EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Improving Greater Caribbean manatee vocalization detection across habitats using neural networks

Eric A Ramos, Amit Galor, Michael Faran iD, Michael M Mishelashvili, Nataly Castelblanco-Martinez iD, Marisa Tellez, Beth Brady iD

PLOS One · 2026

DOI: 10.1371/journal.pone.0341561 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, 2026. Verfasst von Eric A Ramos und 6 weiteren Autor:innen. Beteiligte Einrichtung: Mote Marine Laboratory, Sarasota, Florida, United States of America. Hauptthemen laut MEDLINE: Vocalization, Animal, Neural Networks, Computer, Ecosystem, Trichechus manatus. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The detection and classification of Greater Caribbean manatee vocalizations ( Trichechus manatus manatus ) present unique challenges due to the complexities of underwater acoustic environments. This study explores the application of neural networks for improving the identification and classification of Greater Caribbean manatee vocalizations, which can provide valuable insights into their behavior and aid in conservation efforts. Utilizing a large dataset of underwater recordings, we trained a known CNN architecture without domain-relevant pretraining to identify and classify Greater Caribbean manatee calls. Our approach combined advanced signal processing techniques such as filtering and normalization with deep learning algorithms to account for the dynamic and noisy conditions of marine environments, employing data augmentation and feature extraction strategies to focus on relevant and informative sound characteristics. The neural network demonstrated promising results, with an overall F1 score of 95.6% on the Wildtracks test dataset, and an F1 score of 64.4% on the Placencia dataset after fine-tuning on less than 10 seconds of vocalizations. This highlights the ability of the model to generalize to novel datasets collected in different regions with vastly different noise profiles. Although there is room for improvement in terms of generalization, these findings represent an advancement in the automated detection and classification of Greater Caribbean manatee vocalizations. This could potentially lead to more effective monitoring of their populations and contribute to the development of improved conservation strategies.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Vocalization, AnimalNeural Networks, ComputerEcosystemTrichechus manatusAnimalsConvolutional Neural NetworksCaribbean RegionSignal Processing, Computer-Assisted

Autor:innen und Institutionen

  • Mote Marine Laboratory, Sarasota, Florida, United States of America.
  • Fundación Internacional para la Naturaleza y la Sustentabilidad, Chetumal, Quintana Roo, Mexico.
  • Deep Voice, Tel Aviv, Israel.
  • El Colegio de la Frontera Sur, Department of Systematic and Aquatic Ecology, Laboratory of Aquatic Mammals, Chetumal, Quintana Roo, Mexico.
  • Crocodile Research Coalition, Stann Creek, Belize.
  • Save the Manatee Club, Longwood, Florida.

ORCID-Profile: Michael Faran, Nataly Castelblanco-Martinez, Beth Brady

Zitieren

APA

Ramos, E. A., Galor, A., Faran, M., Mishelashvili, M. M., Castelblanco-Martinez, N., Tellez, M., & Brady, B. (2026). Improving Greater Caribbean manatee vocalization detection across habitats using neural networks. PLOS One. https://doi.org/10.1371/journal.pone.0341561

Vancouver

Ramos EA, Galor A, Faran M, Mishelashvili MM, Castelblanco-Martinez N, Tellez M, et al. Improving Greater Caribbean manatee vocalization detection across habitats using neural networks. PLOS One. 2026. doi:10.1371/journal.pone.0341561

BibTeX
@article{Ramos2026,
  author = {Ramos, Eric A and Galor, Amit and Faran, Michael and Mishelashvili, Michael M and Castelblanco-Martinez, Nataly and Tellez, Marisa and Brady, Beth},
  title = {Improving Greater Caribbean manatee vocalization detection across habitats using neural networks},
  journal = {PLOS One},
  year = {2026},
  doi = {10.1371/journal.pone.0341561}
}
Kostenlos starten →