entire dataset. By automating unwanted sound detection, this approach enables researchers to efficiently identify and remove affected files, streamline data cleaning, and focus on recordings of interest for further analysis.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Biological Sciences, University of New Hampshire, Durham, New Hampshire 03824, USA.
- Center for Acoustics Research and Education, University of New Hampshire, Durham, New Hampshire 03824, USA.
- Institute of Marine Science, University of California, Santa Cruz, Santa Cruz, California 95064, USA.
ORCID-Profile: Danielle T Fradet, Easton R White, Laura N Kloepper
Förderung
- NSF
- National Science Foundation
Zitieren
APA
Fradet, D. T., Cimino, M. A., White, E. R., & Kloepper, L. N. (2025). Designing a BirdNET classifier for high wind detection in passive acoustic recordings to support wildlife monitoring. The Journal of the Acoustical Society of America, 157(6), 4502-4512. https://doi.org/10.1121/10.0036887
Vancouver
Fradet DT, Cimino MA, White ER, Kloepper LN. Designing a BirdNET classifier for high wind detection in passive acoustic recordings to support wildlife monitoring. The Journal of the Acoustical Society of America. 2025;157(6):4502-4512. doi:10.1121/10.0036887
BibTeX
@article{Fradet2025,
author = {Fradet, Danielle T and Cimino, Megan A and White, Easton R and Kloepper, Laura N},
title = {Designing a BirdNET classifier for high wind detection in passive acoustic recordings to support wildlife monitoring},
journal = {The Journal of the Acoustical Society of America},
year = {2025},
volume = {157},
number = {6},
pages = {4502--4512},
doi = {10.1121/10.0036887}
}
EUVIMED