EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Spatial validation of acoustic individual identification models without ground truths: a case study with the cao-vit gibbon population

Paul Best, Angela Dassow, Arik Kershenbaum, Tho Duc Nguyen, Megan Pogson, Aishwarya Maheshwari, Ricard Marxer

PeerJ · 02.03.2026 · 14 · e20655

DOI: 10.7717/peerj.20655 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PeerJ, online erschienen am 02.03.2026. Verfasst von Paul Best und 6 weiteren Autor:innen. Beteiligte Einrichtung: Laboratoire d'Informatique des Systèmes, Université de Toulon, Toulon, France. Hauptthemen laut MEDLINE: Vocalization, Animal, Acoustics, Hylobates. Die Arbeit nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Wisconsin Space Grant Consortium—Research Infrastructure Program, Arcus Foundation, Agence Nationale de la Recherche, Institute of Convergence ILCB supported by grants from France 2030.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Technological progress has made bioacoustics an important tool for research in the ecology and behaviour of sound producing animals. Using an array of synchronised autonomous recorders, we can localise vocalising animals, and for certain species, computational models can acoustically identify individuals (AIID). Knowing both the precise location and identity of vocalising animals enables a more detailed interpretation of long-term bioacoustic data, but assessing the reliability of AIID models is often difficult, especially for populations that evolve over time. Annotated ground truth labels in test sets are commonly used, but they are often limited in size, and there can be a mismatch with the application data (for instance in case of a change in recording system). Here, we formalise a methodology to evaluate AIID models based on localised predictions, thus bypassing the need for ground truth labels. We demonstrate it on a case study with the critically endangered cao-vit gibbons ( Nomascus nasutus ). Using deep-learning, we develop an AIID model for male cao-vit gibbons. Then, we estimate its performance without any ground truths, using a new framework that relies on assumptions of territoriality. Empirical tests with simulated data show that this approach to no-ground-truth AIID evaluation is fairly reliable (0.05 of root mean square error between estimated and real accuracy for models with less than 30% of error rate), and specific flaws of performance estimation are described according to specific types of AIID errors. With this article, we demonstrate how spatialised data might help in the evaluation of AIID models for territorial species, both theoretically, and in practice with the cao-vit gibbon population.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Vocalization, AnimalAcousticsHylobatesAnimalsMaleReproducibility of ResultsDeep LearningEndangered Species

Stichwörter der Autor:innen

Acoustic Identification of Individuals (AIID)BioacousticsDeep learningGibbons

Autor:innen und Institutionen

  • Laboratoire d'Informatique des Systèmes, Université de Toulon, Toulon, France.
  • Institue of Language Communication and the Brain, Marseille, France.
  • Biology Department, Carthage College, Kenosha, WI, United States of America.
  • Girton College and Department of Zoology, University of Cambridge, Cambridge, United Kingdom.
  • Vietnam Programme, Fauna & Flora, Hanoi, Vietnam.
  • Graduate University of Science and Technology, Vietnam Academy of Science and Technology, Hanoi, Vietnam.

Förderung

  • Wisconsin Space Grant Consortium—Research Infrastructure Program
  • Arcus Foundation, Agence Nationale de la Recherche
  • Institute of Convergence ILCB supported by grants from France 2030
  • Excellence Initiative of Aix-Marseille University

Zitieren

APA

Best, P., Dassow, A., Kershenbaum, A., Nguyen, T. D., Pogson, M., Maheshwari, A., & Marxer, R. (2026). Spatial validation of acoustic individual identification models without ground truths: a case study with the cao-vit gibbon population. PeerJ, 14, e20655. https://doi.org/10.7717/peerj.20655

Vancouver

Best P, Dassow A, Kershenbaum A, Nguyen TD, Pogson M, Maheshwari A, et al. Spatial validation of acoustic individual identification models without ground truths: a case study with the cao-vit gibbon population. PeerJ. 2026;14:e20655. doi:10.7717/peerj.20655

BibTeX
@article{Best2026,
  author = {Best, Paul and Dassow, Angela and Kershenbaum, Arik and Nguyen, Tho Duc and Pogson, Megan and Maheshwari, Aishwarya and Marxer, Ricard},
  title = {Spatial validation of acoustic individual identification models without ground truths: a case study with the cao-vit gibbon population},
  journal = {PeerJ},
  year = {2026},
  volume = {14},
  pages = {e20655},
  doi = {10.7717/peerj.20655}
}
Kostenlos starten →