Learned Two-Step Iterative Shrinkage Thresholding Algorithm for Deep Compressive Sensing

Deep unrolling architectures have revitalized compressive sensing (CS) by seamlessly blending deep neural networks with traditional optimization-based reconstruction algorithms. In pursuit of an efficient and deep interpretable approach, we propose LTwIST for CS problem, a novel deep unrolling frame...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	IEEE transactions on circuits and systems for video technology Ročník 34; číslo 5; s. 3943 - 3956
Hlavní autori:	Gan, Hongping, Wang, Xiaoyang, He, Lijun, Liu, Jie
Médium:	Journal Article
Jazyk:	English
Vydavateľské údaje:	New York IEEE 01.05.2024 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Predmet:	Algorithms Artificial neural networks Compressive sensing deep learning Image quality Image reconstruction Iterative algorithms Magnetic resonance imaging Matching pursuit algorithms Optimization Parameters Quality assessment Recovery Regeneration Sensors Transformers TwIST
ISSN:	1051-8215, 1558-2205
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Abstract	Deep unrolling architectures have revitalized compressive sensing (CS) by seamlessly blending deep neural networks with traditional optimization-based reconstruction algorithms. In pursuit of an efficient and deep interpretable approach, we propose LTwIST for CS problem, a novel deep unrolling framework that draws inspiration from the well-known two-step iterative shrinkage thresholding (TwIST) algorithm. LTwIST uses a trainable sensing matrix to adaptively learn structural information in images, and introduces a customized U-block architecture to solve the proximal mapping of nonlinear transformations connected with the sparsity-inducing regularizer. Specifically, each iteration recovery step of LTwIST corresponds to an iterative update step of the traditional TwIST algorithm. Moreover, the proposed method is designed to learn all the parameters end-to-end without manual tuning such as shrinkable thresholds, step sizes, etc. As a result, LTwIST obviates the need for manual parameter optimization, allows for high-quality image recovery and provides unambiguous interpretability. Moreover, our proposed LTwIST is also applicable to CS-based magnetic resonance imaging and exhibits a strong reconstruction performance. Extensive experiments on several public benchmark datasets demonstrate that the proposed LTwIST outperforms existing state-of-the-art deep CS methods by considerable margins in terms of quality evaluation metrics and visual performance. Our code is available on LTwIST.
AbstractList	Deep unrolling architectures have revitalized compressive sensing (CS) by seamlessly blending deep neural networks with traditional optimization-based reconstruction algorithms. In pursuit of an efficient and deep interpretable approach, we propose LTwIST for CS problem, a novel deep unrolling framework that draws inspiration from the well-known two-step iterative shrinkage thresholding (TwIST) algorithm. LTwIST uses a trainable sensing matrix to adaptively learn structural information in images, and introduces a customized U-block architecture to solve the proximal mapping of nonlinear transformations connected with the sparsity-inducing regularizer. Specifically, each iteration recovery step of LTwIST corresponds to an iterative update step of the traditional TwIST algorithm. Moreover, the proposed method is designed to learn all the parameters end-to-end without manual tuning such as shrinkable thresholds, step sizes, etc. As a result, LTwIST obviates the need for manual parameter optimization, allows for high-quality image recovery and provides unambiguous interpretability. Moreover, our proposed LTwIST is also applicable to CS-based magnetic resonance imaging and exhibits a strong reconstruction performance. Extensive experiments on several public benchmark datasets demonstrate that the proposed LTwIST outperforms existing state-of-the-art deep CS methods by considerable margins in terms of quality evaluation metrics and visual performance. Our code is available on LTwIST.
Author	Liu, Jie Wang, Xiaoyang He, Lijun Gan, Hongping
Author_xml	– sequence: 1 givenname: Hongping orcidid: 0000-0002-4853-5077 surname: Gan fullname: Gan, Hongping email: ganhongping@nwpu.edu.cn organization: School of Software, Northwestern Polytechnical University (NPU), Xi'an, China – sequence: 2 givenname: Xiaoyang orcidid: 0009-0001-5599-1740 surname: Wang fullname: Wang, Xiaoyang email: wxyang@mail.nwpu.edu.cn organization: School of Software, Northwestern Polytechnical University (NPU), Xi'an, China – sequence: 3 givenname: Lijun orcidid: 0000-0001-9000-845X surname: He fullname: He, Lijun email: lijunhe@nwpu.edu.cn organization: School of Software, Northwestern Polytechnical University (NPU), Xi'an, China – sequence: 4 givenname: Jie orcidid: 0000-0003-2587-9210 surname: Liu fullname: Liu, Jie email: lucky_jiel@nwpu.edu.cn organization: School of Software, Northwestern Polytechnical University (NPU), Xi'an, China
BookMark	eNp9kLtOw0AQRVcIJJLADyAKS9QO-7R3y8i8IkWiiEF01toexw6O1-w6IP6ezaNAFFRzi3tmNGeMTjvTAUJXBE8Jweo2TZav6ZRiyqaMUcE4PkEjIoQMKcXi1GcsSCgpEedo7NwaY8Ilj0fobQHadlAG6ZcJlwP0wXwAq4fmE4JlbZvuXa8gSGsLrjZt2XSrYNaujG2GehNUxgZ34JnEbHrfcHsKOudrF-is0q2Dy-OcoJeH-zR5ChfPj_NktggLqqIhVAVQlTNKS6lFXBWKMx4xDVEeF5znERY-lFBwwVUOXGMsNeFQMVGRmBPOJujmsLe35mMLbsjWZms7fzJjWDCqpJLMt-ShVVjjnIUqK5rBf2m6weqmzQjOdh6zvcds5zE7evQo_YP2ttlo-_0_dH2AGgD4BVAZiYixH0zxgS8
CODEN	ITCTEM
CitedBy_id	crossref_primary_10_1016_j_patcog_2025_111974 crossref_primary_10_1109_JSEN_2025_3565292 crossref_primary_10_1109_TCSVT_2024_3399764 crossref_primary_10_1109_TCSVT_2024_3397012 crossref_primary_10_1007_s40747_025_01963_0 crossref_primary_10_1109_JSEN_2025_3583291 crossref_primary_10_1109_LGRS_2025_3547408 crossref_primary_10_1016_j_dsp_2025_105002 crossref_primary_10_1109_TCSVT_2025_3543824 crossref_primary_10_1016_j_dsp_2025_105413 crossref_primary_10_1109_TETCI_2024_3437171 crossref_primary_10_1109_TIM_2025_3554865 crossref_primary_10_1016_j_cmpb_2025_108995 crossref_primary_10_1109_TCI_2025_3572286 crossref_primary_10_1109_LSP_2025_3547666 crossref_primary_10_1109_JIOT_2025_3580548 crossref_primary_10_1016_j_patcog_2025_112022 crossref_primary_10_1109_TIP_2025_3533198 crossref_primary_10_3390_jimaging11050139
Cites_doi	10.1109/TCSVT.2016.2587398 10.1109/MSP.2007.914728 10.1109/TCSVT.2018.2879983 10.1109/TCSVT.2016.2540073 10.1109/ACSSC.2008.5074472 10.1109/ICASSP.2008.4518498 10.1007/978-3-319-10602-1_48 10.1109/TCSVT.2018.2886310 10.1109/CVPR.2018.00196 10.1109/CVPR52688.2022.01688 10.1145/3065386 10.1109/TCSVT.2023.3261542 10.1109/TIP.2020.3015545 10.1109/TCI.2019.2948732 10.1109/TIP.2019.2928136 10.1109/TIP.2020.3044472 10.1109/ICCV.2001.937655 10.1007/978-3-319-46475-6_25 10.1109/TIT.2016.2556683 10.1109/TIP.2006.881959 10.1137/080716542 10.1109/TIT.2006.871582 10.1007/978-3-319-24574-4_28 10.1016/j.neucom.2019.05.006 10.1109/CVPR.2016.55 10.1109/JSTSP.2020.2977507 10.1109/TCI.2023.3244396 10.1117/1.3600632 10.1109/TCSVT.2019.2898908 10.1109/CVPR.2018.00070 10.1109/TIT.2011.2177632 10.5244/C.26.135 10.1109/TIP.2007.909319 10.1109/TIP.2015.2419084 10.1109/MSP.2022.3217936 10.1109/TCSVT.2017.2697972 10.1137/S003614450037906X 10.1109/MSP.2019.2953993 10.1109/TIP.2010.2076294 10.1109/TMI.2017.2760978 10.1109/TCYB.2021.3127657 10.1109/TIP.2020.3023629 10.1002/jmri.1183 10.1109/TPAMI.2018.2883941 10.1609/aaai.v32i1.11869 10.1109/TCI.2023.3241551 10.1109/TCSVT.2016.2527181 10.1109/TIP.2023.3263100 10.1109/TIP.2021.3086049 10.48550/ARXIV.1706.03762 10.1109/TIP.2020.3005515 10.1109/TIP.2022.3217365 10.1109/TCSVT.2019.2935127 10.1109/TCSVT.2019.2939370
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2024
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2024
DBID	97E RIA RIE AAYXX CITATION 7SC 7SP 8FD JQ2 L7M L~C L~D
DOI	10.1109/TCSVT.2023.3325340
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef Computer and Information Systems Abstracts Electronics & Communications Abstracts Technology Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional
DatabaseTitle	CrossRef Technology Research Database Computer and Information Systems Abstracts – Academic Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Professional
DatabaseTitleList	Technology Research Database
Database_xml	– sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
EISSN	1558-2205
EndPage	3956
ExternalDocumentID	10_1109_TCSVT_2023_3325340 10286563
Genre	orig-research
GrantInformation_xml	– fundername: Basic Research Programs of Taicang grantid: TC2022JC19 – fundername: National Natural Science Foundation of China grantid: 62101455; 62201463 funderid: 10.13039/501100001809 – fundername: Fundamental Research Funds for the Central Universities grantid: G2022WD01007 funderid: 10.13039/501100012226 – fundername: Natural Science Basic Research Program of Shaanxi Province grantid: 2022JQ-615 funderid: 10.13039/501100017596
GroupedDBID	-~X 0R~ 29I 4.4 5GY 5VS 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACGFO ACGFS ACIWK AENEX AETIX AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD HZ~ H~9 ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 O9- OCL P2P RIA RIE RNS RXW TAE TN5 VH1 AAYXX CITATION 7SC 7SP 8FD JQ2 L7M L~C L~D
ID	FETCH-LOGICAL-c296t-9ce29b322d8a57fc943463ae6b7c44b605b7cdec4549be4a008a14ef35f174143
IEDL.DBID	RIE
ISICitedReferencesCount	25
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001221132000060&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	1051-8215
IngestDate	Mon Sep 01 14:14:55 EDT 2025 Sat Nov 29 01:44:25 EST 2025 Tue Nov 18 21:37:44 EST 2025 Wed Aug 27 02:05:25 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	5
Language	English
License	https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c296t-9ce29b322d8a57fc943463ae6b7c44b605b7cdec4549be4a008a14ef35f174143
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ORCID	0000-0001-9000-845X 0000-0003-2587-9210 0009-0001-5599-1740 0000-0002-4853-5077
PQID	3053298983
PQPubID	85433
PageCount	14
ParticipantIDs	crossref_citationtrail_10_1109_TCSVT_2023_3325340 crossref_primary_10_1109_TCSVT_2023_3325340 ieee_primary_10286563 proquest_journals_3053298983
PublicationCentury	2000
PublicationDate	2024-05-01
PublicationDateYYYYMMDD	2024-05-01
PublicationDate_xml	– month: 05 year: 2024 text: 2024-05-01 day: 01
PublicationDecade	2020
PublicationPlace	New York
PublicationPlace_xml	– name: New York
PublicationTitle	IEEE transactions on circuits and systems for video technology
PublicationTitleAbbrev	TCSVT
PublicationYear	2024
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref13 ref57 ref12 ref56 ref15 ref14 ref58 ref53 ref52 ref11 ref55 ref10 ref54 Zbontar (ref59) 2018 ref17 ref16 ref19 ref18 Metzler (ref47); 30 ref51 ref50 ref46 ref45 ref48 ref41 ref44 ref43 ref49 ref8 Wu (ref42); 97 ref7 ref9 ref4 ref3 ref6 ref5 ref40 ref35 ref34 ref37 ref36 ref31 ref30 ref33 ref32 ref1 Babacan (ref24) ref39 ref38 ref23 Li (ref2) 2010 ref26 ref25 ref20 ref22 ref21 ref28 ref27 ref29
References_xml	– ident: ref15 doi: 10.1109/TCSVT.2016.2587398 – ident: ref40 doi: 10.1109/MSP.2007.914728 – ident: ref4 doi: 10.1109/TCSVT.2018.2879983 – ident: ref12 doi: 10.1109/TCSVT.2016.2540073 – ident: ref22 doi: 10.1109/ACSSC.2008.5074472 – ident: ref25 doi: 10.1109/ICASSP.2008.4518498 – ident: ref51 doi: 10.1007/978-3-319-10602-1_48 – ident: ref13 doi: 10.1109/TCSVT.2018.2886310 – ident: ref33 doi: 10.1109/CVPR.2018.00196 – ident: ref49 doi: 10.1109/CVPR52688.2022.01688 – ident: ref58 doi: 10.1145/3065386 – ident: ref7 doi: 10.1109/TCSVT.2023.3261542 – ident: ref10 doi: 10.1109/TIP.2020.3015545 – volume: 97 start-page: 6850 volume-title: Proc. 36th Int. Conf. Mach. Learn. ident: ref42 article-title: Deep compressed sensing – ident: ref34 doi: 10.1109/TCI.2019.2948732 – ident: ref31 doi: 10.1109/TIP.2019.2928136 – ident: ref45 doi: 10.1109/TIP.2020.3044472 – volume: 30 start-page: 1 volume-title: Proc. Adv. Neural Inf. Process. Syst. (NIPS) ident: ref47 article-title: Learned D-AMP: Principled neural network based compressive image recovery – ident: ref55 doi: 10.1109/ICCV.2001.937655 – ident: ref56 doi: 10.1007/978-3-319-46475-6_25 – volume-title: An Efficient Algorithm for Total Variation Regularization With Applications to the Single Pixel Camera and Compressive Sensing year: 2010 ident: ref2 – ident: ref20 doi: 10.1109/TIT.2016.2556683 – ident: ref54 doi: 10.1109/TIP.2006.881959 – ident: ref19 doi: 10.1137/080716542 – ident: ref1 doi: 10.1109/TIT.2006.871582 – ident: ref50 doi: 10.1007/978-3-319-24574-4_28 – ident: ref41 doi: 10.1016/j.neucom.2019.05.006 – ident: ref30 doi: 10.1109/CVPR.2016.55 – ident: ref35 doi: 10.1109/JSTSP.2020.2977507 – ident: ref32 doi: 10.1109/TCI.2023.3244396 – ident: ref53 doi: 10.1117/1.3600632 – ident: ref9 doi: 10.1109/TCSVT.2019.2898908 – ident: ref57 doi: 10.1109/CVPR.2018.00070 – ident: ref23 doi: 10.1109/TIT.2011.2177632 – ident: ref52 doi: 10.5244/C.26.135 – ident: ref28 doi: 10.1109/TIP.2007.909319 – ident: ref26 doi: 10.1109/TIP.2015.2419084 – ident: ref17 doi: 10.1109/MSP.2022.3217936 – ident: ref27 doi: 10.1109/TCSVT.2017.2697972 – ident: ref18 doi: 10.1137/S003614450037906X – ident: ref6 doi: 10.1109/MSP.2019.2953993 – ident: ref21 doi: 10.1109/TIP.2010.2076294 – start-page: 110 volume-title: Proc. 17th Eur. Signal Process. Conf. ident: ref24 article-title: Non-convex priors in Bayesian compressed sensing – ident: ref36 doi: 10.1109/TMI.2017.2760978 – ident: ref44 doi: 10.1109/TCYB.2021.3127657 – ident: ref43 doi: 10.1109/TIP.2020.3023629 – ident: ref39 doi: 10.1002/jmri.1183 – ident: ref37 doi: 10.1109/TPAMI.2018.2883941 – ident: ref38 doi: 10.1609/aaai.v32i1.11869 – ident: ref3 doi: 10.1109/TCI.2023.3241551 – ident: ref11 doi: 10.1109/TCSVT.2016.2527181 – ident: ref48 doi: 10.1109/TIP.2023.3263100 – ident: ref16 doi: 10.1109/TIP.2021.3086049 – ident: ref29 doi: 10.48550/ARXIV.1706.03762 – ident: ref8 doi: 10.1109/TIP.2020.3005515 – ident: ref46 doi: 10.1109/TIP.2022.3217365 – year: 2018 ident: ref59 article-title: FastMRI: An open dataset and benchmarks for accelerated MRI publication-title: arXiv:1811.08839 – ident: ref5 doi: 10.1109/TCSVT.2019.2935127 – ident: ref14 doi: 10.1109/TCSVT.2019.2939370
SSID	ssj0014847
Score	2.552114
Snippet	Deep unrolling architectures have revitalized compressive sensing (CS) by seamlessly blending deep neural networks with traditional optimization-based...
SourceID	proquest crossref ieee
SourceType	Aggregation Database Enrichment Source Index Database Publisher
StartPage	3943
SubjectTerms	Algorithms Artificial neural networks Compressive sensing deep learning Image quality Image reconstruction Iterative algorithms Magnetic resonance imaging Matching pursuit algorithms Optimization Parameters Quality assessment Recovery Regeneration Sensors Transformers TwIST
Title	Learned Two-Step Iterative Shrinkage Thresholding Algorithm for Deep Compressive Sensing
URI	https://ieeexplore.ieee.org/document/10286563 https://www.proquest.com/docview/3053298983
Volume	34
WOSCitedRecordID	wos001221132000060&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 1558-2205 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014847 issn: 1051-8215 databaseCode: RIE dateStart: 19910101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LSwMxEA62eNCDz4rVKjl4k_Sxye5mj6VaFKQIXaW3ZTc7fWDdSh_6953JbktBFLzlkAlLvmQem5lvGLvBKEPFsYlFql1HKDxTIkFDJnyH2NGM1kliSVyf_F5PDwbBc1GsbmthAMAmn0GdhvYtP52ZFf0qa5AxRP9DlljJ9728WGvzZKC07SaG_kJLaDRk6wqZZtAIO_3XsE6NwutSOq6kPx1bVsi2Vfmhi62B6R7-89OO2EHhSfJ2Dv0x24HshO1v8QuesoFlT4WUh18zQflc_NGSKKOG4_0xznlDbcJDhHNRvELx9nQ0m0-W43eO3iy_A5QhlWGzZUmK8t2zUYW9dO_DzoMoWikI4wTeUgQGnCDBy5vq2PWHhljhPBmDl_hGqQRjGhykYBSGiwmoGD2DuKVgKN0hhizoU52xcjbL4JxxpTGeBumAp-j268Q4Xpriuk3XBNBSVdZab21kCp5xancxjWy80QwiC0dEcEQFHFV2u5H5yFk2_pxdIQC2ZuZ7X2W1NYRRcRMXkaTWF9QkU178InbJ9nB1lWcx1lh5OV_BFds1n8vJYn5tD9k3HLHNng
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT8MwDLZ4ScCB5xCDATlwQxlbk3bpEQETiDEhUdBuVZt6DzE2tA34-9hZhyYhkLj1ELdVnPiR2N8HcEpZhk4Sm8jM-J7UtKZkSo5M1jxGR7PGpKkDcW3Umk3TaoUPebO664VBRFd8hmV-dHf52dC-81HZOTtDij_UIiwzdVbervV9aaCN4xOjiKEqDbmyWY9MJTyPLh-fozJThZeV8nzFZx1zfsgRq_ywxs7F1Df_-XNbsJHHkuJiqvxtWMDBDqzPIQzuQsvhp2Imos-h5IoucetglMnGiccujXkheyIiUug4v4cSF_3OcNSbdF8FxbPiCkmGjYarl2UprngfdArwVL-OLm9kTqYgrRcGExla9MKUtm9mEr_WtowLF6gEg7RmtU4pq6GHDK2mhDFFnVBskFQ1tpXfpqSFoqo9WBoMB7gPQhvKqFF5GGje_ya1XpBl9N6Kb0Os6iJUZ1Mb2xxpnAkv-rHLOCph7NQRszriXB1FOPuWeZvibPw5usAKmBs5nfsilGYqjPO9OI4Vk18wTaY6-EXsBFZvovtG3Lht3h3CGn1JT2saS7A0Gb3jEazYj0lvPDp2C-4L_ZvQ5w
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Learned+Two-Step+Iterative+Shrinkage+Thresholding+Algorithm+for+Deep+Compressive+Sensing&rft.jtitle=IEEE+transactions+on+circuits+and+systems+for+video+technology&rft.au=Gan%2C+Hongping&rft.au=Wang%2C+Xiaoyang&rft.au=He%2C+Lijun&rft.au=Liu%2C+Jie&rft.date=2024-05-01&rft.issn=1051-8215&rft.eissn=1558-2205&rft.volume=34&rft.issue=5&rft.spage=3943&rft.epage=3956&rft_id=info:doi/10.1109%2FTCSVT.2023.3325340&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TCSVT_2023_3325340
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1051-8215&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1051-8215&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1051-8215&client=summon