Same data, different results? Machine learning approaches in bioacoustics
Abstract Automated acoustic analysis is increasingly used in behavioural ecology, and determining caller identity is a key element for many investigations. However, variability in feature extraction and classification methods limits the comparability of results across species and studies, constraini...
Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-08-01
|
| Series: | Methods in Ecology and Evolution |
| Subjects: | |
| Online Access: | https://doi.org/10.1111/2041-210X.70091 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849467035489140736 |
|---|---|
| author | Kaja Wierucka Derek Murphy Stuart K. Watson Nikola Falk Claudia Fichtel Julian León Stephan T. Leu Peter M. Kappeler Elodie F. Briefer Marta B. Manser Nikhil Phaniraj Marina Scheumann Judith M. Burkart |
| author_facet | Kaja Wierucka Derek Murphy Stuart K. Watson Nikola Falk Claudia Fichtel Julian León Stephan T. Leu Peter M. Kappeler Elodie F. Briefer Marta B. Manser Nikhil Phaniraj Marina Scheumann Judith M. Burkart |
| author_sort | Kaja Wierucka |
| collection | DOAJ |
| description | Abstract Automated acoustic analysis is increasingly used in behavioural ecology, and determining caller identity is a key element for many investigations. However, variability in feature extraction and classification methods limits the comparability of results across species and studies, constraining conclusions we can draw about the ecology and evolution of the groups under study. We investigated the impact of using different feature extraction (spectro‐temporal measurements, linear and Mel‐frequency cepstral coefficients (MFCC), as well as highly comparative time‐series analysis) and classification methods (discriminant function analysis, neural networks, random forests (RF), and support vector machines) on the consistency of caller identity classification accuracy across 16 mammalian datasets. We found that MFCCs and RFs yield consistently reliable results across datasets, facilitating a standardised approach across species that generates directly comparable data. These findings remained consistent across vocalisation sample sizes and number of individuals considered. We offer guidelines for processing and analysing mammalian vocalisations, fostering greater comparability and advancing our understanding of the evolutionary significance of acoustic communication in diverse mammalian species. |
| format | Article |
| id | doaj-art-aaef806a00cb417fbdd4c2bbd29efd80 |
| institution | Kabale University |
| issn | 2041-210X |
| language | English |
| publishDate | 2025-08-01 |
| publisher | Wiley |
| record_format | Article |
| series | Methods in Ecology and Evolution |
| spelling | doaj-art-aaef806a00cb417fbdd4c2bbd29efd802025-08-20T03:39:14ZengWileyMethods in Ecology and Evolution2041-210X2025-08-011681574158610.1111/2041-210X.70091Same data, different results? Machine learning approaches in bioacousticsKaja Wierucka0Derek Murphy1Stuart K. Watson2Nikola Falk3Claudia Fichtel4Julian León5Stephan T. Leu6Peter M. Kappeler7Elodie F. Briefer8Marta B. Manser9Nikhil Phaniraj10Marina Scheumann11Judith M. Burkart12Behavioral Ecology and Sociobiology Unit, German Primate Center—Leibniz Institute for Primate Research Göttingen GermanyCognitive Ethology Laboratory German Primate Center—Leibniz Institute for Primate Research Göttingen GermanyDepartment of Comparative Language Science University of Zurich Zurich SwitzerlandCenter for the Interdisciplinary Study of Language Evolution University of Zurich Zurich SwitzerlandBehavioral Ecology and Sociobiology Unit, German Primate Center—Leibniz Institute for Primate Research Göttingen GermanyDepartment for the Ecology of Animal Societies Max Planck Institute of Animal Behaviour Konstanz GermanySchool of Animal and Veterinary Sciences The University of Adelaide Roseworthy South Australia AustraliaBehavioral Ecology and Sociobiology Unit, German Primate Center—Leibniz Institute for Primate Research Göttingen GermanyBehavioural Ecology Group, Section for Ecology and Evolution, Department of Biology University of Copenhagen Copenhagen DenmarkCenter for the Interdisciplinary Study of Language Evolution University of Zurich Zurich SwitzerlandDepartment of Evolutionary Biology and Environmental Studies University of Zurich Zurich SwitzerlandInstitute of Zoology University of Veterinary Medicine Hannover Hannover GermanyDepartment of Evolutionary Anthropology University of Zurich Zurich SwitzerlandAbstract Automated acoustic analysis is increasingly used in behavioural ecology, and determining caller identity is a key element for many investigations. However, variability in feature extraction and classification methods limits the comparability of results across species and studies, constraining conclusions we can draw about the ecology and evolution of the groups under study. We investigated the impact of using different feature extraction (spectro‐temporal measurements, linear and Mel‐frequency cepstral coefficients (MFCC), as well as highly comparative time‐series analysis) and classification methods (discriminant function analysis, neural networks, random forests (RF), and support vector machines) on the consistency of caller identity classification accuracy across 16 mammalian datasets. We found that MFCCs and RFs yield consistently reliable results across datasets, facilitating a standardised approach across species that generates directly comparable data. These findings remained consistent across vocalisation sample sizes and number of individuals considered. We offer guidelines for processing and analysing mammalian vocalisations, fostering greater comparability and advancing our understanding of the evolutionary significance of acoustic communication in diverse mammalian species.https://doi.org/10.1111/2041-210X.70091bioacousticscall distinctivenessindividual identificationmachine learningmethod comparisonreview |
| spellingShingle | Kaja Wierucka Derek Murphy Stuart K. Watson Nikola Falk Claudia Fichtel Julian León Stephan T. Leu Peter M. Kappeler Elodie F. Briefer Marta B. Manser Nikhil Phaniraj Marina Scheumann Judith M. Burkart Same data, different results? Machine learning approaches in bioacoustics Methods in Ecology and Evolution bioacoustics call distinctiveness individual identification machine learning method comparison review |
| title | Same data, different results? Machine learning approaches in bioacoustics |
| title_full | Same data, different results? Machine learning approaches in bioacoustics |
| title_fullStr | Same data, different results? Machine learning approaches in bioacoustics |
| title_full_unstemmed | Same data, different results? Machine learning approaches in bioacoustics |
| title_short | Same data, different results? Machine learning approaches in bioacoustics |
| title_sort | same data different results machine learning approaches in bioacoustics |
| topic | bioacoustics call distinctiveness individual identification machine learning method comparison review |
| url | https://doi.org/10.1111/2041-210X.70091 |
| work_keys_str_mv | AT kajawierucka samedatadifferentresultsmachinelearningapproachesinbioacoustics AT derekmurphy samedatadifferentresultsmachinelearningapproachesinbioacoustics AT stuartkwatson samedatadifferentresultsmachinelearningapproachesinbioacoustics AT nikolafalk samedatadifferentresultsmachinelearningapproachesinbioacoustics AT claudiafichtel samedatadifferentresultsmachinelearningapproachesinbioacoustics AT julianleon samedatadifferentresultsmachinelearningapproachesinbioacoustics AT stephantleu samedatadifferentresultsmachinelearningapproachesinbioacoustics AT petermkappeler samedatadifferentresultsmachinelearningapproachesinbioacoustics AT elodiefbriefer samedatadifferentresultsmachinelearningapproachesinbioacoustics AT martabmanser samedatadifferentresultsmachinelearningapproachesinbioacoustics AT nikhilphaniraj samedatadifferentresultsmachinelearningapproachesinbioacoustics AT marinascheumann samedatadifferentresultsmachinelearningapproachesinbioacoustics AT judithmburkart samedatadifferentresultsmachinelearningapproachesinbioacoustics |