Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears
Emerging technologies support a new era of applied wildlife research, generating data on scales from individuals to populations. Computer vision methods can process large datasets generated through image‐based techniques by automating the detection and identification of species and individuals. With...
Published in: | Ecology and Evolution |
---|---|
Main Authors: | , , , |
Format: | Text |
Language: | English |
Published: |
John Wiley and Sons Inc.
2020
|
Subjects: | |
Online Access: | http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7713984/ https://doi.org/10.1002/ece3.6840 |
id |
ftpubmed:oai:pubmedcentral.nih.gov:7713984 |
---|---|
record_format |
openpolar |
spelling |
ftpubmed:oai:pubmedcentral.nih.gov:7713984 2023-05-15T18:42:12+02:00 Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears Clapham, Melanie Miller, Ed Nguyen, Mary Darimont, Chris T. 2020-11-06 http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7713984/ https://doi.org/10.1002/ece3.6840 en eng John Wiley and Sons Inc. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7713984/ http://dx.doi.org/10.1002/ece3.6840 © 2020 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. This is an open access article under the terms of the http://creativecommons.org/licenses/by/4.0/ License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. CC-BY Ecol Evol Original Research Text 2020 ftpubmed https://doi.org/10.1002/ece3.6840 2020-12-13T01:33:54Z Emerging technologies support a new era of applied wildlife research, generating data on scales from individuals to populations. Computer vision methods can process large datasets generated through image‐based techniques by automating the detection and identification of species and individuals. With the exception of primates, however, there are no objective visual methods of individual identification for species that lack unique and consistent body markings. We apply deep learning approaches of facial recognition using object detection, landmark detection, a similarity comparison network, and an support vector machine‐based classifier to identify individuals in a representative species, the brown bear Ursus arctos. Our open‐source application, BearID, detects a bear’s face in an image, rotates and extracts the face, creates an “embedding” for the face, and uses the embedding to classify the individual. We trained and tested the application using labeled images of 132 known individuals collected from British Columbia, Canada, and Alaska, USA. Based on 4,674 images, with an 80/20% split for training and testing, respectively, we achieved a facial detection (ability to find a face) average precision of 0.98 and an individual classification (ability to identify the individual) accuracy of 83.9%. BearID and its annotated source code provide a replicable methodology for applying deep learning methods of facial recognition applicable to many other species that lack distinguishing markings. Further analyses of performance should focus on the influence of certain parameters on recognition accuracy, such as age and body size. Combining BearID with camera trapping could facilitate fine‐scale behavioral research such as individual spatiotemporal activity patterns, and a cost‐effective method of population monitoring through mark–recapture studies, with implications for species and landscape conservation and management. Applications to practical conservation include identifying problem individuals in human–wildlife ... Text Ursus arctos Alaska PubMed Central (PMC) British Columbia ENVELOPE(-125.003,-125.003,54.000,54.000) Canada Ecology and Evolution 10 23 12883 12892 |
institution |
Open Polar |
collection |
PubMed Central (PMC) |
op_collection_id |
ftpubmed |
language |
English |
topic |
Original Research |
spellingShingle |
Original Research Clapham, Melanie Miller, Ed Nguyen, Mary Darimont, Chris T. Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears |
topic_facet |
Original Research |
description |
Emerging technologies support a new era of applied wildlife research, generating data on scales from individuals to populations. Computer vision methods can process large datasets generated through image‐based techniques by automating the detection and identification of species and individuals. With the exception of primates, however, there are no objective visual methods of individual identification for species that lack unique and consistent body markings. We apply deep learning approaches of facial recognition using object detection, landmark detection, a similarity comparison network, and an support vector machine‐based classifier to identify individuals in a representative species, the brown bear Ursus arctos. Our open‐source application, BearID, detects a bear’s face in an image, rotates and extracts the face, creates an “embedding” for the face, and uses the embedding to classify the individual. We trained and tested the application using labeled images of 132 known individuals collected from British Columbia, Canada, and Alaska, USA. Based on 4,674 images, with an 80/20% split for training and testing, respectively, we achieved a facial detection (ability to find a face) average precision of 0.98 and an individual classification (ability to identify the individual) accuracy of 83.9%. BearID and its annotated source code provide a replicable methodology for applying deep learning methods of facial recognition applicable to many other species that lack distinguishing markings. Further analyses of performance should focus on the influence of certain parameters on recognition accuracy, such as age and body size. Combining BearID with camera trapping could facilitate fine‐scale behavioral research such as individual spatiotemporal activity patterns, and a cost‐effective method of population monitoring through mark–recapture studies, with implications for species and landscape conservation and management. Applications to practical conservation include identifying problem individuals in human–wildlife ... |
format |
Text |
author |
Clapham, Melanie Miller, Ed Nguyen, Mary Darimont, Chris T. |
author_facet |
Clapham, Melanie Miller, Ed Nguyen, Mary Darimont, Chris T. |
author_sort |
Clapham, Melanie |
title |
Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears |
title_short |
Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears |
title_full |
Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears |
title_fullStr |
Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears |
title_full_unstemmed |
Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears |
title_sort |
automated facial recognition for wildlife that lack unique markings: a deep learning approach for brown bears |
publisher |
John Wiley and Sons Inc. |
publishDate |
2020 |
url |
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7713984/ https://doi.org/10.1002/ece3.6840 |
long_lat |
ENVELOPE(-125.003,-125.003,54.000,54.000) |
geographic |
British Columbia Canada |
geographic_facet |
British Columbia Canada |
genre |
Ursus arctos Alaska |
genre_facet |
Ursus arctos Alaska |
op_source |
Ecol Evol |
op_relation |
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7713984/ http://dx.doi.org/10.1002/ece3.6840 |
op_rights |
© 2020 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. This is an open access article under the terms of the http://creativecommons.org/licenses/by/4.0/ License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. |
op_rightsnorm |
CC-BY |
op_doi |
https://doi.org/10.1002/ece3.6840 |
container_title |
Ecology and Evolution |
container_volume |
10 |
container_issue |
23 |
container_start_page |
12883 |
op_container_end_page |
12892 |
_version_ |
1766231831885119488 |