A practical guide for generating unsupervised, spectrogram-based latent space representations of animal vocalizations
A practical guide for generating unsupervised, spectrogram-based latent space representations of animal vocalizations
ABSTRACT The manual detection, analysis, and classification of animal vocalizations in acoustic recordings is laborious and requires expert knowledge. Hence, there is a need for objective, generalizable methods that detect underlying patterns in these data, categorize sounds into distinct groups, and quantify similarities between them. Among all computational methods that have been proposed to accomplish this, neighborhood-based dimensionality reduction of spectrograms to produce a latent-space representation of calls stands out for its conceptual simplicity and effectiveness. Using a dataset of manually annotated meerkat (Suricata suricatta) vocalizations, we demonstrate how this method can be used to obtain meaningful latent space representations that reflect the established taxonomy of call types. We analyze strengths and weaknesses of the proposed approach, give recommendations for its usage and show application examples, such as the classification of ambiguous calls and the detection of mislabeled calls. All analyses are accompanied by example code to help researchers realize the potential of this method for the study of animal vocalizations.
Thomas Mara、Averly Baptiste、Manser Marta B.、Sainburg Tim、Jensen Frants H.、Demartsev Vlad、Roch Marie A.、Strandburg-Peshkin Ariana
Department for the Ecology of Animal Societies, Max Planck Institute of Animal Behavior||Biology Department, University of Konstanz, Universit?tsstrasse 10Department for the Ecology of Animal Societies, Max Planck Institute of Animal Behavior||Biology Department, University of Konstanz, Universit?tsstrasse 10Kalahari Meerkat Project||Department of Evolutionary Biology and Environmental Studies, University of Z¨1richDepartment of Psychology, University of California San DiegoBiology Department, Woods Hole Oceanographic Institution||Department of Biology, Syracuse UniversityDepartment for the Ecology of Animal Societies, Max Planck Institute of Animal Behavior||Biology Department, University of Konstanz, Universit?tsstrasse 10Department of Computer Science, San Diego State UniversityDepartment for the Ecology of Animal Societies, Max Planck Institute of Animal Behavior||Biology Department, University of Konstanz, Universit?tsstrasse 10||Kalahari Meerkat Project||Centre for the Advanced Study of Collective Behavior, University of Konstanz
动物学生物科学研究方法、生物科学研究技术
Thomas Mara,Averly Baptiste,Manser Marta B.,Sainburg Tim,Jensen Frants H.,Demartsev Vlad,Roch Marie A.,Strandburg-Peshkin Ariana.A practical guide for generating unsupervised, spectrogram-based latent space representations of animal vocalizations[EB/OL].(2025-03-28)[2025-06-08].https://www.biorxiv.org/content/10.1101/2021.12.16.472881.点此复制
评论