Show simple item record

dc.contributor.authorHansen, OLP
dc.contributor.authorSvenning, J
dc.contributor.authorOlsen, K
dc.contributor.authorDupont, Steen
dc.contributor.authorgarner, beulah
dc.contributor.authorIosifidis, A
dc.contributor.authorPrice, BW
dc.contributor.authorHøye, TT
dc.date.accessioned2020-04-01T14:03:53Z
dc.date.available2020-04-01T14:03:53Z
dc.date.issued2019-12-24
dc.date.submitted2020-03-31
dc.identifier.citationHansen, OLP, Svenning, J‐C, Olsen, K, et al. Species‐level image classification with convolutional neural network enables insect identification from habitus images. Ecol Evol. 2020; 10: 737– 747. https://doi.org/10.1002/ece3.5921en_US
dc.identifier.issn2045-7758
dc.identifier.doi10.1002/ece3.5921
dc.identifier.urihttp://hdl.handle.net/10141/622674
dc.description.abstract1. Changes in insect biomass, abundance, and diversity are challenging to track at sufficient spatial, temporal, and taxonomic resolution. Camera traps can capture habitus images of ground-dwelling insects. However, currently sampling involves manually detecting and identifying specimens. Here, we test whether a convolutional neural network (CNN) can classify habitus images of ground beetles to species level, and estimate how correct classification relates to body size, number of species inside genera, and species identity. 2. We created an image database of 65,841 museum specimens comprising 361 carabid beetle species from the British Isles and fine-tuned the parameters of a pretrained CNN from a training dataset. By summing up class confidence values within genus, tribe, and subfamily and setting a confidence threshold, we trade-off between classification accuracy, precision, and recall and taxonomic resolution. 3. The CNN classified 51.9% of 19,164 test images correctly to species level and 74.9% to genus level. Average classification recall on species level was 50.7%. Applying a threshold of 0.5 increased the average classification recall to 74.6% at the expense of taxonomic resolution. Higher top value from the output layer and larger sized species were more often classified correctly, as were images of species in genera with few species. 4. Fine-tuning enabled us to classify images with a high mean recall for the whole test dataset to species or higher taxonomic levels, however, with high variability. This indicates that some species are more difficult to identify because of properties such as their body size or the number of related species. 5. Together, species-level image classification of arthropods from museum collections and ecological monitoring can substantially increase the amount of occurrence data that can feasibly be collected. These tools thus provide new opportunities in understanding and predicting ecological responses to environmental change.en_US
dc.language.isoenen_US
dc.publisherWileyen_US
dc.rightsopenAccessen_US
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.titleSpecies‐level image classification with convolutional neural network enables insect identification from habitus imagesen_US
dc.typeJournal Articleen_US
dc.identifier.eissn2045-7758
dc.identifier.journalEcology and Evolutionen_US
dc.identifier.volume10en_US
dc.identifier.issue2en_US
dc.identifier.startpage737 - 747en_US
pubs.organisational-group/Natural History Museum
pubs.organisational-group/Natural History Museum/Science Group
pubs.organisational-group/Natural History Museum/Science Group/Functional groups
pubs.organisational-group/Natural History Museum/Science Group/Functional groups/Other Support
pubs.organisational-group/Natural History Museum/Science Group/Life Sciences
dc.embargoNot knownen_US
elements.import.authorHansen, OLPen_US
elements.import.authorSvenning, Jen_US
elements.import.authorOlsen, Ken_US
elements.import.authorDupont, Sen_US
elements.import.authorGarner, BHen_US
elements.import.authorIosifidis, Aen_US
elements.import.authorPrice, BWen_US
elements.import.authorHøye, TTen_US
dc.description.nhmThis is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2019 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. The attached file is the published pdf.en_US
dc.description.nhmNHM Repository
dc.subject.nhmarthropod samplingen_US
dc.subject.nhmautomatic species identificationen_US
dc.subject.nhmcamera trapen_US
dc.subject.nhmentomological collectionen_US
dc.subject.nhmimage classificationen_US
dc.subject.nhmimage databaseen_US
refterms.dateFOA2020-04-01T14:03:54Z


Files in this item

Thumbnail
Name:
Species-level image classification ...
Size:
727.8Kb
Format:
PDF
Description:
Published/publisher's PDF version
Thumbnail
Name:
ece35921-sup-0001-figs1-s4.pdf
Size:
785.5Kb
Format:
PDF
Description:
Supporting information
Thumbnail
Name:
ece35921-sup-0002-tables1.pdf
Size:
350.9Kb
Format:
PDF
Description:
Supporting information

This item appears in the following Collection(s)

Show simple item record

openAccess
Except where otherwise noted, this item's license is described as openAccess