File Download
  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Few-Shot Hyperspectral Image Classification With Unknown Classes Using Multitask Deep Learning

TitleFew-Shot Hyperspectral Image Classification With Unknown Classes Using Multitask Deep Learning
Authors
KeywordsClassification with unknown classes
convolutional neural network (CNN)
deep learning
hyperspectral image classification
multitask learning
Issue Date2021
PublisherInstitute of Electrical and Electronics Engineers. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=36
Citation
IEEE Transactions on Geoscience and Remote Sensing, 2021, v. 59 n. 6, p. 5085-5102 How to Cite?
AbstractCurrent hyperspectral image classification assumes that a predefined classification system is closed and complete, and there are no unknown or novel classes in the unseen data. However, this assumption may be too strict for the real world. Often, novel classes are overlooked when the classification system is constructed. The closed nature forces a model to assign a label given a new sample and may lead to overestimation of known land covers (e.g., crop area). To tackle this issue, we propose a multitask deep learning method that simultaneously conducts classification and reconstruction in the open world (named MDL4OW) where unknown classes may exist. The reconstructed data are compared with the original data; those failing to be reconstructed are considered unknown based on the assumption that they are not well represented in the latent features due to the lack of labels. A threshold needs to be defined to separate the unknown and known classes; we propose two strategies based on the extreme value theory for few- and many-shot scenarios. The proposed method was tested on real-world hyperspectral images; state-of-the-art results were achieved, e.g., improving the overall accuracy by 4.94% for the Salinas data. By considering the existence of unknown classes in the open world, our method achieved more accurate hyperspectral image classification, especially under the few-shot context.
Persistent Identifierhttp://hdl.handle.net/10722/300807
ISSN
2023 Impact Factor: 7.5
2023 SCImago Journal Rankings: 2.403
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorLiu, S-
dc.contributor.authorShi, Q-
dc.contributor.authorZhang, L-
dc.date.accessioned2021-07-06T03:10:32Z-
dc.date.available2021-07-06T03:10:32Z-
dc.date.issued2021-
dc.identifier.citationIEEE Transactions on Geoscience and Remote Sensing, 2021, v. 59 n. 6, p. 5085-5102-
dc.identifier.issn0196-2892-
dc.identifier.urihttp://hdl.handle.net/10722/300807-
dc.description.abstractCurrent hyperspectral image classification assumes that a predefined classification system is closed and complete, and there are no unknown or novel classes in the unseen data. However, this assumption may be too strict for the real world. Often, novel classes are overlooked when the classification system is constructed. The closed nature forces a model to assign a label given a new sample and may lead to overestimation of known land covers (e.g., crop area). To tackle this issue, we propose a multitask deep learning method that simultaneously conducts classification and reconstruction in the open world (named MDL4OW) where unknown classes may exist. The reconstructed data are compared with the original data; those failing to be reconstructed are considered unknown based on the assumption that they are not well represented in the latent features due to the lack of labels. A threshold needs to be defined to separate the unknown and known classes; we propose two strategies based on the extreme value theory for few- and many-shot scenarios. The proposed method was tested on real-world hyperspectral images; state-of-the-art results were achieved, e.g., improving the overall accuracy by 4.94% for the Salinas data. By considering the existence of unknown classes in the open world, our method achieved more accurate hyperspectral image classification, especially under the few-shot context.-
dc.languageeng-
dc.publisherInstitute of Electrical and Electronics Engineers. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=36-
dc.relation.ispartofIEEE Transactions on Geoscience and Remote Sensing-
dc.rightsIEEE Transactions on Geoscience and Remote Sensing. Copyright © Institute of Electrical and Electronics Engineers.-
dc.rights©2021 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.-
dc.subjectClassification with unknown classes-
dc.subjectconvolutional neural network (CNN)-
dc.subjectdeep learning-
dc.subjecthyperspectral image classification-
dc.subjectmultitask learning-
dc.titleFew-Shot Hyperspectral Image Classification With Unknown Classes Using Multitask Deep Learning-
dc.typeArticle-
dc.identifier.emailLiu, S: liusj@hku.hk-
dc.description.naturepostprint-
dc.identifier.doi10.1109/TGRS.2020.3018879-
dc.identifier.scopuseid_2-s2.0-85106694020-
dc.identifier.hkuros323083-
dc.identifier.volume59-
dc.identifier.issue6-
dc.identifier.spage5085-
dc.identifier.epage5102-
dc.identifier.isiWOS:000652834200044-
dc.publisher.placeUnited States-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats