Loading…

Feedback Attention-Based Dense CNN for Hyperspectral Image Classification

Hyperspectral image classification (HSIC) methods based on convolutional neural network (CNN) continue to progress in recent years. However, high complexity, information redundancy, and inefficient description still are the main barriers to the current HSIC networks. To address the mentioned problem...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on geoscience and remote sensing 2022, Vol.60, p.1-16
Main Authors: Yu, Chunyan, Han, Rui, Song, Meiping, Liu, Caiyu, Chang, Chein-I
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c293t-de87baf8fcbe8cab5f1b7e1e02be2545517d70506e24be968c3787d3d022f9f53
cites cdi_FETCH-LOGICAL-c293t-de87baf8fcbe8cab5f1b7e1e02be2545517d70506e24be968c3787d3d022f9f53
container_end_page 16
container_issue
container_start_page 1
container_title IEEE transactions on geoscience and remote sensing
container_volume 60
creator Yu, Chunyan
Han, Rui
Song, Meiping
Liu, Caiyu
Chang, Chein-I
description Hyperspectral image classification (HSIC) methods based on convolutional neural network (CNN) continue to progress in recent years. However, high complexity, information redundancy, and inefficient description still are the main barriers to the current HSIC networks. To address the mentioned problems, we present a spatial-spectral dense CNN framework with a feedback attention mechanism called FADCNN for HSIC in this article. The proposed architecture assembles the spectral-spatial feature in a compact connection style to extract sufficient information independently with two separate dense CNN networks. Specifically, the feedback attention modules are developed for the first time to enhance the attention map with the semantic knowledge from the high-level layer of the dense model, and we strengthen the spatial attention module by considering multiscale spatial information. To further improve the computation efficiency and the discrimination of the feature representation, the band attention module is designed to emphasize the weight of the bands that participated in the classification training. Besides, the spatial-spectral features are integrated and mined intensely for better refinement in the feature mining network. The extensive experimental results on real hyperspectral images (HSI) demonstrate that the proposed FADCNN architecture has significant advantages compared with other state-of-the-art methods.
doi_str_mv 10.1109/TGRS.2021.3058549
format article
fullrecord <record><control><sourceid>proquest_ieee_</sourceid><recordid>TN_cdi_ieee_primary_9361752</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9361752</ieee_id><sourcerecordid>2607876252</sourcerecordid><originalsourceid>FETCH-LOGICAL-c293t-de87baf8fcbe8cab5f1b7e1e02be2545517d70506e24be968c3787d3d022f9f53</originalsourceid><addsrcrecordid>eNo9kMtOwzAQRS0EEqXwAYhNJNYpthO_lqXQUqkqEpS1ZTtjlNImwU4X_XsctWI1izn3zuggdE_whBCsnjaLj88JxZRMCswkK9UFGhHGZI55WV6iESaK51Qqeo1uYtxiTEpGxAgt5wCVNe4nm_Y9NH3dNvmziVBlL9BEyGbrdebbkL0dOwixA9cHs8uWe_OddjsTY-1rZ4bYLbryZhfh7jzH6Gv-upm95av3xXI2XeWOqqLPK5DCGi-9syCdscwTK4AAphYoK1n6qhKYYQ60tKC4dIWQoioqTKlXnhVj9Hjq7UL7e4DY6217CE06qSnHieWU0USRE-VCG2MAr7tQ7004aoL1YEwPxvRgTJ-NpczDKVMDwD-vCk5EavwDrvJnAA</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2607876252</pqid></control><display><type>article</type><title>Feedback Attention-Based Dense CNN for Hyperspectral Image Classification</title><source>IEEE Electronic Library (IEL) Journals</source><creator>Yu, Chunyan ; Han, Rui ; Song, Meiping ; Liu, Caiyu ; Chang, Chein-I</creator><creatorcontrib>Yu, Chunyan ; Han, Rui ; Song, Meiping ; Liu, Caiyu ; Chang, Chein-I</creatorcontrib><description>Hyperspectral image classification (HSIC) methods based on convolutional neural network (CNN) continue to progress in recent years. However, high complexity, information redundancy, and inefficient description still are the main barriers to the current HSIC networks. To address the mentioned problems, we present a spatial-spectral dense CNN framework with a feedback attention mechanism called FADCNN for HSIC in this article. The proposed architecture assembles the spectral-spatial feature in a compact connection style to extract sufficient information independently with two separate dense CNN networks. Specifically, the feedback attention modules are developed for the first time to enhance the attention map with the semantic knowledge from the high-level layer of the dense model, and we strengthen the spatial attention module by considering multiscale spatial information. To further improve the computation efficiency and the discrimination of the feature representation, the band attention module is designed to emphasize the weight of the bands that participated in the classification training. Besides, the spatial-spectral features are integrated and mined intensely for better refinement in the feature mining network. The extensive experimental results on real hyperspectral images (HSI) demonstrate that the proposed FADCNN architecture has significant advantages compared with other state-of-the-art methods.</description><identifier>ISSN: 0196-2892</identifier><identifier>EISSN: 1558-0644</identifier><identifier>DOI: 10.1109/TGRS.2021.3058549</identifier><identifier>CODEN: IGRSD2</identifier><language>eng</language><publisher>New York: IEEE</publisher><subject>Artificial neural networks ; Attention map ; Classification ; Computation ; Computational modeling ; Computer architecture ; convolutional neural network (CNN) ; Data mining ; dense feature ; Feature extraction ; Feedback ; Frequency modulation ; hyperspectral image classification (HSIC) ; Hyperspectral imaging ; Image classification ; Information processing ; Methods ; Modules ; Neural networks ; Redundancy ; Spatial data ; spatial feature extraction ; Spectra ; spectral feature extraction ; Training</subject><ispartof>IEEE transactions on geoscience and remote sensing, 2022, Vol.60, p.1-16</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c293t-de87baf8fcbe8cab5f1b7e1e02be2545517d70506e24be968c3787d3d022f9f53</citedby><cites>FETCH-LOGICAL-c293t-de87baf8fcbe8cab5f1b7e1e02be2545517d70506e24be968c3787d3d022f9f53</cites><orcidid>0000-0002-5450-4891 ; 0000-0002-9260-6629</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9361752$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,780,784,4024,27923,27924,27925,54796</link.rule.ids></links><search><creatorcontrib>Yu, Chunyan</creatorcontrib><creatorcontrib>Han, Rui</creatorcontrib><creatorcontrib>Song, Meiping</creatorcontrib><creatorcontrib>Liu, Caiyu</creatorcontrib><creatorcontrib>Chang, Chein-I</creatorcontrib><title>Feedback Attention-Based Dense CNN for Hyperspectral Image Classification</title><title>IEEE transactions on geoscience and remote sensing</title><addtitle>TGRS</addtitle><description>Hyperspectral image classification (HSIC) methods based on convolutional neural network (CNN) continue to progress in recent years. However, high complexity, information redundancy, and inefficient description still are the main barriers to the current HSIC networks. To address the mentioned problems, we present a spatial-spectral dense CNN framework with a feedback attention mechanism called FADCNN for HSIC in this article. The proposed architecture assembles the spectral-spatial feature in a compact connection style to extract sufficient information independently with two separate dense CNN networks. Specifically, the feedback attention modules are developed for the first time to enhance the attention map with the semantic knowledge from the high-level layer of the dense model, and we strengthen the spatial attention module by considering multiscale spatial information. To further improve the computation efficiency and the discrimination of the feature representation, the band attention module is designed to emphasize the weight of the bands that participated in the classification training. Besides, the spatial-spectral features are integrated and mined intensely for better refinement in the feature mining network. The extensive experimental results on real hyperspectral images (HSI) demonstrate that the proposed FADCNN architecture has significant advantages compared with other state-of-the-art methods.</description><subject>Artificial neural networks</subject><subject>Attention map</subject><subject>Classification</subject><subject>Computation</subject><subject>Computational modeling</subject><subject>Computer architecture</subject><subject>convolutional neural network (CNN)</subject><subject>Data mining</subject><subject>dense feature</subject><subject>Feature extraction</subject><subject>Feedback</subject><subject>Frequency modulation</subject><subject>hyperspectral image classification (HSIC)</subject><subject>Hyperspectral imaging</subject><subject>Image classification</subject><subject>Information processing</subject><subject>Methods</subject><subject>Modules</subject><subject>Neural networks</subject><subject>Redundancy</subject><subject>Spatial data</subject><subject>spatial feature extraction</subject><subject>Spectra</subject><subject>spectral feature extraction</subject><subject>Training</subject><issn>0196-2892</issn><issn>1558-0644</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><recordid>eNo9kMtOwzAQRS0EEqXwAYhNJNYpthO_lqXQUqkqEpS1ZTtjlNImwU4X_XsctWI1izn3zuggdE_whBCsnjaLj88JxZRMCswkK9UFGhHGZI55WV6iESaK51Qqeo1uYtxiTEpGxAgt5wCVNe4nm_Y9NH3dNvmziVBlL9BEyGbrdebbkL0dOwixA9cHs8uWe_OddjsTY-1rZ4bYLbryZhfh7jzH6Gv-upm95av3xXI2XeWOqqLPK5DCGi-9syCdscwTK4AAphYoK1n6qhKYYQ60tKC4dIWQoioqTKlXnhVj9Hjq7UL7e4DY6217CE06qSnHieWU0USRE-VCG2MAr7tQ7004aoL1YEwPxvRgTJ-NpczDKVMDwD-vCk5EavwDrvJnAA</recordid><startdate>2022</startdate><enddate>2022</enddate><creator>Yu, Chunyan</creator><creator>Han, Rui</creator><creator>Song, Meiping</creator><creator>Liu, Caiyu</creator><creator>Chang, Chein-I</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/0000-0002-5450-4891</orcidid><orcidid>https://orcid.org/0000-0002-9260-6629</orcidid></search><sort><creationdate>2022</creationdate><title>Feedback Attention-Based Dense CNN for Hyperspectral Image Classification</title><author>Yu, Chunyan ; Han, Rui ; Song, Meiping ; Liu, Caiyu ; Chang, Chein-I</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c293t-de87baf8fcbe8cab5f1b7e1e02be2545517d70506e24be968c3787d3d022f9f53</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Artificial neural networks</topic><topic>Attention map</topic><topic>Classification</topic><topic>Computation</topic><topic>Computational modeling</topic><topic>Computer architecture</topic><topic>convolutional neural network (CNN)</topic><topic>Data mining</topic><topic>dense feature</topic><topic>Feature extraction</topic><topic>Feedback</topic><topic>Frequency modulation</topic><topic>hyperspectral image classification (HSIC)</topic><topic>Hyperspectral imaging</topic><topic>Image classification</topic><topic>Information processing</topic><topic>Methods</topic><topic>Modules</topic><topic>Neural networks</topic><topic>Redundancy</topic><topic>Spatial data</topic><topic>spatial feature extraction</topic><topic>Spectra</topic><topic>spectral feature extraction</topic><topic>Training</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Yu, Chunyan</creatorcontrib><creatorcontrib>Han, Rui</creatorcontrib><creatorcontrib>Song, Meiping</creatorcontrib><creatorcontrib>Liu, Caiyu</creatorcontrib><creatorcontrib>Chang, Chein-I</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) Online</collection><collection>IEEE/IET Electronic Library</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 &amp; Fisheries Abstracts (ASFA) 2: Ocean Technology, Policy &amp; Non-Living Resources</collection><collection>Civil Engineering Abstracts</collection><collection>Aquatic Science &amp; 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</fulltext></delivery><addata><au>Yu, Chunyan</au><au>Han, Rui</au><au>Song, Meiping</au><au>Liu, Caiyu</au><au>Chang, Chein-I</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Feedback Attention-Based Dense CNN for Hyperspectral Image Classification</atitle><jtitle>IEEE transactions on geoscience and remote sensing</jtitle><stitle>TGRS</stitle><date>2022</date><risdate>2022</risdate><volume>60</volume><spage>1</spage><epage>16</epage><pages>1-16</pages><issn>0196-2892</issn><eissn>1558-0644</eissn><coden>IGRSD2</coden><abstract>Hyperspectral image classification (HSIC) methods based on convolutional neural network (CNN) continue to progress in recent years. However, high complexity, information redundancy, and inefficient description still are the main barriers to the current HSIC networks. To address the mentioned problems, we present a spatial-spectral dense CNN framework with a feedback attention mechanism called FADCNN for HSIC in this article. The proposed architecture assembles the spectral-spatial feature in a compact connection style to extract sufficient information independently with two separate dense CNN networks. Specifically, the feedback attention modules are developed for the first time to enhance the attention map with the semantic knowledge from the high-level layer of the dense model, and we strengthen the spatial attention module by considering multiscale spatial information. To further improve the computation efficiency and the discrimination of the feature representation, the band attention module is designed to emphasize the weight of the bands that participated in the classification training. Besides, the spatial-spectral features are integrated and mined intensely for better refinement in the feature mining network. The extensive experimental results on real hyperspectral images (HSI) demonstrate that the proposed FADCNN architecture has significant advantages compared with other state-of-the-art methods.</abstract><cop>New York</cop><pub>IEEE</pub><doi>10.1109/TGRS.2021.3058549</doi><tpages>16</tpages><orcidid>https://orcid.org/0000-0002-5450-4891</orcidid><orcidid>https://orcid.org/0000-0002-9260-6629</orcidid></addata></record>
fulltext fulltext
identifier ISSN: 0196-2892
ispartof IEEE transactions on geoscience and remote sensing, 2022, Vol.60, p.1-16
issn 0196-2892
1558-0644
language eng
recordid cdi_ieee_primary_9361752
source IEEE Electronic Library (IEL) Journals
subjects Artificial neural networks
Attention map
Classification
Computation
Computational modeling
Computer architecture
convolutional neural network (CNN)
Data mining
dense feature
Feature extraction
Feedback
Frequency modulation
hyperspectral image classification (HSIC)
Hyperspectral imaging
Image classification
Information processing
Methods
Modules
Neural networks
Redundancy
Spatial data
spatial feature extraction
Spectra
spectral feature extraction
Training
title Feedback Attention-Based Dense CNN for Hyperspectral Image Classification
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T20%3A34%3A17IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_ieee_&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Feedback%20Attention-Based%20Dense%20CNN%20for%20Hyperspectral%20Image%20Classification&rft.jtitle=IEEE%20transactions%20on%20geoscience%20and%20remote%20sensing&rft.au=Yu,%20Chunyan&rft.date=2022&rft.volume=60&rft.spage=1&rft.epage=16&rft.pages=1-16&rft.issn=0196-2892&rft.eissn=1558-0644&rft.coden=IGRSD2&rft_id=info:doi/10.1109/TGRS.2021.3058549&rft_dat=%3Cproquest_ieee_%3E2607876252%3C/proquest_ieee_%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c293t-de87baf8fcbe8cab5f1b7e1e02be2545517d70506e24be968c3787d3d022f9f53%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=2607876252&rft_id=info:pmid/&rft_ieee_id=9361752&rfr_iscdi=true