A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean species
Abstract Researchers can investigate many aspects of animal ecology through noninvasive photo–identification. Photo–identification is becoming more efficient as matching individuals between photos is increasingly automated. However, the convolutional neural network models that have facilitated this...
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2023-10-01
|
| Series: | Methods in Ecology and Evolution |
| Subjects: | |
| Online Access: | https://doi.org/10.1111/2041-210X.14167 |
| _version_ | 1827800205059162112 |
|---|---|
| author | Philip T. Patton Ted Cheeseman Kenshin Abe Taiki Yamaguchi Walter Reade Ken Southerland Addison Howard Erin M. Oleson Jason B. Allen Erin Ashe Aline Athayde Robin W. Baird Charla Basran Elsa Cabrera John Calambokidis Júlio Cardoso Emma L. Carroll Amina Cesario Barbara J. Cheney Enrico Corsi Jens Currie John W. Durban Erin A. Falcone Holly Fearnbach Kiirsten Flynn Trish Franklin Wally Franklin Bárbara Galletti Vernazzani Tilen Genov Marie Hill David R. Johnston Erin L. Keene Sabre D. Mahaffy Tamara L. McGuire Liah McPherson Catherine Meyer Robert Michaud Anastasia Miliou Dara N. Orbach Heidi C. Pearson Marianne H. Rasmussen William J. Rayment Caroline Rinaldi Renato Rinaldi Salvatore Siciliano Stephanie Stack Beatriz Tintore Leigh G. Torres Jared R. Towers Cameron Trotter Reny Tyson Moore Caroline R. Weir Rebecca Wellard Randall Wells Kymberly M. Yano Jochen R. Zaeschmar Lars Bejder |
| author_facet | Philip T. Patton Ted Cheeseman Kenshin Abe Taiki Yamaguchi Walter Reade Ken Southerland Addison Howard Erin M. Oleson Jason B. Allen Erin Ashe Aline Athayde Robin W. Baird Charla Basran Elsa Cabrera John Calambokidis Júlio Cardoso Emma L. Carroll Amina Cesario Barbara J. Cheney Enrico Corsi Jens Currie John W. Durban Erin A. Falcone Holly Fearnbach Kiirsten Flynn Trish Franklin Wally Franklin Bárbara Galletti Vernazzani Tilen Genov Marie Hill David R. Johnston Erin L. Keene Sabre D. Mahaffy Tamara L. McGuire Liah McPherson Catherine Meyer Robert Michaud Anastasia Miliou Dara N. Orbach Heidi C. Pearson Marianne H. Rasmussen William J. Rayment Caroline Rinaldi Renato Rinaldi Salvatore Siciliano Stephanie Stack Beatriz Tintore Leigh G. Torres Jared R. Towers Cameron Trotter Reny Tyson Moore Caroline R. Weir Rebecca Wellard Randall Wells Kymberly M. Yano Jochen R. Zaeschmar Lars Bejder |
| author_sort | Philip T. Patton |
| collection | DOAJ |
| description | Abstract Researchers can investigate many aspects of animal ecology through noninvasive photo–identification. Photo–identification is becoming more efficient as matching individuals between photos is increasingly automated. However, the convolutional neural network models that have facilitated this change need many training images to generalize well. As a result, they have often been developed for individual species that meet this threshold. These single‐species methods might underperform, as they ignore potential similarities in identifying characteristics and the photo–identification process among species. In this paper, we introduce a multi‐species photo–identification model based on a state‐of‐the‐art method in human facial recognition, the ArcFace classification head. Our model uses two such heads to jointly classify species and identities, allowing species to share information and parameters within the network. As a demonstration, we trained this model with 50,796 images from 39 catalogues of 24 cetacean species, evaluating its predictive performance on 21,192 test images from the same catalogues. We further evaluated its predictive performance with two external catalogues entirely composed of identities that the model did not see during training. The model achieved a mean average precision (MAP) of 0.869 on the test set. Of these, 10 catalogues representing seven species achieved a MAP score over 0.95. For some species, there was notable variation in performance among catalogues, largely explained by variation in photo quality. Finally, the model appeared to generalize well, with the two external catalogues scoring similarly to their species' counterparts in the larger test set. From our cetacean application, we provide a list of recommendations for potential users of this model, focusing on those with cetacean photo–identification catalogues. For example, users with high quality images of animals identified by dorsal nicks and notches should expect near optimal performance. Users can expect decreasing performance for catalogues with higher proportions of indistinct individuals or poor quality photos. Finally, we note that this model is currently freely available as code in a GitHub repository and as a graphical user interface, with additional functionality for collaborative data management, via Happywhale.com. |
| first_indexed | 2024-03-11T20:04:58Z |
| format | Article |
| id | doaj.art-f5f26c5e992e4bbfbb39d716613aac91 |
| institution | Directory Open Access Journal |
| issn | 2041-210X |
| language | English |
| last_indexed | 2024-03-11T20:04:58Z |
| publishDate | 2023-10-01 |
| publisher | Wiley |
| record_format | Article |
| series | Methods in Ecology and Evolution |
| spelling | doaj.art-f5f26c5e992e4bbfbb39d716613aac912023-10-04T06:42:59ZengWileyMethods in Ecology and Evolution2041-210X2023-10-0114102611262510.1111/2041-210X.14167A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean speciesPhilip T. Patton0Ted Cheeseman1Kenshin Abe2Taiki Yamaguchi3Walter Reade4Ken Southerland5Addison Howard6Erin M. Oleson7Jason B. Allen8Erin Ashe9Aline Athayde10Robin W. Baird11Charla Basran12Elsa Cabrera13John Calambokidis14Júlio Cardoso15Emma L. Carroll16Amina Cesario17Barbara J. Cheney18Enrico Corsi19Jens Currie20John W. Durban21Erin A. Falcone22Holly Fearnbach23Kiirsten Flynn24Trish Franklin25Wally Franklin26Bárbara Galletti Vernazzani27Tilen Genov28Marie Hill29David R. Johnston30Erin L. Keene31Sabre D. Mahaffy32Tamara L. McGuire33Liah McPherson34Catherine Meyer35Robert Michaud36Anastasia Miliou37Dara N. Orbach38Heidi C. Pearson39Marianne H. Rasmussen40William J. Rayment41Caroline Rinaldi42Renato Rinaldi43Salvatore Siciliano44Stephanie Stack45Beatriz Tintore46Leigh G. Torres47Jared R. Towers48Cameron Trotter49Reny Tyson Moore50Caroline R. Weir51Rebecca Wellard52Randall Wells53Kymberly M. Yano54Jochen R. Zaeschmar55Lars Bejder56Marine Mammal Research Program, Hawai'i Institute of Marine Biology University of Hawai‘i at Mānoa Kāne'ohe Hawai'i USAMarine Ecological Research Centre Southern Cross University Lismore New South Wales AustraliaPreferred Networks, Inc. Chiyoda‐ku Tokyo JapanPreferred Networks, Inc. Chiyoda‐ku Tokyo JapanGoogle, Kaggle San Francisco California USAHappywhale.com Santa Cruz California USAGoogle, Kaggle San Francisco California USANOAA Fisheries Pacific Islands Fisheries Science Center Honolulu Hawai'i USAChicago Zoological Society's Sarasota Dolphin Research Program c/o Mote Marine Laboratory Sarasota Florida USAOceans Initiative Seattle Washington USAProjeto Baleia à Vista (ProBaV) Ilhabela BrazilCascadia Research Collective Olympia Washington USAResearch Center in Húsavík University of Iceland Húsavík IcelandCentro de Conservación Cetacea (CCC) Santiago ChileCascadia Research Collective Olympia Washington USAProjeto Baleia à Vista (ProBaV) Ilhabela BrazilSchool of Biological Sciences University of Auckland‐Waipapa Taumata Rau Auckland New ZealandTethys Research Institute Milan ItalySchool of Biological Sciences University of Aberdeen Cromarty UKCascadia Research Collective Olympia Washington USAMarine Mammal Research Program, Hawai'i Institute of Marine Biology University of Hawai‘i at Mānoa Kāne'ohe Hawai'i USASR3, SeaLife Response, Rehabilitation and Research Des Moines Washington USAMarine Ecology and Telemetry Research Seabeck Washington USASR3, SeaLife Response, Rehabilitation and Research Des Moines Washington USACascadia Research Collective Olympia Washington USAMarine Ecological Research Centre Southern Cross University Lismore New South Wales AustraliaMarine Ecological Research Centre Southern Cross University Lismore New South Wales AustraliaCentro de Conservación Cetacea (CCC) Santiago ChileMorigenos‐Slovenian Marine Mammal Society Piran SloveniaNOAA Fisheries Pacific Islands Fisheries Science Center Honolulu Hawai'i USAMarine Science Department, Te Tari Putaiao Taimoana University of Otago Otago New ZealandMarine Ecology and Telemetry Research Seabeck Washington USACascadia Research Collective Olympia Washington USAThe Cook Inlet Beluga Whale Photo–ID Project Anchorage Alaska USAMarine Mammal Research Program, Hawai'i Institute of Marine Biology University of Hawai‘i at Mānoa Kāne'ohe Hawai'i USASchool of Biological Sciences, Te Kura Mātauranga Koiora University of Auckland Auckland New ZealandGroupe de Recherche et D'éducation sur les Mammifères Marins (GREMM) Tadoussac Québec CanadaArchipelagos Institute of Marine Conservation Samos Island GreeceDepartment of Life Sciences Texas A&M University‐Corpus Christi Corpus Christi Texas USADepartment of Natural Sciences University of Alaska Southeast Juneau Alaska USAResearch Center in Húsavík University of Iceland Húsavík IcelandDepartment of Marine Science‐Te Tari Pūtaiao Taimoana University of Otago Dunedin New ZealandL'association Evasion Tropicale Bouillante GuadeloupeL'association Evasion Tropicale Bouillante GuadeloupeDepartamento de Ciências Biológicas Escola Nacional de Saúde Pública/Fiocruz Rio de Janeiro BrazilPacific Whale Foundation Wailuku Hawai'i USAArchipelagos Institute of Marine Conservation Samos Island GreeceMarine Mammal Institute, Oregon State University Newport Oregon USABay Cetology Alert Bay British Columbia CanadaSchool of Engineering Newcastle University Newcastle UKChicago Zoological Society's Sarasota Dolphin Research Program c/o Mote Marine Laboratory Sarasota Florida USAFalklands Conservation Stanley Falkland IslandsCentre for Marine Science and Technology Curtin University Bentley Western Australia AustraliaChicago Zoological Society's Sarasota Dolphin Research Program c/o Mote Marine Laboratory Sarasota Florida USANOAA Fisheries Pacific Islands Fisheries Science Center Honolulu Hawai'i USAFar Out Ocean Research Collective Paihia New ZealandMarine Mammal Research Program, Hawai'i Institute of Marine Biology University of Hawai‘i at Mānoa Kāne'ohe Hawai'i USAAbstract Researchers can investigate many aspects of animal ecology through noninvasive photo–identification. Photo–identification is becoming more efficient as matching individuals between photos is increasingly automated. However, the convolutional neural network models that have facilitated this change need many training images to generalize well. As a result, they have often been developed for individual species that meet this threshold. These single‐species methods might underperform, as they ignore potential similarities in identifying characteristics and the photo–identification process among species. In this paper, we introduce a multi‐species photo–identification model based on a state‐of‐the‐art method in human facial recognition, the ArcFace classification head. Our model uses two such heads to jointly classify species and identities, allowing species to share information and parameters within the network. As a demonstration, we trained this model with 50,796 images from 39 catalogues of 24 cetacean species, evaluating its predictive performance on 21,192 test images from the same catalogues. We further evaluated its predictive performance with two external catalogues entirely composed of identities that the model did not see during training. The model achieved a mean average precision (MAP) of 0.869 on the test set. Of these, 10 catalogues representing seven species achieved a MAP score over 0.95. For some species, there was notable variation in performance among catalogues, largely explained by variation in photo quality. Finally, the model appeared to generalize well, with the two external catalogues scoring similarly to their species' counterparts in the larger test set. From our cetacean application, we provide a list of recommendations for potential users of this model, focusing on those with cetacean photo–identification catalogues. For example, users with high quality images of animals identified by dorsal nicks and notches should expect near optimal performance. Users can expect decreasing performance for catalogues with higher proportions of indistinct individuals or poor quality photos. Finally, we note that this model is currently freely available as code in a GitHub repository and as a graphical user interface, with additional functionality for collaborative data management, via Happywhale.com.https://doi.org/10.1111/2041-210X.14167artificial intelligencecetaceancomputer visionconvolutional neural networkdeep learningdolphin |
| spellingShingle | Philip T. Patton Ted Cheeseman Kenshin Abe Taiki Yamaguchi Walter Reade Ken Southerland Addison Howard Erin M. Oleson Jason B. Allen Erin Ashe Aline Athayde Robin W. Baird Charla Basran Elsa Cabrera John Calambokidis Júlio Cardoso Emma L. Carroll Amina Cesario Barbara J. Cheney Enrico Corsi Jens Currie John W. Durban Erin A. Falcone Holly Fearnbach Kiirsten Flynn Trish Franklin Wally Franklin Bárbara Galletti Vernazzani Tilen Genov Marie Hill David R. Johnston Erin L. Keene Sabre D. Mahaffy Tamara L. McGuire Liah McPherson Catherine Meyer Robert Michaud Anastasia Miliou Dara N. Orbach Heidi C. Pearson Marianne H. Rasmussen William J. Rayment Caroline Rinaldi Renato Rinaldi Salvatore Siciliano Stephanie Stack Beatriz Tintore Leigh G. Torres Jared R. Towers Cameron Trotter Reny Tyson Moore Caroline R. Weir Rebecca Wellard Randall Wells Kymberly M. Yano Jochen R. Zaeschmar Lars Bejder A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean species Methods in Ecology and Evolution artificial intelligence cetacean computer vision convolutional neural network deep learning dolphin |
| title | A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean species |
| title_full | A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean species |
| title_fullStr | A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean species |
| title_full_unstemmed | A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean species |
| title_short | A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean species |
| title_sort | deep learning approach to photo identification demonstrates high performance on two dozen cetacean species |
| topic | artificial intelligence cetacean computer vision convolutional neural network deep learning dolphin |
| url | https://doi.org/10.1111/2041-210X.14167 |
| work_keys_str_mv | AT philiptpatton adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT tedcheeseman adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT kenshinabe adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT taikiyamaguchi adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT walterreade adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT kensoutherland adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT addisonhoward adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT erinmoleson adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT jasonballen adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT erinashe adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT alineathayde adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT robinwbaird adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT charlabasran adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT elsacabrera adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT johncalambokidis adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT juliocardoso adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT emmalcarroll adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT aminacesario adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT barbarajcheney adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT enricocorsi adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT jenscurrie adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT johnwdurban adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT erinafalcone adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT hollyfearnbach adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT kiirstenflynn adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT trishfranklin adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT wallyfranklin adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT barbaragallettivernazzani adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT tilengenov adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT mariehill adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT davidrjohnston adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT erinlkeene adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT sabredmahaffy adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT tamaralmcguire adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT liahmcpherson adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT catherinemeyer adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT robertmichaud adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT anastasiamiliou adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT daranorbach adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT heidicpearson adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT mariannehrasmussen adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT williamjrayment adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT carolinerinaldi adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT renatorinaldi adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT salvatoresiciliano adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT stephaniestack adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT beatriztintore adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT leighgtorres adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT jaredrtowers adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT camerontrotter adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT renytysonmoore adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT carolinerweir adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT rebeccawellard adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT randallwells adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT kymberlymyano adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT jochenrzaeschmar adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT larsbejder adeeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT philiptpatton deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT tedcheeseman deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT kenshinabe deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT taikiyamaguchi deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT walterreade deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT kensoutherland deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT addisonhoward deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT erinmoleson deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT jasonballen deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT erinashe deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT alineathayde deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT robinwbaird deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT charlabasran deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT elsacabrera deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT johncalambokidis deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT juliocardoso deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT emmalcarroll deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT aminacesario deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT barbarajcheney deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT enricocorsi deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT jenscurrie deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT johnwdurban deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT erinafalcone deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT hollyfearnbach deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT kiirstenflynn deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT trishfranklin deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT wallyfranklin deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT barbaragallettivernazzani deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT tilengenov deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT mariehill deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT davidrjohnston deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT erinlkeene deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT sabredmahaffy deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT tamaralmcguire deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT liahmcpherson deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT catherinemeyer deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT robertmichaud deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT anastasiamiliou deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT daranorbach deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT heidicpearson deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT mariannehrasmussen deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT williamjrayment deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT carolinerinaldi deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT renatorinaldi deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT salvatoresiciliano deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT stephaniestack deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT beatriztintore deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT leighgtorres deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT jaredrtowers deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT camerontrotter deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT renytysonmoore deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT carolinerweir deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT rebeccawellard deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT randallwells deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT kymberlymyano deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT jochenrzaeschmar deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies AT larsbejder deeplearningapproachtophotoidentificationdemonstrateshighperformanceontwodozencetaceanspecies |