相似性学习网络能独特识别四种海洋和陆地物种的个体

IF 2.7 3区 环境科学与生态学 Q2 ECOLOGY
Ecosphere Pub Date : 2024-10-08 DOI:10.1002/ecs2.70012
Emmanuel Kabuga, Izzy Langley, Monica Arso Civil, John Measey, Bubacarr Bah, Ian Durbach
{"title":"相似性学习网络能独特识别四种海洋和陆地物种的个体","authors":"Emmanuel Kabuga,&nbsp;Izzy Langley,&nbsp;Monica Arso Civil,&nbsp;John Measey,&nbsp;Bubacarr Bah,&nbsp;Ian Durbach","doi":"10.1002/ecs2.70012","DOIUrl":null,"url":null,"abstract":"<p>Estimating the size of animal populations plays an important role in evidence-based conservation and management. Some methods for estimating population size rely on animals being individually identifiable. Traditionally, this has been done by marking physically captured animals, but increasingly, animals with distinctive natural markings are surveyed noninvasively using cameras. Animal reidentification from photographs is usually done manually, which is expensive, laborious, and requires considerable skill. An alternative is to develop computer vision methods that can support or replace the manual identification task. We developed an automated approach using deep learning to identify whether a pair of photographs is of the same individual or not. The core of the approach is a similarity learning network that uses paired convolutional neural networks with a triplet loss function to summarize image pairs and decide whether they are from the same individual. Prior to the main matching step, two additional convolutional neural networks perform image segmentation, cropping the animal object within the image, and orientation prediction, deciding which side of the animal was photographed. We applied the approach to four species, with images of the same individual often spanning several years: systematic surveys of bottlenose dolphins (<i>Tursiops truncatus</i>, 2008–2019) and harbor seals (<i>Phoca vitulina</i>, 2015–2019), a citizen science dataset of western leopard toads (<i>Sclerophrys pantherina</i>, unknown dates), and a publicly available repository of humpback whale images (<i>Megaptera novaeangliae</i>, unknown dates). For these species, our best-performing models were able to identify whether a pair of images were from the same individual or different individuals in 95.8%, 94.6%, 88.2%, and 83.8% of the cases, respectively. We found that triplet loss functions outperformed binary cross-entropy loss functions and that data augmentation and additional manual curation of training data provided small but consistent improvements in performance. These results demonstrate the potential of deep learning to replace or, more likely, support and facilitate manual individual identification efforts.</p>","PeriodicalId":48930,"journal":{"name":"Ecosphere","volume":"15 10","pages":""},"PeriodicalIF":2.7000,"publicationDate":"2024-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/ecs2.70012","citationCount":"0","resultStr":"{\"title\":\"Similarity learning networks uniquely identify individuals of four marine and terrestrial species\",\"authors\":\"Emmanuel Kabuga,&nbsp;Izzy Langley,&nbsp;Monica Arso Civil,&nbsp;John Measey,&nbsp;Bubacarr Bah,&nbsp;Ian Durbach\",\"doi\":\"10.1002/ecs2.70012\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>Estimating the size of animal populations plays an important role in evidence-based conservation and management. Some methods for estimating population size rely on animals being individually identifiable. Traditionally, this has been done by marking physically captured animals, but increasingly, animals with distinctive natural markings are surveyed noninvasively using cameras. Animal reidentification from photographs is usually done manually, which is expensive, laborious, and requires considerable skill. An alternative is to develop computer vision methods that can support or replace the manual identification task. We developed an automated approach using deep learning to identify whether a pair of photographs is of the same individual or not. The core of the approach is a similarity learning network that uses paired convolutional neural networks with a triplet loss function to summarize image pairs and decide whether they are from the same individual. Prior to the main matching step, two additional convolutional neural networks perform image segmentation, cropping the animal object within the image, and orientation prediction, deciding which side of the animal was photographed. We applied the approach to four species, with images of the same individual often spanning several years: systematic surveys of bottlenose dolphins (<i>Tursiops truncatus</i>, 2008–2019) and harbor seals (<i>Phoca vitulina</i>, 2015–2019), a citizen science dataset of western leopard toads (<i>Sclerophrys pantherina</i>, unknown dates), and a publicly available repository of humpback whale images (<i>Megaptera novaeangliae</i>, unknown dates). For these species, our best-performing models were able to identify whether a pair of images were from the same individual or different individuals in 95.8%, 94.6%, 88.2%, and 83.8% of the cases, respectively. We found that triplet loss functions outperformed binary cross-entropy loss functions and that data augmentation and additional manual curation of training data provided small but consistent improvements in performance. These results demonstrate the potential of deep learning to replace or, more likely, support and facilitate manual individual identification efforts.</p>\",\"PeriodicalId\":48930,\"journal\":{\"name\":\"Ecosphere\",\"volume\":\"15 10\",\"pages\":\"\"},\"PeriodicalIF\":2.7000,\"publicationDate\":\"2024-10-08\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://onlinelibrary.wiley.com/doi/epdf/10.1002/ecs2.70012\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Ecosphere\",\"FirstCategoryId\":\"93\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1002/ecs2.70012\",\"RegionNum\":3,\"RegionCategory\":\"环境科学与生态学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"ECOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Ecosphere","FirstCategoryId":"93","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/ecs2.70012","RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ECOLOGY","Score":null,"Total":0}
引用次数: 0

摘要

估算动物种群数量在循证保护和管理中发挥着重要作用。一些估算种群数量的方法依赖于可单独识别的动物。传统上,这是通过对捕获的动物进行标记来实现的,但现在越来越多地使用照相机对具有独特自然标记的动物进行非侵入式调查。从照片上重新识别动物通常是人工完成的,这种方法既昂贵又费力,而且需要相当高的技能。另一种方法是开发计算机视觉方法,以支持或取代人工识别任务。我们开发了一种使用深度学习的自动方法,用于识别一对照片是否为同一个体。该方法的核心是一个相似性学习网络,它使用带有三重损失函数的成对卷积神经网络来总结图像对,并判断它们是否来自同一个人。在主要匹配步骤之前,另外两个卷积神经网络会进行图像分割,裁剪图像中的动物对象,并进行方向预测,决定拍摄动物的哪一面。我们将该方法应用于四个物种,同一个体的图像往往跨越数年:瓶鼻海豚(Tursiops truncatus,2008-2019 年)和港海豹(Phoca vitulina,2015-2019 年)的系统调查、西豹蟾蜍(Sclerophrys pantherina,日期不详)的公民科学数据集以及座头鲸图像的公开资料库(Megaptera novaeangliae,日期不详)。对于这些物种,我们表现最好的模型能够分别在 95.8%、94.6%、88.2% 和 83.8% 的情况下识别出一对图像是来自同一个体还是不同个体。我们发现,三重损失函数的性能优于二元交叉熵损失函数,而且数据扩充和额外的手动训练数据整理能带来微小但一致的性能提升。这些结果表明,深度学习有潜力取代或更有可能支持和促进人工个体识别工作。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Similarity learning networks uniquely identify individuals of four marine and terrestrial species

Similarity learning networks uniquely identify individuals of four marine and terrestrial species

Estimating the size of animal populations plays an important role in evidence-based conservation and management. Some methods for estimating population size rely on animals being individually identifiable. Traditionally, this has been done by marking physically captured animals, but increasingly, animals with distinctive natural markings are surveyed noninvasively using cameras. Animal reidentification from photographs is usually done manually, which is expensive, laborious, and requires considerable skill. An alternative is to develop computer vision methods that can support or replace the manual identification task. We developed an automated approach using deep learning to identify whether a pair of photographs is of the same individual or not. The core of the approach is a similarity learning network that uses paired convolutional neural networks with a triplet loss function to summarize image pairs and decide whether they are from the same individual. Prior to the main matching step, two additional convolutional neural networks perform image segmentation, cropping the animal object within the image, and orientation prediction, deciding which side of the animal was photographed. We applied the approach to four species, with images of the same individual often spanning several years: systematic surveys of bottlenose dolphins (Tursiops truncatus, 2008–2019) and harbor seals (Phoca vitulina, 2015–2019), a citizen science dataset of western leopard toads (Sclerophrys pantherina, unknown dates), and a publicly available repository of humpback whale images (Megaptera novaeangliae, unknown dates). For these species, our best-performing models were able to identify whether a pair of images were from the same individual or different individuals in 95.8%, 94.6%, 88.2%, and 83.8% of the cases, respectively. We found that triplet loss functions outperformed binary cross-entropy loss functions and that data augmentation and additional manual curation of training data provided small but consistent improvements in performance. These results demonstrate the potential of deep learning to replace or, more likely, support and facilitate manual individual identification efforts.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Ecosphere
Ecosphere ECOLOGY-
CiteScore
4.70
自引率
3.70%
发文量
378
审稿时长
15 weeks
期刊介绍: The scope of Ecosphere is as broad as the science of ecology itself. The journal welcomes submissions from all sub-disciplines of ecological science, as well as interdisciplinary studies relating to ecology. The journal''s goal is to provide a rapid-publication, online-only, open-access alternative to ESA''s other journals, while maintaining the rigorous standards of peer review for which ESA publications are renowned.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信