Learning Image Formation and Regularization in Unrolling AMP for Lensless Image Reconstruction

This paper proposes an unrolling learnable approximate message passing recurrent neural network (called ULAMP-Net) for lensless image reconstruction. By unrolling the optimization iterations, key modules and parameters are made learnable to achieve high reconstruction quality. Specifically, observat...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on computational imaging Jg. 8; S. 479 - 489
Hauptverfasser: Yang, Jingyu, Yin, Xiangjun, Zhang, Mengxi, Yue, Huihui, Cui, Xingyu, Yue, Huanjing
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Piscataway IEEE 2022
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
ISSN:2573-0436, 2333-9403
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract This paper proposes an unrolling learnable approximate message passing recurrent neural network (called ULAMP-Net) for lensless image reconstruction. By unrolling the optimization iterations, key modules and parameters are made learnable to achieve high reconstruction quality. Specifically, observation matrices are rectified on the fly through network learning to suppress systematic errors in the measurement of the point spread function. We devise a domain transformation structure to achieve a more powerful representation and propose a learnable multistage threshold function to accommodate a much richer family of priors with only a small amount of parameters. Finally, we introduce a multi-layer perceptron (MLP) module to enhance the input and an attention mechanism as an output module to refine the final results. Experimental results on display captured dataset and real scene data demonstrate that, compared with the state-of-the-art methods, our method achieves the best reconstruction quality with low computational complexity and the tiny model size on the testing set. Our code will be released in https://github.com/Xiangjun-TJU/ULAMP-NET .
AbstractList This paper proposes an unrolling learnable approximate message passing recurrent neural network (called ULAMP-Net) for lensless image reconstruction. By unrolling the optimization iterations, key modules and parameters are made learnable to achieve high reconstruction quality. Specifically, observation matrices are rectified on the fly through network learning to suppress systematic errors in the measurement of the point spread function. We devise a domain transformation structure to achieve a more powerful representation and propose a learnable multistage threshold function to accommodate a much richer family of priors with only a small amount of parameters. Finally, we introduce a multi-layer perceptron (MLP) module to enhance the input and an attention mechanism as an output module to refine the final results. Experimental results on display captured dataset and real scene data demonstrate that, compared with the state-of-the-art methods, our method achieves the best reconstruction quality with low computational complexity and the tiny model size on the testing set. Our code will be released in https://github.com/Xiangjun-TJU/ULAMP-NET .
Author Yue, Huihui
Yin, Xiangjun
Cui, Xingyu
Yue, Huanjing
Yang, Jingyu
Zhang, Mengxi
Author_xml – sequence: 1
  givenname: Jingyu
  orcidid: 0000-0002-7521-7920
  surname: Yang
  fullname: Yang, Jingyu
  email: yjy@tju.edu.cn
  organization: School of Electrical and Information Engineering, Tianjin University, Tianjin, China
– sequence: 2
  givenname: Xiangjun
  orcidid: 0000-0002-4829-9019
  surname: Yin
  fullname: Yin, Xiangjun
  email: yinxiangjun@tju.edu.cn
  organization: School of Electrical and Information Engineering, Tianjin University, Tianjin, China
– sequence: 3
  givenname: Mengxi
  surname: Zhang
  fullname: Zhang, Mengxi
  email: mengxizhang@tju.edu.cn
  organization: School of Electrical and Information Engineering, Tianjin University, Tianjin, China
– sequence: 4
  givenname: Huihui
  surname: Yue
  fullname: Yue, Huihui
  email: yuehuihui@tju.edu.cn
  organization: School of Electrical and Information Engineering, Tianjin University, Tianjin, China
– sequence: 5
  givenname: Xingyu
  surname: Cui
  fullname: Cui, Xingyu
  email: cuixingyu@tju.edu.cn
  organization: School of Electrical and Information Engineering, Tianjin University, Tianjin, China
– sequence: 6
  givenname: Huanjing
  orcidid: 0000-0003-2517-9783
  surname: Yue
  fullname: Yue, Huanjing
  email: huanjing.yue@tju.edu.cn
  organization: School of Electrical and Information Engineering, Tianjin University, Tianjin, China
BookMark eNp9kE1PwkAQhjcGExG5m3hp4rk4s7P9OhIiSoLRELjabMuUlJRd3C0H_fUWIR48eNrJ5nneybzXomesYSFuEUaIkD0sJ7ORBClHhCmqhC5EXxJRmCmgXjdHCYWgKL4SQ--3AIAqk5TGffE-Z-1MbTbBbKc3HEyt2-m2tibQZh0seHNotKu_Tl-1CVbG2aY58uOXt6CyLpiz8Q17fw5YcGmNb92hPCo34rLSjefh-R2I1fRxOXkO569Ps8l4HpYywzZMkbMCqZBxUqZUxajTkrTWqkJJRaJpLZVSqAAYNWOkU0SudMSgVKEqoIG4P-Xunf04sG_zrT04063Mu8yMMhmB7Kj4RJXOeu-4ysu6_TmtdbpucoT8WGfe1Zkf68zPdXYi_BH3rt5p9_mfcndSamb-xbMUMI6AvgG7LoFh
CODEN ITCIAJ
CitedBy_id crossref_primary_10_1109_TETCI_2024_3375022
crossref_primary_10_1145_3701731
crossref_primary_10_1109_TIM_2024_3375987
crossref_primary_10_1109_TCI_2023_3315853
Cites_doi 10.1364/OL.411228
10.1109/TMI.2021.3054167
10.1109/JSTSP.2020.2977507
10.1109/TSP.2017.2708040
10.1023/A:1007966830741
10.1109/TCI.2020.3010360
10.1109/ICIEA48937.2020.9248182
10.1364/OE.19.004294
10.1007/978-3-319-46475-6_43
10.1109/TCI.2016.2593662
10.1109/MSP.2016.2581921
10.1109/TPAMI.2020.2987489
10.1109/TPAMI.2020.3033882
10.1016/j.imavis.2020.103871
10.1364/OPTICA.6.001185
10.1364/ao.56.00f189
10.1038/s41377-020-00380-x
10.1007/978-3-030-60636-7_32
10.1162/NECO_a_00907
10.1364/AO.28.004996
10.1364/ao.58.000509
10.1109/ICCV.2019.00795
10.1109/TPAMI.2018.2873610
10.1364/OPTICA.397214
10.1109/TIP.2020.3044472
10.1007/978-3-319-24574-4_28
10.1364/OE.27.028075
10.1364/OE.424075
10.1109/CVPR.2018.00196
10.1109/ICASSP39728.2021.9414728
10.1364/OPTICA.4.001117
10.1364/OL.390810
10.1137/18M1169655
10.1126/sciadv.1701548
10.1364/OPTICA.5.000001
10.1109/WACV48630.2021.00045
10.1109/I2MTC50364.2021.9460040
10.1109/CVPR.2019.01257
10.1109/CVPR46437.2021.01595
10.1038/s41377-020-0289-9
10.1007/BF00171998
10.1109/JSTSP.2020.3037516
10.1109/TCI.2021.3114542
ContentType Journal Article
Copyright Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022
Copyright_xml – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022
DBID 97E
RIA
RIE
AAYXX
CITATION
7SC
8FD
JQ2
L7M
L~C
L~D
DOI 10.1109/TCI.2022.3181473
DatabaseName IEEE Xplore (IEEE)
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE Electronic Library (IEL)
CrossRef
Computer and Information Systems 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
Computer and Information Systems Abstracts
Technology Research Database
Computer and Information Systems Abstracts – Academic
Advanced Technologies Database with Aerospace
ProQuest Computer Science Collection
Computer and Information Systems Abstracts Professional
DatabaseTitleList
Computer and Information Systems Abstracts
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
EISSN 2333-9403
EndPage 489
ExternalDocumentID 10_1109_TCI_2022_3181473
9801650
Genre orig-research
GrantInformation_xml – fundername: National Natural Science Foundation of China
  grantid: 62072331; 61771339
  funderid: 10.13039/501100001809
GroupedDBID 0R~
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABJNI
ABQJQ
ABVLG
ACGFS
AGQYO
AGSQL
AHBIQ
AKJIK
AKQYR
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
EBS
EJD
IFIPE
IPLJI
JAVBF
O9-
OCL
PQQKQ
RIA
RIE
AAYXX
CITATION
7SC
8FD
JQ2
L7M
L~C
L~D
ID FETCH-LOGICAL-c291t-81e9b13b267c83f61a8c3aaa4f123b7a3d24441400e1ae15a811efa5e044b4f03
IEDL.DBID RIE
ISICitedReferencesCount 8
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000838376400001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2573-0436
IngestDate Mon Jun 30 05:12:10 EDT 2025
Sat Nov 29 05:06:23 EST 2025
Tue Nov 18 21:46:37 EST 2025
Wed Aug 27 02:23:54 EDT 2025
IsPeerReviewed false
IsScholarly true
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-c291t-81e9b13b267c83f61a8c3aaa4f123b7a3d24441400e1ae15a811efa5e044b4f03
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ORCID 0000-0003-2517-9783
0000-0002-7521-7920
0000-0002-4829-9019
PQID 2679392502
PQPubID 2040412
PageCount 11
ParticipantIDs crossref_citationtrail_10_1109_TCI_2022_3181473
ieee_primary_9801650
crossref_primary_10_1109_TCI_2022_3181473
proquest_journals_2679392502
PublicationCentury 2000
PublicationDate 20220000
2022-00-00
20220101
PublicationDateYYYYMMDD 2022-01-01
PublicationDate_xml – year: 2022
  text: 20220000
PublicationDecade 2020
PublicationPlace Piscataway
PublicationPlace_xml – name: Piscataway
PublicationTitle IEEE transactions on computational imaging
PublicationTitleAbbrev TCI
PublicationYear 2022
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
ref35
ref12
ref34
ref15
Stork (ref20) 2013
ref14
ref31
ref30
ref11
ref33
ref10
ref32
Diederik (ref43) 2014
ref2
ref1
ref17
ref39
ref16
ref38
ref19
ref18
Busboom (ref36) 1998; 8
ref24
ref46
ref23
ref45
ref26
ref25
ref42
ref41
ref22
ref44
ref21
ref28
ref27
Amir (ref37) 2007
ref29
ref8
ref7
ref9
ref4
ref3
ref6
ref5
ref40
References_xml – ident: ref12
  doi: 10.1364/OL.411228
– start-page: 186
  volume-title: Proc. 7th Int. Conf. Sensor Technol. Appl.
  year: 2013
  ident: ref20
  article-title: Lensless ultra-miniature CMOS computational imagers and sensors
– ident: ref45
  doi: 10.1109/TMI.2021.3054167
– ident: ref26
  doi: 10.1109/JSTSP.2020.2977507
– ident: ref38
  doi: 10.1109/TSP.2017.2708040
– volume: 8
  start-page: 97
  year: 1998
  ident: ref36
  article-title: Uniformly redundant arrays
  publication-title: Exp. Astron.
  doi: 10.1023/A:1007966830741
– ident: ref18
  doi: 10.1109/TCI.2020.3010360
– ident: ref33
  doi: 10.1109/ICIEA48937.2020.9248182
– ident: ref21
  doi: 10.1364/OE.19.004294
– ident: ref42
  doi: 10.1007/978-3-319-46475-6_43
– ident: ref9
  doi: 10.1109/TCI.2016.2593662
– ident: ref3
  doi: 10.1109/MSP.2016.2581921
– ident: ref10
  doi: 10.1109/TPAMI.2020.2987489
– ident: ref14
  doi: 10.1109/TPAMI.2020.3033882
– ident: ref46
  doi: 10.1016/j.imavis.2020.103871
– ident: ref6
  doi: 10.1364/OPTICA.6.001185
– ident: ref4
  doi: 10.1364/ao.56.00f189
– ident: ref7
  doi: 10.1038/s41377-020-00380-x
– ident: ref27
  doi: 10.1007/978-3-030-60636-7_32
– ident: ref41
  doi: 10.1162/NECO_a_00907
– ident: ref2
  doi: 10.1364/AO.28.004996
– ident: ref5
  doi: 10.1364/ao.58.000509
– ident: ref1
  doi: 10.1109/ICCV.2019.00795
– ident: ref29
  doi: 10.1109/TPAMI.2018.2873610
– ident: ref22
  doi: 10.1364/OPTICA.397214
– ident: ref39
  doi: 10.1109/TIP.2020.3044472
– ident: ref44
  doi: 10.1007/978-3-319-24574-4_28
– ident: ref16
  doi: 10.1364/OE.27.028075
– start-page: 1
  volume-title: Proc. 3rd Int. Conf. Learn. Representations
  year: 2014
  ident: ref43
  article-title: Adam: A method for stochastic optimization
– ident: ref35
  doi: 10.1364/OE.424075
– ident: ref40
  doi: 10.1109/CVPR.2018.00196
– ident: ref23
  doi: 10.1109/ICASSP39728.2021.9414728
– ident: ref32
  doi: 10.1364/OPTICA.4.001117
– ident: ref15
  doi: 10.1364/OL.390810
– ident: ref24
  doi: 10.1137/18M1169655
– ident: ref19
  doi: 10.1126/sciadv.1701548
– start-page: 553
  volume-title: Proc. Int. Conf. Neural Inf. Process. Syst.
  year: 2007
  ident: ref37
  article-title: Fixing max-product: Convergent message passing algorithms for MAP LP-Relaxations
– ident: ref11
  doi: 10.1364/OPTICA.5.000001
– ident: ref13
  doi: 10.1109/WACV48630.2021.00045
– ident: ref25
  doi: 10.1109/I2MTC50364.2021.9460040
– ident: ref28
  doi: 10.1109/CVPR.2019.01257
– ident: ref30
  doi: 10.1109/CVPR46437.2021.01595
– ident: ref8
  doi: 10.1038/s41377-020-0289-9
– ident: ref17
  doi: 10.1007/BF00171998
– ident: ref31
  doi: 10.1109/JSTSP.2020.3037516
– ident: ref34
  doi: 10.1109/TCI.2021.3114542
SSID ssj0001492386
Score 2.2297816
Snippet This paper proposes an unrolling learnable approximate message passing recurrent neural network (called ULAMP-Net) for lensless image reconstruction. By...
SourceID proquest
crossref
ieee
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 479
SubjectTerms approximate message passing algorithm
Cameras
Current measurement
deep unfolding method
Image reconstruction
Imaging
Learning
Lensless imaging
Measurement uncertainty
Message passing
Modules
Multilayer perceptrons
Multilayers
Optimization
Parameters
Point spread functions
Prototypes
Recurrent neural networks
Regularization
spatial-channel attention
Systematic errors
Training
Title Learning Image Formation and Regularization in Unrolling AMP for Lensless Image Reconstruction
URI https://ieeexplore.ieee.org/document/9801650
https://www.proquest.com/docview/2679392502
Volume 8
WOSCitedRecordID wos000838376400001&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: 2333-9403
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0001492386
  issn: 2573-0436
  databaseCode: RIE
  dateStart: 20150101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PS8MwFA46PHhx_sT5ixy8CNY1TbI0xyEOBR1DJniyvKapCNrJNv37fUnTDVEEbz0koc2XvPe-pO97hJwmEtCNCBtJ4HkkZK4iSCSPCgW5Ait7mhe-2IQaDtPHRz1aIeeLXBhrrf_5zF64R3-XX0zMhzsq6-rUJd8gQV9VStW5WsvzFKc05gs74iLkkVNWb24lY90dX94gF0wSpKgpE4p_80K-rMoPW-wdzKD9v1fbJBshkKT9GvktsmKrbdIOQSUNW3a2Q56CguozvXlD20EHTbYihaqg974U_TQkY9KXij5U01qnm_bvRhRDWnqLTPcVDWIYwBHWpezsLnkYXI0vr6NQVCEyiWbzKGVW54znSU-ZlJc9BqnhACBK9GEIDy_Q4QukXbFlYJmElDFbgrSxELkoY75HWtWksvuEllqD0BAbBUiqlNKGldZVkIHSsCLXHdJtJjkzQXHcFb54zTzziHWGsGQOlizA0iFnix7vtdrGH213HAyLdgGBDjlqcMzCdpxl-LEaA0EZJwe_9zok627s-mzliLRwEu0xWTOf85fZ9MSvtC9wY9AM
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NTxsxEB1FtBJcSsuHSEtbH7ggsWS9tuP1EaFGRE0ihILEidWs14uQYFMlgd_fsdcJqkBIve3B3g8_2zNvvPMG4ChTSGZEukShKBOpSp1gpkRSaSw1OtU3ogrFJvRkkt_cmMsOnKxzYZxz4eczd-ovw1l-NbNPPlTWM7lPviGC_kFJmfE2W-slouK1xkJpR5qGIvHa6qtzydT0pudDYoNZRiQ151KLf-xQKKzyajcOJmaw_X8v9xk-RVeSnbXYf4GOa3ZgO7qVLC7axS7cRg3VOzZ8pN2DDVb5igybil2FYvTzmI7J7ht23cxbpW52Nr5k5NSyEXHdB9oS4w08ZX0Rnt2D68Gv6flFEssqJDYzfJnk3JmSizLra5uLus8xtwIRZU1WjAASFZl8ScQrdRwdV5hz7mpULpWylHUq9mGjmTXuAFhtDEqDqdVItEprY3ntfA0ZrC2vStOF3mqQCxs1x33pi4cicI_UFARL4WEpIixdOF73-NPqbbzTdtfDsG4XEejC4QrHIi7IRUEfa8gVVGn29e1eP2HzYjoeFaPh5Pc32PLPaSMth7BBA-q-w0f7vLxfzH-EWfcXbHzTUw
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=Learning+Image+Formation+and+Regularization+in+Unrolling+AMP+for+Lensless+Image+Reconstruction&rft.jtitle=IEEE+transactions+on+computational+imaging&rft.au=Yang%2C+Jingyu&rft.au=Yin%2C+Xiangjun&rft.au=Zhang%2C+Mengxi&rft.au=Yue%2C+Huihui&rft.date=2022&rft.issn=2573-0436&rft.eissn=2333-9403&rft.volume=8&rft.spage=479&rft.epage=489&rft_id=info:doi/10.1109%2FTCI.2022.3181473&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TCI_2022_3181473
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2573-0436&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2573-0436&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2573-0436&client=summon