基于图像的动物检测和品种识别的神经网络

{"title":"基于图像的动物检测和品种识别的神经网络","authors":"","doi":"10.46243/jst.2020.v5.i5.pp130-134","DOIUrl":null,"url":null,"abstract":"Having accurate, detailed, and up-to-date information about the behaviour of animals in the wild world would\nimprove our ability to study and conserve ecosystems. We investigate the ability to automatically, accurately, and inexpensively\ncollect such data through various sources, which could help catalyse the transformation of many fields of ecology, wildlife\nbiology, zoology, conservation biology, animal behaviour into “big data” sciences and many more. So extracting information\nfrom the pictures remains an expensive, time-consuming, and manual task for us. We demonstrate that such information can be\nautomatically extracted by deep learning and convolutional neural network. Leveraging on recent advances in deep learning\ntechniques in computer vision, we propose in this project a framework to build automated animal recognition in the wild,\naiming at an automated wildlife monitoring system. In particular, we use a single-labelled dataset done by citizen scientists,\nand the state-of-the-art deep convolutional neural network architectures, face biometrics, to train a computational system\ncapable of filtering animal images and identifying species automatically and counting the number of species. Our results\nsuggest that deep learning could enable the inexpensive, unobtrusive, high-volume, and even real-time collection of a wealth of\ninformation about vast numbers of animals in the wild and this, in turn, can, therefore, speed up research findings, construct\nmore efficient citizen science-based monitoring systems and subsequent management decisions, having the potential to make\nsignificant impacts to the world of ecology and trap camera images analysis .","PeriodicalId":23534,"journal":{"name":"Volume 5, Issue 4","volume":"24 14","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2020-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"Image-Based Animal Detection and Breed Identification Using Neural\\nNetworks\",\"authors\":\"\",\"doi\":\"10.46243/jst.2020.v5.i5.pp130-134\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Having accurate, detailed, and up-to-date information about the behaviour of animals in the wild world would\\nimprove our ability to study and conserve ecosystems. We investigate the ability to automatically, accurately, and inexpensively\\ncollect such data through various sources, which could help catalyse the transformation of many fields of ecology, wildlife\\nbiology, zoology, conservation biology, animal behaviour into “big data” sciences and many more. So extracting information\\nfrom the pictures remains an expensive, time-consuming, and manual task for us. We demonstrate that such information can be\\nautomatically extracted by deep learning and convolutional neural network. Leveraging on recent advances in deep learning\\ntechniques in computer vision, we propose in this project a framework to build automated animal recognition in the wild,\\naiming at an automated wildlife monitoring system. In particular, we use a single-labelled dataset done by citizen scientists,\\nand the state-of-the-art deep convolutional neural network architectures, face biometrics, to train a computational system\\ncapable of filtering animal images and identifying species automatically and counting the number of species. Our results\\nsuggest that deep learning could enable the inexpensive, unobtrusive, high-volume, and even real-time collection of a wealth of\\ninformation about vast numbers of animals in the wild and this, in turn, can, therefore, speed up research findings, construct\\nmore efficient citizen science-based monitoring systems and subsequent management decisions, having the potential to make\\nsignificant impacts to the world of ecology and trap camera images analysis .\",\"PeriodicalId\":23534,\"journal\":{\"name\":\"Volume 5, Issue 4\",\"volume\":\"24 14\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-10-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Volume 5, Issue 4\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.46243/jst.2020.v5.i5.pp130-134\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Volume 5, Issue 4","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.46243/jst.2020.v5.i5.pp130-134","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2

摘要

拥有关于野生动物行为的准确、详细和最新的信息将提高我们研究和保护生态系统的能力。我们研究通过各种来源自动、准确、廉价地收集这些数据的能力,这将有助于促进生态学、野生动物生物学、动物学、保护生物学、动物行为等许多领域向“大数据”科学的转变。因此,从图片中提取信息对我们来说仍然是一项昂贵、耗时和手动的任务。我们证明了这些信息可以通过深度学习和卷积神经网络自动提取。利用计算机视觉中深度学习技术的最新进展,我们在本项目中提出了一个框架,用于在野外建立自动动物识别,旨在实现自动野生动物监测系统。特别是,我们使用由公民科学家完成的单标签数据集,以及最先进的深度卷积神经网络架构,面部生物识别技术,来训练能够过滤动物图像并自动识别物种并计算物种数量的计算系统。我们的研究结果表明,深度学习可以实现廉价、不显眼、大批量甚至实时收集大量野生动物的信息,从而可以加快研究成果,构建更有效的基于公民科学的监测系统和随后的管理决策,有可能对生态世界和陷阱相机图像分析产生重大影响。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Image-Based Animal Detection and Breed Identification Using Neural Networks
Having accurate, detailed, and up-to-date information about the behaviour of animals in the wild world would improve our ability to study and conserve ecosystems. We investigate the ability to automatically, accurately, and inexpensively collect such data through various sources, which could help catalyse the transformation of many fields of ecology, wildlife biology, zoology, conservation biology, animal behaviour into “big data” sciences and many more. So extracting information from the pictures remains an expensive, time-consuming, and manual task for us. We demonstrate that such information can be automatically extracted by deep learning and convolutional neural network. Leveraging on recent advances in deep learning techniques in computer vision, we propose in this project a framework to build automated animal recognition in the wild, aiming at an automated wildlife monitoring system. In particular, we use a single-labelled dataset done by citizen scientists, and the state-of-the-art deep convolutional neural network architectures, face biometrics, to train a computational system capable of filtering animal images and identifying species automatically and counting the number of species. Our results suggest that deep learning could enable the inexpensive, unobtrusive, high-volume, and even real-time collection of a wealth of information about vast numbers of animals in the wild and this, in turn, can, therefore, speed up research findings, construct more efficient citizen science-based monitoring systems and subsequent management decisions, having the potential to make significant impacts to the world of ecology and trap camera images analysis .
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
自引率
0.00%
发文量
0
×
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