Domain randomization-enhanced deep learning models for bird detection

Abstract Automatic bird detection in ornithological analyses is limited by the accuracy of existing models, due to the lack of training data and the difficulties in extracting the fine-grained features required to distinguish bird species. Here we apply the domain randomization strategy to enhance t...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Xin Mao, Jun Kang Chow, Pin Siang Tan, Kuan-fu Liu, Jimmy Wu, Zhaoyu Su, Ye Hur Cheong, Ghee Leng Ooi, Chun Chiu Pang, Yu-Hsing Wang
Formato: article
Lenguaje:EN
Publicado: Nature Portfolio 2021
Materias:
R
Q
Acceso en línea:https://doaj.org/article/d86c15a5955843a886c09d5595dd32d7
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:d86c15a5955843a886c09d5595dd32d7
record_format dspace
spelling oai:doaj.org-article:d86c15a5955843a886c09d5595dd32d72021-12-02T14:12:43ZDomain randomization-enhanced deep learning models for bird detection10.1038/s41598-020-80101-x2045-2322https://doaj.org/article/d86c15a5955843a886c09d5595dd32d72021-01-01T00:00:00Zhttps://doi.org/10.1038/s41598-020-80101-xhttps://doaj.org/toc/2045-2322Abstract Automatic bird detection in ornithological analyses is limited by the accuracy of existing models, due to the lack of training data and the difficulties in extracting the fine-grained features required to distinguish bird species. Here we apply the domain randomization strategy to enhance the accuracy of the deep learning models in bird detection. Trained with virtual birds of sufficient variations in different environments, the model tends to focus on the fine-grained features of birds and achieves higher accuracies. Based on the 100 terabytes of 2-month continuous monitoring data of egrets, our results cover the findings using conventional manual observations, e.g., vertical stratification of egrets according to body size, and also open up opportunities of long-term bird surveys requiring intensive monitoring that is impractical using conventional methods, e.g., the weather influences on egrets, and the relationship of the migration schedules between the great egrets and little egrets.Xin MaoJun Kang ChowPin Siang TanKuan-fu LiuJimmy WuZhaoyu SuYe Hur CheongGhee Leng OoiChun Chiu PangYu-Hsing WangNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-13 (2021)
institution DOAJ
collection DOAJ
language EN
topic Medicine
R
Science
Q
spellingShingle Medicine
R
Science
Q
Xin Mao
Jun Kang Chow
Pin Siang Tan
Kuan-fu Liu
Jimmy Wu
Zhaoyu Su
Ye Hur Cheong
Ghee Leng Ooi
Chun Chiu Pang
Yu-Hsing Wang
Domain randomization-enhanced deep learning models for bird detection
description Abstract Automatic bird detection in ornithological analyses is limited by the accuracy of existing models, due to the lack of training data and the difficulties in extracting the fine-grained features required to distinguish bird species. Here we apply the domain randomization strategy to enhance the accuracy of the deep learning models in bird detection. Trained with virtual birds of sufficient variations in different environments, the model tends to focus on the fine-grained features of birds and achieves higher accuracies. Based on the 100 terabytes of 2-month continuous monitoring data of egrets, our results cover the findings using conventional manual observations, e.g., vertical stratification of egrets according to body size, and also open up opportunities of long-term bird surveys requiring intensive monitoring that is impractical using conventional methods, e.g., the weather influences on egrets, and the relationship of the migration schedules between the great egrets and little egrets.
format article
author Xin Mao
Jun Kang Chow
Pin Siang Tan
Kuan-fu Liu
Jimmy Wu
Zhaoyu Su
Ye Hur Cheong
Ghee Leng Ooi
Chun Chiu Pang
Yu-Hsing Wang
author_facet Xin Mao
Jun Kang Chow
Pin Siang Tan
Kuan-fu Liu
Jimmy Wu
Zhaoyu Su
Ye Hur Cheong
Ghee Leng Ooi
Chun Chiu Pang
Yu-Hsing Wang
author_sort Xin Mao
title Domain randomization-enhanced deep learning models for bird detection
title_short Domain randomization-enhanced deep learning models for bird detection
title_full Domain randomization-enhanced deep learning models for bird detection
title_fullStr Domain randomization-enhanced deep learning models for bird detection
title_full_unstemmed Domain randomization-enhanced deep learning models for bird detection
title_sort domain randomization-enhanced deep learning models for bird detection
publisher Nature Portfolio
publishDate 2021
url https://doaj.org/article/d86c15a5955843a886c09d5595dd32d7
work_keys_str_mv AT xinmao domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT junkangchow domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT pinsiangtan domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT kuanfuliu domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT jimmywu domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT zhaoyusu domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT yehurcheong domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT gheelengooi domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT chunchiupang domainrandomizationenhanceddeeplearningmodelsforbirddetection
AT yuhsingwang domainrandomizationenhanceddeeplearningmodelsforbirddetection
_version_ 1718391815510425600