Inference-Reconstruction Variational Autoencoder for Light Field Image Reconstruction

Light field cameras can capture the radiance and direction of light rays by a single exposure, providing a new perspective to photography and 3D geometry perception. However, existing sub-aperture based light field cameras are limited by their sensor resolution to obtain high spatial and angular res...

Full description

Saved in:

Bibliographic Details
Published in:	IEEE transactions on image processing Vol. 31; pp. 5629 - 5644
Main Authors:	Han, Kang, Xiang, Wei
Format:	Journal Article
Language:	English
Published:	New York IEEE 2022 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:	adaptive convolution Angular resolution Cameras Convolution Distance measurement Feature extraction Field cameras Image reconstruction indirect view synthesis Inference Information flow Light Light field image reconstruction Spatial resolution statistic distance measurement Superresolution Training variational autoencoder
ISSN:	1057-7149, 1941-0042, 1941-0042
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Abstract	Light field cameras can capture the radiance and direction of light rays by a single exposure, providing a new perspective to photography and 3D geometry perception. However, existing sub-aperture based light field cameras are limited by their sensor resolution to obtain high spatial and angular resolution images simultaneously. In this paper, we propose an inference-reconstruction variational autoencoder (IR-VAE) to reconstruct a dense light field image out of four corner reference views in a light field image. The proposed IR-VAE is comprised of one inference network and one reconstruction network, where the inference network infers novel views from existing reference views and viewpoint conditions, and the reconstruction network reconstructs novel views from a latent variable that contains the information of reference views, novel views, and viewpoints. The conditional latent variable in the inference network is regularized by the latent variable in the reconstruction network to facilitate information flow between the conditional latent variable and novel views. We also propose a statistic distance measurement dubbed the mean local maximum mean discrepancy (MLMMD) to enable the measurement of the statistic distance between two distributions with high-resolution latent variables, which can capture richer information than their low-resolution counterparts. Finally, we propose a viewpoint-dependent indirect view synthesis method to synthesize novel views more efficiently by leveraging adaptive convolution. Experimental results show that our proposed methods outperform state-of-the-art methods on different light field datasets.
AbstractList	Light field cameras can capture the radiance and direction of light rays by a single exposure, providing a new perspective to photography and 3D geometry perception. However, existing sub-aperture based light field cameras are limited by their sensor resolution to obtain high spatial and angular resolution images simultaneously. In this paper, we propose an inference-reconstruction variational autoencoder (IR-VAE) to reconstruct a dense light field image out of four corner reference views in a light field image. The proposed IR-VAE is comprised of one inference network and one reconstruction network, where the inference network infers novel views from existing reference views and viewpoint conditions, and the reconstruction network reconstructs novel views from a latent variable that contains the information of reference views, novel views, and viewpoints. The conditional latent variable in the inference network is regularized by the latent variable in the reconstruction network to facilitate information flow between the conditional latent variable and novel views. We also propose a statistic distance measurement dubbed the mean local maximum mean discrepancy (MLMMD) to enable the measurement of the statistic distance between two distributions with high-resolution latent variables, which can capture richer information than their low-resolution counterparts. Finally, we propose a viewpoint-dependent indirect view synthesis method to synthesize novel views more efficiently by leveraging adaptive convolution. Experimental results show that our proposed methods outperform state-of-the-art methods on different light field datasets. Light field cameras can capture the radiance and direction of light rays by a single exposure, providing a new perspective to photography and 3D geometry perception. However, existing sub-aperture based light field cameras are limited by their sensor resolution to obtain high spatial and angular resolution images simultaneously. In this paper, we propose an inference-reconstruction variational autoencoder (IR-VAE) to reconstruct a dense light field image out of four corner reference views in a light field image. The proposed IR-VAE is comprised of one inference network and one reconstruction network, where the inference network infers novel views from existing reference views and viewpoint conditions, and the reconstruction network reconstructs novel views from a latent variable that contains the information of reference views, novel views, and viewpoints. The conditional latent variable in the inference network is regularized by the latent variable in the reconstruction network to facilitate information flow between the conditional latent variable and novel views. We also propose a statistic distance measurement dubbed the mean local maximum mean discrepancy (MLMMD) to enable the measurement of the statistic distance between two distributions with high-resolution latent variables, which can capture richer information than their low-resolution counterparts. Finally, we propose a viewpoint-dependent indirect view synthesis method to synthesize novel views more efficiently by leveraging adaptive convolution. Experimental results show that our proposed methods outperform state-of-the-art methods on different light field datasets.Light field cameras can capture the radiance and direction of light rays by a single exposure, providing a new perspective to photography and 3D geometry perception. However, existing sub-aperture based light field cameras are limited by their sensor resolution to obtain high spatial and angular resolution images simultaneously. In this paper, we propose an inference-reconstruction variational autoencoder (IR-VAE) to reconstruct a dense light field image out of four corner reference views in a light field image. The proposed IR-VAE is comprised of one inference network and one reconstruction network, where the inference network infers novel views from existing reference views and viewpoint conditions, and the reconstruction network reconstructs novel views from a latent variable that contains the information of reference views, novel views, and viewpoints. The conditional latent variable in the inference network is regularized by the latent variable in the reconstruction network to facilitate information flow between the conditional latent variable and novel views. We also propose a statistic distance measurement dubbed the mean local maximum mean discrepancy (MLMMD) to enable the measurement of the statistic distance between two distributions with high-resolution latent variables, which can capture richer information than their low-resolution counterparts. Finally, we propose a viewpoint-dependent indirect view synthesis method to synthesize novel views more efficiently by leveraging adaptive convolution. Experimental results show that our proposed methods outperform state-of-the-art methods on different light field datasets.
Author	Xiang, Wei Han, Kang
Author_xml	– sequence: 1 givenname: Kang orcidid: 0000-0003-3626-7818 surname: Han fullname: Han, Kang email: kang.han@my.jcu.edu.au organization: College of Science and Engineering, James Cook University, Cairns, QLD, Australia – sequence: 2 givenname: Wei orcidid: 0000-0002-0608-065X surname: Xiang fullname: Xiang, Wei email: w.xiang@latrobe.edu.au organization: School of Computing, Engineering and Mathematical Sciences, La Trobe University, Melbourne, VIC, Australia
BookMark	eNp9kDtPwzAURi1URB-wI7FEYmFJ8StxPFYVhUqVQKhljRznprhK42InA_8e9yEkOjDdO5zv6rtniHqNbQChW4LHhGD5uJy_jSmmdMyIFFKkF2hAJCcxxpz2wo4TEQvCZR8Nvd9gTHhC0ivUZ4mUPGFkgFbzpgIHjYb4HbRtfOs63RrbRB_KGbXfVB1NutYGxpbgosq6aGHWn200M1CX0Xyr1hD9DV-jy0rVHm5Oc4RWs6fl9CVevD7Pp5NFrBnlbUxTwnHJKYYirTJVSMUzAJyJUC7RQmBZZoTR0LrQIHTKE80yVSWEaZBFqdgIPRzv7pz96sC3-dZ4DXWtGrCdz6kIBrhMJQ3o_Rm6sZ0Lzx2oLOUsYTxQ6ZHSznrvoMq1aQ8WWqdMnROc78XnQXy-F5-fxIcgPgvunNkq9_1f5O4YMQDwi8tQhWaM_QDdg437
CODEN	IIPRE4
CitedBy_id	crossref_primary_10_1016_j_nucengdes_2023_112712 crossref_primary_10_1109_TIP_2023_3329663 crossref_primary_10_1109_TMM_2023_3328176 crossref_primary_10_1016_j_image_2023_117031 crossref_primary_10_1016_j_nucengdes_2025_114433 crossref_primary_10_1109_TCI_2024_3507634 crossref_primary_10_1371_journal_pone_0316642 crossref_primary_10_1109_JSEN_2023_3324220 crossref_primary_10_1109_TIP_2025_3592546
Cites_doi	10.1007/978-3-030-58545-7_4 10.1007/978-3-030-01216-8_21 10.1109/CVPR.2016.90 10.1109/TPAMI.2020.2968521 10.1609/aaai.v34i07.6847 10.1609/aaai.v33i01.33015885 10.1109/CVPR.2018.00938 10.1109/ICCV.2013.89 10.1126/science.aar6170 10.1145/1141911.1141976 10.1109/ICCV.2017.37 10.1109/CVPR.2018.00262 10.1109/CVPR.2018.00931 10.1109/ICCVW.2015.17 10.1109/TPAMI.2022.3152488 10.1109/TIP.2021.3122089 10.1038/nature14539 10.1109/TIP.2003.819861 10.1109/JBHI.2021.3087407 10.1109/TPAMI.2020.2983686 10.1109/CVPR46437.2021.00466 10.1109/ICIP42928.2021.9506586 10.1007/s10489-022-03759-y 10.1109/TIP.2021.3066293 10.1109/TPAMI.2019.2945027 10.1109/ICCV.2017.478 10.1145/3306346.3322980 10.1109/CVPR.2018.00068 10.1109/TIP.2018.2791864 10.1109/CVPR.2017.178 10.1145/3197517.3201388 10.1109/CVPR.2018.00183 10.1109/TPAMI.2020.3026039 10.1007/978-3-319-46493-0_18 10.1109/ICMEW.2018.8551583 10.1109/TPAMI.2019.2941941 10.1145/2980179.2980251 10.1109/TIP.2019.2895463 10.1109/CVPR.2015.7298762 10.1109/JSTSP.2017.2747126 10.1109/CVPR42600.2020.00263 10.1109/TIP.2019.2923323 10.1109/ICCV.2015.398
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 7SP 8FD JQ2 L7M L~C L~D 7X8
DOI	10.1109/TIP.2022.3197976
DatabaseName	IEEE Xplore (IEEE) 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 MEDLINE - Academic
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 MEDLINE - Academic
DatabaseTitleList	Technology Research Database MEDLINE - Academic
Database_xml	– sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher – sequence: 2 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database
DeliveryMethod	fulltext_linktorsrc
Discipline	Applied Sciences Engineering
EISSN	1941-0042
EndPage	5644
ExternalDocumentID	10_1109_TIP_2022_3197976 9864283
Genre	orig-research
GrantInformation_xml	– fundername: Australian Government funderid: 10.13039/100015539 – fundername: Australian Research Council’s Discovery Projects Funding Scheme grantid: DP220101634 funderid: 10.13039/501100000923
GroupedDBID	--- -~X .DC 0R~ 29I 4.4 53G 5GY 5VS 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABFSI 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 E.L EBS EJD F5P HZ~ H~9 ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 MS~ O9- OCL P2P RIA RIE RNS TAE TN5 VH1 AAYXX CITATION 7SC 7SP 8FD JQ2 L7M L~C L~D 7X8
ID	FETCH-LOGICAL-c324t-26140d420eb6f8ab9a48ee0879455c7709d8132145bce7c645c38af513ce9bda3
IEDL.DBID	RIE
ISICitedReferencesCount	13
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000848262600005&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	1057-7149 1941-0042
IngestDate	Sun Sep 28 02:02:13 EDT 2025 Mon Jun 30 10:14:01 EDT 2025 Sat Nov 29 03:21:16 EST 2025 Tue Nov 18 22:35:39 EST 2025 Wed Aug 27 02:29:22 EDT 2025
IsPeerReviewed	true
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-c324t-26140d420eb6f8ab9a48ee0879455c7709d8132145bce7c645c38af513ce9bda3
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23
ORCID	0000-0002-0608-065X 0000-0003-3626-7818
PMID	35994531
PQID	2708643534
PQPubID	85429
PageCount	16
ParticipantIDs	proquest_journals_2708643534 crossref_citationtrail_10_1109_TIP_2022_3197976 ieee_primary_9864283 crossref_primary_10_1109_TIP_2022_3197976 proquest_miscellaneous_2705749692
PublicationCentury	2000
PublicationDate	20220000 2022-00-00 20220101
PublicationDateYYYYMMDD	2022-01-01
PublicationDate_xml	– year: 2022 text: 20220000
PublicationDecade	2020
PublicationPlace	New York
PublicationPlace_xml	– name: New York
PublicationTitle	IEEE transactions on image processing
PublicationTitleAbbrev	TIP
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	ref57 ref13 lecun (ref19) 2015; 521 ref12 ref59 ref58 ref14 ref53 ref52 gretton (ref39) 2006 ref54 ref10 ref17 ref16 ng (ref3) 2005 wang (ref41) 2020; 43 hinton (ref18) 2015 kingma (ref55) 2014 ref46 ref45 ref48 ref47 paszke (ref56) 2017 ref44 niklaus (ref15) 2017 ref49 goodfellow (ref35) 2014; 27 ref8 ref7 ref9 ref4 ref6 ref5 ref34 ref37 ref36 johnson (ref43) 2016 ref30 ref33 honauer (ref60) 2016 ref2 ref1 ref38 simonyan (ref42) 2014 (ref51) 2016 li (ref40) 2015 sohn (ref32) 2015 ref24 ref23 ref26 ref25 ref20 ref22 ref21 ref28 ref27 wing fung yeung (ref29) 2018 rerabek (ref50) 2016 kingma (ref11) 2013 ivanov (ref31) 2018
References_xml	– ident: ref34 doi: 10.1007/978-3-030-58545-7_4 – year: 2015 ident: ref18 article-title: Distilling the knowledge in a neural network publication-title: ArXiv 1503 02531 – start-page: 3483 year: 2015 ident: ref32 article-title: Learning structured output representation using deep conditional generative models publication-title: Proc Adv Neural Inf Process Syst (NIPS) – ident: ref28 doi: 10.1007/978-3-030-01216-8_21 – ident: ref20 doi: 10.1109/CVPR.2016.90 – ident: ref49 doi: 10.1109/TPAMI.2020.2968521 – year: 2014 ident: ref55 article-title: Adam: A method for stochastic optimization publication-title: arXiv 1412 6980 – start-page: 1 year: 2005 ident: ref3 article-title: Light field photography with a hand-held plenoptic camera – ident: ref7 doi: 10.1609/aaai.v34i07.6847 – start-page: 670 year: 2017 ident: ref15 article-title: Video frame interpolation via adaptive convolution publication-title: Proc IEEE Int Conf Comput Vis Pattern Recognit (CVPR) – ident: ref33 doi: 10.1609/aaai.v33i01.33015885 – start-page: 1 year: 2016 ident: ref50 article-title: New light field image dataset publication-title: Proc 8th Int Conf Quality Multimedia Exper – ident: ref16 doi: 10.1109/CVPR.2018.00938 – ident: ref5 doi: 10.1109/ICCV.2013.89 – ident: ref12 doi: 10.1126/science.aar6170 – ident: ref53 doi: 10.1145/1141911.1141976 – ident: ref36 doi: 10.1109/ICCV.2017.37 – volume: 27 start-page: 1 year: 2014 ident: ref35 article-title: Generative adversarial nets publication-title: Proc Adv Neural Inf Process Syst – start-page: 1 year: 2017 ident: ref56 article-title: Automatic differentiation in PyTorch publication-title: Proc Annu Conf Neural Inf Process Syst – ident: ref48 doi: 10.1109/CVPR.2018.00262 – ident: ref47 doi: 10.1109/CVPR.2018.00931 – ident: ref26 doi: 10.1109/ICCVW.2015.17 – ident: ref30 doi: 10.1109/TPAMI.2022.3152488 – ident: ref59 doi: 10.1109/TIP.2021.3122089 – volume: 521 start-page: 436 year: 2015 ident: ref19 article-title: Deep learning publication-title: Nature doi: 10.1038/nature14539 – ident: ref57 doi: 10.1109/TIP.2003.819861 – ident: ref45 doi: 10.1109/JBHI.2021.3087407 – year: 2013 ident: ref11 article-title: Auto-encoding variational Bayes publication-title: arXiv 1312 6114 – volume: 43 start-page: 3349 year: 2020 ident: ref41 article-title: Deep high-resolution representation learning for visual recognition publication-title: IEEE Trans Pattern Anal Mach Intell doi: 10.1109/TPAMI.2020.2983686 – ident: ref25 doi: 10.1109/CVPR46437.2021.00466 – ident: ref22 doi: 10.1109/ICIP42928.2021.9506586 – year: 2014 ident: ref42 article-title: Very deep convolutional networks for large-scale image recognition publication-title: arXiv 1409 1556 – year: 2016 ident: ref51 publication-title: The Stanford Light Field Archive – ident: ref58 doi: 10.1007/s10489-022-03759-y – ident: ref10 doi: 10.1109/TIP.2021.3066293 – start-page: 1718 year: 2015 ident: ref40 article-title: Generative moment matching networks publication-title: Proc Int Conf Mach Learn (ICML) – start-page: 513 year: 2006 ident: ref39 article-title: A kernel method for the two-sample-problem publication-title: Proc Adv Neural Inf Process Syst – ident: ref8 doi: 10.1109/TPAMI.2019.2945027 – ident: ref14 doi: 10.1109/ICCV.2017.478 – start-page: 694 year: 2016 ident: ref43 article-title: Perceptual losses for real-time style transfer and super-resolution publication-title: Proc 14th Eur Conf Comput Vis (ECCV) – ident: ref24 doi: 10.1145/3306346.3322980 – ident: ref44 doi: 10.1109/CVPR.2018.00068 – ident: ref52 doi: 10.1109/TIP.2018.2791864 – ident: ref27 doi: 10.1109/CVPR.2017.178 – ident: ref46 doi: 10.1145/3197517.3201388 – start-page: 19 year: 2016 ident: ref60 article-title: A dataset and evaluation methodology for depth estimation on 4D light fields publication-title: Proc Asian Conf Comput Vis – start-page: 1 year: 2018 ident: ref31 article-title: Variational autoencoder with arbitrary conditioning publication-title: Proc Int Conf Learn Represent – ident: ref37 doi: 10.1109/CVPR.2018.00183 – ident: ref23 doi: 10.1109/TPAMI.2020.3026039 – ident: ref17 doi: 10.1007/978-3-319-46493-0_18 – ident: ref38 doi: 10.1109/ICMEW.2018.8551583 – ident: ref13 doi: 10.1109/TPAMI.2019.2941941 – ident: ref6 doi: 10.1145/2980179.2980251 – ident: ref9 doi: 10.1109/TIP.2019.2895463 – ident: ref2 doi: 10.1109/CVPR.2015.7298762 – ident: ref4 doi: 10.1109/JSTSP.2017.2747126 – ident: ref21 doi: 10.1109/CVPR42600.2020.00263 – ident: ref54 doi: 10.1109/TIP.2019.2923323 – start-page: 137 year: 2018 ident: ref29 article-title: Fast light field reconstruction with deep coarse-to-fine modeling of spatial-angular clues publication-title: Proc Eur Conf Comput Vis (ECCV) – ident: ref1 doi: 10.1109/ICCV.2015.398
SSID	ssj0014516
Score	2.4477594
Snippet	Light field cameras can capture the radiance and direction of light rays by a single exposure, providing a new perspective to photography and 3D geometry...
SourceID	proquest crossref ieee
SourceType	Aggregation Database Enrichment Source Index Database Publisher
StartPage	5629
SubjectTerms	adaptive convolution Angular resolution Cameras Convolution Distance measurement Feature extraction Field cameras Image reconstruction indirect view synthesis Inference Information flow Light Light field image reconstruction Spatial resolution statistic distance measurement Superresolution Training variational autoencoder
Title	Inference-Reconstruction Variational Autoencoder for Light Field Image Reconstruction
URI	https://ieeexplore.ieee.org/document/9864283 https://www.proquest.com/docview/2708643534 https://www.proquest.com/docview/2705749692
Volume	31
WOSCitedRecordID	wos000848262600005&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: 1941-0042 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014516 issn: 1057-7149 databaseCode: RIE dateStart: 19920101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3fSxwxEB5UpNgHbf2BV61E6Euh8fZ2k03mUaRHD0R8ULm3JZudBcHeFe-uf38nudyqVATfNmyyu2QyO9-XycwAfGtVWTMNaKTRSFLVqpbYEklnbBMSjpVtjEq7uzRXV3Y8xus1-NHFwhBRPHxGZ-Ey-vKbqV-ErbJ-SCXO5nAd1o0pl7FanccgFJyNnk1tpGHYv3JJZti_GV0zEcxz5qdo2PxuwYdCIypdDF5Yo1he5b9_cjQ0w533feIn2E6AUpwvV8BnWKPJLuwkcCmS6s524eOzzIN7cDtaBfrJQECf0siKOybPaYNQnC_m05DosqFHweBWXAYmL4bh0JsY_eY_kXg5eB9uhz9vLn7JVGFBegZSc8n0SWWNyjOqy9a6Gp2yRJllJdXaG5NhYwehlJGuPRlfKu0L61o9KDxh3bjiADYm0wkdgnAKm1Y7hi_OK8wKdLbkluM2MmdxPeivZrryKf14qILxUEUakmHFYqqCmKokph5870b8WabeeKPvXpBF1y-JoQfHK2FWSTdnVW6YxjFKLFQPTrvbrFXBVeImNF3EPtooLDH_8vqTj2ArvH-5GXMMGzzR9BU2_d_5_ezxhBfo2J7EBfoP_QPgMw
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1RT9swED4xNm3wMAYdoowNI_EyaV7TxI5zj2haRbVS8VAQb5HjXKRJ0CLa7vfv7LoZaGjS3mLFjiKfz3efz_cdwGmj8ophQC2NRpKqUpXEhkhaU9SecCxvQlba9ciMx8XNDV5uwJc2F4aIwuUz-uofQyy_nrmlPyrreSpxNocv4KVWKk1W2VptzMCXnA2xTW2kYcd_HZRMsDcZXjIUTFNGqGjYAG_B60wjKp31n9ijUGDlr105mJrBzv_95Dt4G11KcbZaA7uwQdM92InupYjKO9-D7Ufcgx24Gq5T_aSHoH-IZMU1w-d4RCjOlouZp7qs6UGweytGHsuLgb_2JoZ3vBeJp4Pfw9Xg--TbuYw1FqRjV2ohGUCppObJpCpvCluhVQVRUrCaau2MSbAu-r6Yka4cGZcr7bLCNrqfOcKqttk-bE5nUzoAYRXWjbbswFinMMnQFjm3LLeRUYvtQm8906WLBOS-DsZtGYBIgiWLqfRiKqOYuvC5HXG_It_4R9-Ol0XbL4qhC0drYZZRO-dlahjIsZ-YqS6ctK9Zr3ywxE5ptgx9tFGYY3r4_JeP4c355GJUjobjHx9gy__L6mjmCDZ50ukjvHK_Fj_nD5_CMv0NhITikg
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=Inference-Reconstruction+Variational+Autoencoder+for+Light+Field+Image+Reconstruction&rft.jtitle=IEEE+transactions+on+image+processing&rft.au=Han%2C+Kang&rft.au=Xiang%2C+Wei&rft.date=2022&rft.pub=IEEE&rft.issn=1057-7149&rft.volume=31&rft.spage=5629&rft.epage=5644&rft_id=info:doi/10.1109%2FTIP.2022.3197976&rft_id=info%3Apmid%2F35994531&rft.externalDocID=9864283
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1057-7149&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1057-7149&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1057-7149&client=summon