Neural Architecture Search-Based Few-Shot Learning for Hyperspectral Image Classification
Few-shot learning (FSL) has achieved promising performance in hyperspectral image classification (HSIC) with few labeled samples by designing a proper embedding feature extractor. However, the performance of embedding feature extractors relies on the design of efficient deep convolutional neural net...
Gespeichert in:
Veröffentlicht in: | IEEE transactions on geoscience and remote sensing 2024, Vol.62, p.1-15 |
---|---|
Hauptverfasser: | , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | 15 |
---|---|
container_issue | |
container_start_page | 1 |
container_title | IEEE transactions on geoscience and remote sensing |
container_volume | 62 |
creator | Xiao, Fen Xiang, Han Cao, Chunhong Gao, Xieping |
description | Few-shot learning (FSL) has achieved promising performance in hyperspectral image classification (HSIC) with few labeled samples by designing a proper embedding feature extractor. However, the performance of embedding feature extractors relies on the design of efficient deep convolutional neural network (CNN) architectures, which heavily depends on expertise knowledge. Particularly, FSL requires extracting discriminative features effectively across different domains, which makes the construction even more challenging. In this article, we propose a novel neural architecture search-based FSL model for HSI classification (HCFSL-NAS). Three novel strategies are proposed in this work. First, a neural architecture search (NAS)-based embedding feature extractor is developed to the FSL in HSIC, whose search space includes a group of proposed multiscale convolutions with channel attention. Second, a multisource learning framework is employed to aggregate abundant heterogeneous and homogeneous source data, which enables the powerful generalization of the network to the HSIC with only few labeled samples. Finally, the pointwise-based cross-entropy (CE) loss and the pairwise-based adaptive sparse loss are jointly optimized to maximize interclass distance and minimize the distance within a class simultaneously. Experimental results on four publicly hyperspectral datasets demonstrate that HCFSL-NAS outperforms both the exiting FSL methods and supervised learning methods for HSI classification with only few labeled samples. The code is available at https://github.com/xh-captain/HCFSL-NAS . |
doi_str_mv | 10.1109/TGRS.2024.3385478 |
format | Article |
fullrecord | <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_ieee_primary_10493163</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>10493163</ieee_id><sourcerecordid>3040055138</sourcerecordid><originalsourceid>FETCH-LOGICAL-c246t-e9ab0f3949530ddfa079ca4cec480681bf296c53fa6121e772e9d90b2583e1dc3</originalsourceid><addsrcrecordid>eNpNkE1Lw0AQhhdRsFZ_gOAh4Dl19ivZPdZiP6Ao2HrwFLab2TalTeJugvTfm9AePA3v8D4z8BDySGFEKeiX9exzNWLAxIhzJUWqrsiASqliSIS4JgOgOomZ0uyW3IWwB6BC0nRAvt-x9eYQjb3dFQ3apvUYrdB0MX41AfNoir_xalc10bLblkW5jVzlo_mpRh_qDujpxdFsMZocTAiFK6xpiqq8JzfOHAI-XOaQfE3f1pN5vPyYLSbjZWyZSJoYtdmA41poySHPnYFUWyMsWqEgUXTjmE6s5M4klFFMU4Y617BhUnGkueVD8ny-W_vqp8XQZPuq9WX3MuMgAKSkXHUtem5ZX4Xg0WW1L47GnzIKWW8w6w1mvcHsYrBjns5MgYj_-kJzmnD-Bx7RbSU</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>3040055138</pqid></control><display><type>article</type><title>Neural Architecture Search-Based Few-Shot Learning for Hyperspectral Image Classification</title><source>IEEE Electronic Library (IEL)</source><creator>Xiao, Fen ; Xiang, Han ; Cao, Chunhong ; Gao, Xieping</creator><creatorcontrib>Xiao, Fen ; Xiang, Han ; Cao, Chunhong ; Gao, Xieping</creatorcontrib><description>Few-shot learning (FSL) has achieved promising performance in hyperspectral image classification (HSIC) with few labeled samples by designing a proper embedding feature extractor. However, the performance of embedding feature extractors relies on the design of efficient deep convolutional neural network (CNN) architectures, which heavily depends on expertise knowledge. Particularly, FSL requires extracting discriminative features effectively across different domains, which makes the construction even more challenging. In this article, we propose a novel neural architecture search-based FSL model for HSI classification (HCFSL-NAS). Three novel strategies are proposed in this work. First, a neural architecture search (NAS)-based embedding feature extractor is developed to the FSL in HSIC, whose search space includes a group of proposed multiscale convolutions with channel attention. Second, a multisource learning framework is employed to aggregate abundant heterogeneous and homogeneous source data, which enables the powerful generalization of the network to the HSIC with only few labeled samples. Finally, the pointwise-based cross-entropy (CE) loss and the pairwise-based adaptive sparse loss are jointly optimized to maximize interclass distance and minimize the distance within a class simultaneously. Experimental results on four publicly hyperspectral datasets demonstrate that HCFSL-NAS outperforms both the exiting FSL methods and supervised learning methods for HSI classification with only few labeled samples. The code is available at https://github.com/xh-captain/HCFSL-NAS .</description><identifier>ISSN: 0196-2892</identifier><identifier>EISSN: 1558-0644</identifier><identifier>DOI: 10.1109/TGRS.2024.3385478</identifier><identifier>CODEN: IGRSD2</identifier><language>eng</language><publisher>New York: IEEE</publisher><subject>Adaptation models ; Artificial neural networks ; Classification ; Computer architecture ; Data mining ; Distance ; Embedding ; Feature extraction ; Few-shot learning (FSL) ; hyperspectral image (HSI) classification ; Hyperspectral imaging ; Image classification ; Learning ; Machine learning ; multisource learning ; neural architecture search (NAS) ; Neural networks ; Searching ; Supervised learning ; Taxonomy ; Training</subject><ispartof>IEEE transactions on geoscience and remote sensing, 2024, Vol.62, p.1-15</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2024</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><cites>FETCH-LOGICAL-c246t-e9ab0f3949530ddfa079ca4cec480681bf296c53fa6121e772e9d90b2583e1dc3</cites><orcidid>0000-0001-7511-9418 ; 0000-0002-3812-6536 ; 0000-0002-7764-3616 ; 0000-0002-3922-3696</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/10493163$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,776,780,792,4010,27900,27901,27902,54733</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/10493163$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Xiao, Fen</creatorcontrib><creatorcontrib>Xiang, Han</creatorcontrib><creatorcontrib>Cao, Chunhong</creatorcontrib><creatorcontrib>Gao, Xieping</creatorcontrib><title>Neural Architecture Search-Based Few-Shot Learning for Hyperspectral Image Classification</title><title>IEEE transactions on geoscience and remote sensing</title><addtitle>TGRS</addtitle><description>Few-shot learning (FSL) has achieved promising performance in hyperspectral image classification (HSIC) with few labeled samples by designing a proper embedding feature extractor. However, the performance of embedding feature extractors relies on the design of efficient deep convolutional neural network (CNN) architectures, which heavily depends on expertise knowledge. Particularly, FSL requires extracting discriminative features effectively across different domains, which makes the construction even more challenging. In this article, we propose a novel neural architecture search-based FSL model for HSI classification (HCFSL-NAS). Three novel strategies are proposed in this work. First, a neural architecture search (NAS)-based embedding feature extractor is developed to the FSL in HSIC, whose search space includes a group of proposed multiscale convolutions with channel attention. Second, a multisource learning framework is employed to aggregate abundant heterogeneous and homogeneous source data, which enables the powerful generalization of the network to the HSIC with only few labeled samples. Finally, the pointwise-based cross-entropy (CE) loss and the pairwise-based adaptive sparse loss are jointly optimized to maximize interclass distance and minimize the distance within a class simultaneously. Experimental results on four publicly hyperspectral datasets demonstrate that HCFSL-NAS outperforms both the exiting FSL methods and supervised learning methods for HSI classification with only few labeled samples. The code is available at https://github.com/xh-captain/HCFSL-NAS .</description><subject>Adaptation models</subject><subject>Artificial neural networks</subject><subject>Classification</subject><subject>Computer architecture</subject><subject>Data mining</subject><subject>Distance</subject><subject>Embedding</subject><subject>Feature extraction</subject><subject>Few-shot learning (FSL)</subject><subject>hyperspectral image (HSI) classification</subject><subject>Hyperspectral imaging</subject><subject>Image classification</subject><subject>Learning</subject><subject>Machine learning</subject><subject>multisource learning</subject><subject>neural architecture search (NAS)</subject><subject>Neural networks</subject><subject>Searching</subject><subject>Supervised learning</subject><subject>Taxonomy</subject><subject>Training</subject><issn>0196-2892</issn><issn>1558-0644</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNpNkE1Lw0AQhhdRsFZ_gOAh4Dl19ivZPdZiP6Ao2HrwFLab2TalTeJugvTfm9AePA3v8D4z8BDySGFEKeiX9exzNWLAxIhzJUWqrsiASqliSIS4JgOgOomZ0uyW3IWwB6BC0nRAvt-x9eYQjb3dFQ3apvUYrdB0MX41AfNoir_xalc10bLblkW5jVzlo_mpRh_qDujpxdFsMZocTAiFK6xpiqq8JzfOHAI-XOaQfE3f1pN5vPyYLSbjZWyZSJoYtdmA41poySHPnYFUWyMsWqEgUXTjmE6s5M4klFFMU4Y617BhUnGkueVD8ny-W_vqp8XQZPuq9WX3MuMgAKSkXHUtem5ZX4Xg0WW1L47GnzIKWW8w6w1mvcHsYrBjns5MgYj_-kJzmnD-Bx7RbSU</recordid><startdate>2024</startdate><enddate>2024</enddate><creator>Xiao, Fen</creator><creator>Xiang, Han</creator><creator>Cao, Chunhong</creator><creator>Gao, Xieping</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7UA</scope><scope>8FD</scope><scope>C1K</scope><scope>F1W</scope><scope>FR3</scope><scope>H8D</scope><scope>H96</scope><scope>KR7</scope><scope>L.G</scope><scope>L7M</scope><orcidid>https://orcid.org/0000-0001-7511-9418</orcidid><orcidid>https://orcid.org/0000-0002-3812-6536</orcidid><orcidid>https://orcid.org/0000-0002-7764-3616</orcidid><orcidid>https://orcid.org/0000-0002-3922-3696</orcidid></search><sort><creationdate>2024</creationdate><title>Neural Architecture Search-Based Few-Shot Learning for Hyperspectral Image Classification</title><author>Xiao, Fen ; Xiang, Han ; Cao, Chunhong ; Gao, Xieping</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c246t-e9ab0f3949530ddfa079ca4cec480681bf296c53fa6121e772e9d90b2583e1dc3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Adaptation models</topic><topic>Artificial neural networks</topic><topic>Classification</topic><topic>Computer architecture</topic><topic>Data mining</topic><topic>Distance</topic><topic>Embedding</topic><topic>Feature extraction</topic><topic>Few-shot learning (FSL)</topic><topic>hyperspectral image (HSI) classification</topic><topic>Hyperspectral imaging</topic><topic>Image classification</topic><topic>Learning</topic><topic>Machine learning</topic><topic>multisource learning</topic><topic>neural architecture search (NAS)</topic><topic>Neural networks</topic><topic>Searching</topic><topic>Supervised learning</topic><topic>Taxonomy</topic><topic>Training</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Xiao, Fen</creatorcontrib><creatorcontrib>Xiang, Han</creatorcontrib><creatorcontrib>Cao, Chunhong</creatorcontrib><creatorcontrib>Gao, Xieping</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>CrossRef</collection><collection>Water Resources Abstracts</collection><collection>Technology Research Database</collection><collection>Environmental Sciences and Pollution Management</collection><collection>ASFA: Aquatic Sciences and Fisheries Abstracts</collection><collection>Engineering Research Database</collection><collection>Aerospace Database</collection><collection>Aquatic Science & Fisheries Abstracts (ASFA) 2: Ocean Technology, Policy & Non-Living Resources</collection><collection>Civil Engineering Abstracts</collection><collection>Aquatic Science & Fisheries Abstracts (ASFA) Professional</collection><collection>Advanced Technologies Database with Aerospace</collection><jtitle>IEEE transactions on geoscience and remote sensing</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Xiao, Fen</au><au>Xiang, Han</au><au>Cao, Chunhong</au><au>Gao, Xieping</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Neural Architecture Search-Based Few-Shot Learning for Hyperspectral Image Classification</atitle><jtitle>IEEE transactions on geoscience and remote sensing</jtitle><stitle>TGRS</stitle><date>2024</date><risdate>2024</risdate><volume>62</volume><spage>1</spage><epage>15</epage><pages>1-15</pages><issn>0196-2892</issn><eissn>1558-0644</eissn><coden>IGRSD2</coden><abstract>Few-shot learning (FSL) has achieved promising performance in hyperspectral image classification (HSIC) with few labeled samples by designing a proper embedding feature extractor. However, the performance of embedding feature extractors relies on the design of efficient deep convolutional neural network (CNN) architectures, which heavily depends on expertise knowledge. Particularly, FSL requires extracting discriminative features effectively across different domains, which makes the construction even more challenging. In this article, we propose a novel neural architecture search-based FSL model for HSI classification (HCFSL-NAS). Three novel strategies are proposed in this work. First, a neural architecture search (NAS)-based embedding feature extractor is developed to the FSL in HSIC, whose search space includes a group of proposed multiscale convolutions with channel attention. Second, a multisource learning framework is employed to aggregate abundant heterogeneous and homogeneous source data, which enables the powerful generalization of the network to the HSIC with only few labeled samples. Finally, the pointwise-based cross-entropy (CE) loss and the pairwise-based adaptive sparse loss are jointly optimized to maximize interclass distance and minimize the distance within a class simultaneously. Experimental results on four publicly hyperspectral datasets demonstrate that HCFSL-NAS outperforms both the exiting FSL methods and supervised learning methods for HSI classification with only few labeled samples. The code is available at https://github.com/xh-captain/HCFSL-NAS .</abstract><cop>New York</cop><pub>IEEE</pub><doi>10.1109/TGRS.2024.3385478</doi><tpages>15</tpages><orcidid>https://orcid.org/0000-0001-7511-9418</orcidid><orcidid>https://orcid.org/0000-0002-3812-6536</orcidid><orcidid>https://orcid.org/0000-0002-7764-3616</orcidid><orcidid>https://orcid.org/0000-0002-3922-3696</orcidid></addata></record> |
fulltext | fulltext_linktorsrc |
identifier | ISSN: 0196-2892 |
ispartof | IEEE transactions on geoscience and remote sensing, 2024, Vol.62, p.1-15 |
issn | 0196-2892 1558-0644 |
language | eng |
recordid | cdi_ieee_primary_10493163 |
source | IEEE Electronic Library (IEL) |
subjects | Adaptation models Artificial neural networks Classification Computer architecture Data mining Distance Embedding Feature extraction Few-shot learning (FSL) hyperspectral image (HSI) classification Hyperspectral imaging Image classification Learning Machine learning multisource learning neural architecture search (NAS) Neural networks Searching Supervised learning Taxonomy Training |
title | Neural Architecture Search-Based Few-Shot Learning for Hyperspectral Image Classification |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-15T16%3A38%3A52IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Neural%20Architecture%20Search-Based%20Few-Shot%20Learning%20for%20Hyperspectral%20Image%20Classification&rft.jtitle=IEEE%20transactions%20on%20geoscience%20and%20remote%20sensing&rft.au=Xiao,%20Fen&rft.date=2024&rft.volume=62&rft.spage=1&rft.epage=15&rft.pages=1-15&rft.issn=0196-2892&rft.eissn=1558-0644&rft.coden=IGRSD2&rft_id=info:doi/10.1109/TGRS.2024.3385478&rft_dat=%3Cproquest_RIE%3E3040055138%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=3040055138&rft_id=info:pmid/&rft_ieee_id=10493163&rfr_iscdi=true |