Domain knowledge integration into deep learning for typhoon intensity classification

Abstract In this report, we propose a deep learning technique for high-accuracy estimation of the intensity class of a typhoon from a single satellite image, by incorporating meteorological domain knowledge. By using the Visual Geometric Group’s model, VGG-16, with images preprocessed with fisheye d...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Maiki Higa, Shinya Tanahara, Yoshitaka Adachi, Natsumi Ishiki, Shin Nakama, Hiroyuki Yamada, Kosuke Ito, Asanobu Kitamoto, Ryota Miyata
Formato: article
Lenguaje:EN
Publicado: Nature Portfolio 2021
Materias:
R
Q
Acceso en línea:https://doaj.org/article/4ce72493e0a74592836482ed6f7f41f9
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:4ce72493e0a74592836482ed6f7f41f9
record_format dspace
spelling oai:doaj.org-article:4ce72493e0a74592836482ed6f7f41f92021-12-02T17:45:17ZDomain knowledge integration into deep learning for typhoon intensity classification10.1038/s41598-021-92286-w2045-2322https://doaj.org/article/4ce72493e0a74592836482ed6f7f41f92021-06-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-92286-whttps://doaj.org/toc/2045-2322Abstract In this report, we propose a deep learning technique for high-accuracy estimation of the intensity class of a typhoon from a single satellite image, by incorporating meteorological domain knowledge. By using the Visual Geometric Group’s model, VGG-16, with images preprocessed with fisheye distortion, which enhances a typhoon’s eye, eyewall, and cloud distribution, we achieved much higher classification accuracy than that of a previous study, even with sequential-split validation. Through comparison of t-distributed stochastic neighbor embedding (t-SNE) plots for the feature maps of VGG with the original satellite images, we also verified that the fisheye preprocessing facilitated cluster formation, suggesting that our model could successfully extract image features related to the typhoon intensity class. Moreover, gradient-weighted class activation mapping (Grad-CAM) was applied to highlight the eye and the cloud distributions surrounding the eye, which are important regions for intensity classification; the results suggest that our model qualitatively gained a viewpoint similar to that of domain experts. A series of analyses revealed that the data-driven approach using only deep learning has limitations, and the integration of domain knowledge could bring new breakthroughs.Maiki HigaShinya TanaharaYoshitaka AdachiNatsumi IshikiShin NakamaHiroyuki YamadaKosuke ItoAsanobu KitamotoRyota MiyataNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-10 (2021)
institution DOAJ
collection DOAJ
language EN
topic Medicine
R
Science
Q
spellingShingle Medicine
R
Science
Q
Maiki Higa
Shinya Tanahara
Yoshitaka Adachi
Natsumi Ishiki
Shin Nakama
Hiroyuki Yamada
Kosuke Ito
Asanobu Kitamoto
Ryota Miyata
Domain knowledge integration into deep learning for typhoon intensity classification
description Abstract In this report, we propose a deep learning technique for high-accuracy estimation of the intensity class of a typhoon from a single satellite image, by incorporating meteorological domain knowledge. By using the Visual Geometric Group’s model, VGG-16, with images preprocessed with fisheye distortion, which enhances a typhoon’s eye, eyewall, and cloud distribution, we achieved much higher classification accuracy than that of a previous study, even with sequential-split validation. Through comparison of t-distributed stochastic neighbor embedding (t-SNE) plots for the feature maps of VGG with the original satellite images, we also verified that the fisheye preprocessing facilitated cluster formation, suggesting that our model could successfully extract image features related to the typhoon intensity class. Moreover, gradient-weighted class activation mapping (Grad-CAM) was applied to highlight the eye and the cloud distributions surrounding the eye, which are important regions for intensity classification; the results suggest that our model qualitatively gained a viewpoint similar to that of domain experts. A series of analyses revealed that the data-driven approach using only deep learning has limitations, and the integration of domain knowledge could bring new breakthroughs.
format article
author Maiki Higa
Shinya Tanahara
Yoshitaka Adachi
Natsumi Ishiki
Shin Nakama
Hiroyuki Yamada
Kosuke Ito
Asanobu Kitamoto
Ryota Miyata
author_facet Maiki Higa
Shinya Tanahara
Yoshitaka Adachi
Natsumi Ishiki
Shin Nakama
Hiroyuki Yamada
Kosuke Ito
Asanobu Kitamoto
Ryota Miyata
author_sort Maiki Higa
title Domain knowledge integration into deep learning for typhoon intensity classification
title_short Domain knowledge integration into deep learning for typhoon intensity classification
title_full Domain knowledge integration into deep learning for typhoon intensity classification
title_fullStr Domain knowledge integration into deep learning for typhoon intensity classification
title_full_unstemmed Domain knowledge integration into deep learning for typhoon intensity classification
title_sort domain knowledge integration into deep learning for typhoon intensity classification
publisher Nature Portfolio
publishDate 2021
url https://doaj.org/article/4ce72493e0a74592836482ed6f7f41f9
work_keys_str_mv AT maikihiga domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
AT shinyatanahara domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
AT yoshitakaadachi domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
AT natsumiishiki domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
AT shinnakama domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
AT hiroyukiyamada domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
AT kosukeito domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
AT asanobukitamoto domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
AT ryotamiyata domainknowledgeintegrationintodeeplearningfortyphoonintensityclassification
_version_ 1718379584959807488