Deep Joint Source Channel Coding for Privacy-Aware End-to-End Image Transmission

Deep neural network (DNN)-based joint source and channel coding is proposed for privacy-aware end-to-end image transmission against multiple eavesdroppers. Both scenarios of colluding and non-colluding eavesdroppers are considered. Unlike prior works that assume perfectly known and independent ident...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on machine learning in communications and networking Vol. 3; pp. 568 - 584
Main Authors: Letafati, Mehdi, Amirhossein Ameli Kalkhoran, Seyyed, Erdemir, Ecenaz, Hossein Khalaj, Babak, Behroozi, Hamid, Gunduz, Deniz
Format: Journal Article
Language:English
Published: IEEE 2025
Subjects:
adversarial neural networks
Autoencoders
Channel coding
Communication system security
deep learning
DeepJSCC
Eavesdropping
end-to-end learning
Image communication
Image reconstruction
privacy-utility trade-off
secure image transmission
Security
Training
Wireless networks
Wireless sensor networks
ISSN:2831-316X, 2831-316X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Deep neural network (DNN)-based joint source and channel coding is proposed for privacy-aware end-to-end image transmission against multiple eavesdroppers. Both scenarios of colluding and non-colluding eavesdroppers are considered. Unlike prior works that assume perfectly known and independent identically distributed (i.i.d.) source and channel statistics, the proposed scheme operates under unknown and non-i.i.d. conditions, making it more applicable to real-world scenarios. The goal is to transmit images with minimum distortion, while simultaneously preventing eavesdroppers from inferring certain private attributes of images. Simultaneously generalizing the ideas of privacy funnel and wiretap coding, a multi-objective optimization framework is expressed that characterizes the trade-off between image reconstruction quality and information leakage to eavesdroppers, taking into account the structural similarity index (SSIM) for improving the perceptual quality of image reconstruction. Extensive experiments on the CIFAR-10 and CelebA, along with ablation studies, demonstrate significant performance improvements in terms of SSIM, adversarial accuracy, and the mutual information leakage compared to benchmarks. Experiments show that the proposed scheme restrains the adversarially-trained eavesdroppers from intercepting privatized data for both cases of eavesdropping a common secret, as well as the case in which eavesdroppers are interested in different secrets. Furthermore, useful insights on the privacy-utility trade-off are also provided.
AbstractList Deep neural network (DNN)-based joint source and channel coding is proposed for privacy-aware end-to-end image transmission against multiple eavesdroppers. Both scenarios of colluding and non-colluding eavesdroppers are considered. Unlike prior works that assume perfectly known and independent identically distributed (i.i.d.) source and channel statistics, the proposed scheme operates under unknown and non-i.i.d. conditions, making it more applicable to real-world scenarios. The goal is to transmit images with minimum distortion, while simultaneously preventing eavesdroppers from inferring certain private attributes of images. Simultaneously generalizing the ideas of privacy funnel and wiretap coding, a multi-objective optimization framework is expressed that characterizes the trade-off between image reconstruction quality and information leakage to eavesdroppers, taking into account the structural similarity index (SSIM) for improving the perceptual quality of image reconstruction. Extensive experiments on the CIFAR-10 and CelebA, along with ablation studies, demonstrate significant performance improvements in terms of SSIM, adversarial accuracy, and the mutual information leakage compared to benchmarks. Experiments show that the proposed scheme restrains the adversarially-trained eavesdroppers from intercepting privatized data for both cases of eavesdropping a common secret, as well as the case in which eavesdroppers are interested in different secrets. Furthermore, useful insights on the privacy-utility trade-off are also provided.
Author Amirhossein Ameli Kalkhoran, Seyyed
Behroozi, Hamid
Gunduz, Deniz
Erdemir, Ecenaz
Hossein Khalaj, Babak
Letafati, Mehdi
Author_xml – sequence: 1
  givenname: Mehdi
  orcidid: 0000-0003-3731-3943
  surname: Letafati
  fullname: Letafati, Mehdi
  email: mletafati@ee.sharif.edu
  organization: Department of Electrical Engineering, Sharif University of Technology, Tehran, Iran
– sequence: 2
  givenname: Seyyed
  orcidid: 0009-0004-9612-9585
  surname: Amirhossein Ameli Kalkhoran
  fullname: Amirhossein Ameli Kalkhoran, Seyyed
  organization: Computer Engineering Department, Sharif University of Technology, Tehran, Iran
– sequence: 3
  givenname: Ecenaz
  orcidid: 0000-0002-2170-4985
  surname: Erdemir
  fullname: Erdemir, Ecenaz
  organization: Department of Electrical and Electronic Engineering, Imperial College London, London, U.K
– sequence: 4
  givenname: Babak
  orcidid: 0000-0002-9289-2338
  surname: Hossein Khalaj
  fullname: Hossein Khalaj, Babak
  organization: Department of Electrical Engineering, Sharif University of Technology, Tehran, Iran
– sequence: 5
  givenname: Hamid
  orcidid: 0000-0001-9294-3134
  surname: Behroozi
  fullname: Behroozi, Hamid
  organization: Department of Electrical Engineering, Sharif University of Technology, Tehran, Iran
– sequence: 6
  givenname: Deniz
  orcidid: 0000-0002-7725-395X
  surname: Gunduz
  fullname: Gunduz, Deniz
  organization: Department of Electrical and Electronic Engineering, Imperial College London, London, U.K
BookMark eNpNkMtKAzEARYMoWLU_IC7yA1PzmmSyLOOrUrVgBXchz5rSJiVTlf69oxVxdS8X7lmcE3CYcvIAnGM0whjJy_nDtH0cEUTqEa05k0gcgAFpKK4o5q-H__oxGHbdEiFEKRKCkwGYXXm_gfc5pi18zu_Feti-6ZT8CrbZxbSAIRc4K_FD2101_tTFw-vkqm2u-oCTtV54OC86devYdTGnM3AU9Krzw988BS831_P2rpo-3U7a8bSypJbbyhjqHdHU1t7VRjchMEYd5dpa3PQLE6YWpkGSB95IShpntOCcYsdFUxtCT8Fkz3VZL9WmxLUuO5V1VD9DLgulyzbalVeGUcwkYbgxmjHcR0DeoWACsbWgsmeRPcuW3HXFhz8eRupbsfpRrL4Vq1_F_elif4re-38HKSRjhH4BOR15Ag
CODEN ITMLBB
Cites_doi 10.1109/TCOMM.2021.3128423
10.1109/LWC.2023.3275383
10.1109/MCOM.001.1900511
10.1109/MSP.2014.2330661
10.1109/ICIP.2017.8297089
10.1109/TIFS.2019.2945619
10.1109/ITW.2014.6970882
10.1109/ICASSP40776.2020.9053216
10.1145/3395352.3402654
10.1109/MCOM.006.2200878
10.1109/ICCWorkshops50388.2021.9473492
10.1109/TIT.2005.860427
10.1109/ICCV.2015.425
10.1109/GLOBECOM46510.2021.9685537
10.1109/OJVT.2023.3315216
10.1109/ICC45041.2023.10278612
10.1109/TMLCN.2024.3522872
10.1109/JIOT.2020.3026475
10.1109/GLOBECOM54140.2023.10436928
10.1109/JSAC.2020.3036955
10.1109/ALLERTON.2019.8919758
10.1109/TCCN.2023.3306851
10.1109/WCNC57260.2024.10570533
10.1109/TCCN.2019.2919300
10.1109/TCI.2016.2644865
10.1109/JSAIT.2020.2987203
10.1109/jproc.2024.3477331
10.1109/LWC.2022.3147952
10.1109/ICASSP43922.2022.9747068
10.1109/ICCV.2015.123
10.1109/MCOMSTD.0001.2000082
10.1109/TIFS.2023.3262112
10.1109/TWC.2020.3006012
10.1109/MCOM.004.2200819
10.1109/GLOBECOM52923.2024.10901326
10.1109/LWC.2024.3457008
10.1109/ICASSP.2019.8683409
10.1109/VTC2022-Spring54318.2022.9860947
ContentType Journal Article
DBID 97E
ESBDL
RIA
RIE
AAYXX
CITATION
DOA
DOI 10.1109/TMLCN.2025.3564907
DatabaseName IEEE Xplore (IEEE)
IEEE Xplore Open Access Journals (LUT & LAB)
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE Electronic Library (IEL)
CrossRef
DOAJ Directory of Open Access Journals
DatabaseTitle CrossRef
DatabaseTitleList

Database_xml – sequence: 1
  dbid: DOA
  name: DOAJ Directory of Open Access Journals
  url: https://www.doaj.org/
  sourceTypes: Open Website
– sequence: 2
  dbid: RIE
  name: IEEE Xplore
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISSN 2831-316X
EndPage 584
ExternalDocumentID oai_doaj_org_article_b431492418ba44118bf0ed0fbf2c5739
10_1109_TMLCN_2025_3564907
10979442
Genre orig-research
GrantInformation_xml – fundername: Institute for Research in Fundamental Sciences (IPM)
  funderid: 10.13039/501100006115
GroupedDBID 0R~
97E
ABVLG
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
EBS
ESBDL
GROUPED_DOAJ
JAVBF
M~E
OCL
RIA
RIE
AAYXX
CITATION
ID FETCH-LOGICAL-c259t-bb3ed2a3c5ed5ba8ff443d36acc18d5b47b57b8096f689328dba76631d6785b23
IEDL.DBID DOA
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001484660900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2831-316X
IngestDate Fri Oct 03 12:46:32 EDT 2025
Sat Nov 29 07:56:38 EST 2025
Thu May 29 05:57:20 EDT 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Language English
License https://creativecommons.org/licenses/by/4.0/legalcode
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c259t-bb3ed2a3c5ed5ba8ff443d36acc18d5b47b57b8096f689328dba76631d6785b23
ORCID 0000-0003-3731-3943
0009-0004-9612-9585
0000-0001-9294-3134
0000-0002-7725-395X
0000-0002-2170-4985
0000-0002-9289-2338
OpenAccessLink https://doaj.org/article/b431492418ba44118bf0ed0fbf2c5739
PageCount 17
ParticipantIDs crossref_primary_10_1109_TMLCN_2025_3564907
doaj_primary_oai_doaj_org_article_b431492418ba44118bf0ed0fbf2c5739
ieee_primary_10979442
PublicationCentury 2000
PublicationDate 20250000
2025-00-00
2025-01-01
PublicationDateYYYYMMDD 2025-01-01
PublicationDate_xml – year: 2025
  text: 20250000
PublicationDecade 2020
PublicationTitle IEEE transactions on machine learning in communications and networking
PublicationTitleAbbrev TMLCN
PublicationYear 2025
Publisher IEEE
Publisher_xml – name: IEEE
References ref13
ref35
ref12
ref15
ref14
ref36
ref31
ref30
ref11
ref33
ref10
ref32
ref2
ref1
ref17
ref39
ref16
ref19
ref18
Ballé (ref27) 2025
ref24
ref23
ref26
ref25
ref20
Ba (ref34) 2016
Kingma (ref40)
ref42
ref41
ref22
ref21
ref43
ref28
ref29
ref8
ref7
ref9
ref4
ref3
ref6
ref5
Krizhevsky (ref38) 2009
Abadi (ref37) 2025
References_xml – ident: ref11
  doi: 10.1109/TCOMM.2021.3128423
– ident: ref24
  doi: 10.1109/LWC.2023.3275383
– ident: ref2
  doi: 10.1109/MCOM.001.1900511
– ident: ref30
  doi: 10.1109/MSP.2014.2330661
– ident: ref35
  doi: 10.1109/ICIP.2017.8297089
– ident: ref20
  doi: 10.1109/TIFS.2019.2945619
– ident: ref29
  doi: 10.1109/ITW.2014.6970882
– ident: ref16
  doi: 10.1109/ICASSP40776.2020.9053216
– start-page: 1
  volume-title: Proc. Int. Conf. Learn. Represent.
  ident: ref40
  article-title: Adam: A method for stochastic optimization
– ident: ref21
  doi: 10.1145/3395352.3402654
– ident: ref26
  doi: 10.1109/MCOM.006.2200878
– ident: ref12
  doi: 10.1109/ICCWorkshops50388.2021.9473492
– ident: ref28
  doi: 10.1109/TIT.2005.860427
– ident: ref39
  doi: 10.1109/ICCV.2015.425
– ident: ref4
  doi: 10.1109/GLOBECOM46510.2021.9685537
– year: 2016
  ident: ref34
  article-title: Layer normalization
  publication-title: arXiv:1607.06450
– ident: ref5
  doi: 10.1109/OJVT.2023.3315216
– ident: ref22
  doi: 10.1109/ICC45041.2023.10278612
– ident: ref41
  doi: 10.1109/TMLCN.2024.3522872
– year: 2009
  ident: ref38
  article-title: Learning multiple layers of features from tiny images
– ident: ref9
  doi: 10.1109/JIOT.2020.3026475
– ident: ref1
  doi: 10.1109/GLOBECOM54140.2023.10436928
– ident: ref3
  doi: 10.1109/JSAC.2020.3036955
– ident: ref32
  doi: 10.1109/ALLERTON.2019.8919758
– ident: ref23
  doi: 10.1109/TCCN.2023.3306851
– ident: ref43
  doi: 10.1109/WCNC57260.2024.10570533
– ident: ref14
  doi: 10.1109/TCCN.2019.2919300
– ident: ref36
  doi: 10.1109/TCI.2016.2644865
– ident: ref15
  doi: 10.1109/JSAIT.2020.2987203
– ident: ref13
  doi: 10.1109/jproc.2024.3477331
– ident: ref7
  doi: 10.1109/LWC.2022.3147952
– year: 2025
  ident: ref27
  article-title: Density modeling of images using a generalized normalization transformation
  publication-title: arXiv:1511.06281
– volume-title: TensorFlow: Large-Scale Machine Learning on Heterogeneous Systems
  year: 2025
  ident: ref37
– ident: ref17
  doi: 10.1109/ICASSP43922.2022.9747068
– ident: ref33
  doi: 10.1109/ICCV.2015.123
– ident: ref10
  doi: 10.1109/MCOMSTD.0001.2000082
– ident: ref31
  doi: 10.1109/TIFS.2023.3262112
– ident: ref6
  doi: 10.1109/TWC.2020.3006012
– ident: ref18
  doi: 10.1109/MCOM.004.2200819
– ident: ref25
  doi: 10.1109/GLOBECOM52923.2024.10901326
– ident: ref42
  doi: 10.1109/LWC.2024.3457008
– ident: ref19
  doi: 10.1109/ICASSP.2019.8683409
– ident: ref8
  doi: 10.1109/VTC2022-Spring54318.2022.9860947
SSID ssj0003307762
Score 2.278735
Snippet Deep neural network (DNN)-based joint source and channel coding is proposed for privacy-aware end-to-end image transmission against multiple eavesdroppers....
SourceID doaj
crossref
ieee
SourceType Open Website
Index Database
Publisher
StartPage 568
SubjectTerms adversarial neural networks
Autoencoders
Channel coding
Communication system security
deep learning
DeepJSCC
Eavesdropping
end-to-end learning
Image communication
Image reconstruction
privacy-utility trade-off
secure image transmission
Security
Training
Wireless networks
Wireless sensor networks
SummonAdditionalLinks – databaseName: IEEE Electronic Library (IEL)
  dbid: RIE
  link: http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NT9wwELUK6gEOhbYgtkDlQ2-VIRvbsX1cFlBb0dVKpRW3yB9jaaU2QUug6r9n7GTR9sCBk6ORpVhvHM-bmcyYkE-Om8BNBUxpYZiIMTItpWRIBqqx08gRckb315WazfTNjZkPxeq5FgYA8s9ncJIecy4_tP4-hcpOU7bUCIEn7oZSVV-s9RRQQcccheWqMKYwp9ffr6YzdAFLecJlJUy6MnbN-OQe_f9dqpJtyuXOC1ezS94M5JFOem2_Ja-geUe211oKvifzc4Bb-q1dNB39kQPzNBUQNPCbTttkpyiyVDpfLh6s_8cmf-0S6EUTWNcyHOjXP3i-0GzAcAOkSNoe-Xl5cT39woZbE5hHV6ZjznEIpeVeQpDO6hiF4IFX1vuxRolQTiqn0XWJFZKVUgdnFfKOcUC7JV3J98lm0zZwQChEJyJYqwrvBM53pdYVlNZVSnGu_Ih8XsFZ3_bNMersVBSmzuDXCfx6AH9EzhLiTzNTY-ssQFTr4TupHRIagT7hWDuLTA2HWEAoooull4qbEdlLmlh7Xa-ED8_ID8lWWkMfNDkim93yHo7Ja__QLe6WH_MOegS7usP2
  priority: 102
  providerName: IEEE
Title Deep Joint Source Channel Coding for Privacy-Aware End-to-End Image Transmission
URI https://ieeexplore.ieee.org/document/10979442
https://doaj.org/article/b431492418ba44118bf0ed0fbf2c5739
Volume 3
WOSCitedRecordID wos001484660900001&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: PRVAON
  databaseName: DOAJ Directory of Open Access Journals
  customDbUrl:
  eissn: 2831-316X
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0003307762
  issn: 2831-316X
  databaseCode: DOA
  dateStart: 20230101
  isFulltext: true
  titleUrlDefault: https://www.doaj.org/
  providerName: Directory of Open Access Journals
– providerCode: PRVHPJ
  databaseName: ROAD: Directory of Open Access Scholarly Resources
  customDbUrl:
  eissn: 2831-316X
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0003307762
  issn: 2831-316X
  databaseCode: M~E
  dateStart: 20230101
  isFulltext: true
  titleUrlDefault: https://road.issn.org
  providerName: ISSN International Centre
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV3PS8MwFA4iHvQg_pg4f5GDN4l2TdIkR50TlW0MnOKtJGkCA-3GrIoX_3Zf0in15MVLAyG04Xttvvclfe8hdGyoKqjKHBGSKcK890Ryzgk4A1nHSPAR4onuQ18Mh_LxUY0apb7CP2F1euAauDMDDMdAJHSk0UDd0PjEFYk3PrVc0Bi6lwjVEFNhDQaVLuAz_46SSdTZeNDvDkEPpvyU8oypUD-2wUQxYf-vCiuRYK420PrCM8Tn9Yw20ZIrt9BaI1_gNhpdOjfDt9NJWeG7uOuOQ3RA6Z5wdxpICIMLikfzyZu2H-T8Xc8d7pUFqaYEGnzzDIsHjuwE1g3bZC10f9Ubd6_JoiQCsaBTKmIMdUWqqeWu4EZL7xmjBc20tR0JPUwYLowEXeIz8ERSWRgtwKnoFEBK3KR0By2X09LtIuy8Yd5pLRJrGIw3qZSZS7XJhKBU2DY6-YYnn9WZL_KoGBKVRzDzAGa-ALONLgKCPyND1urYAbbMF7bM_7JlG7UC_o3HKVguWLr3HzffR6thwvX2yQFaruav7hCt2Ldq8jI_iq8PXAefvaMYBPgFyirIFw
linkProvider Directory of Open Access Journals
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LTxwxDI4qqEQ5tJSC2D4gh95QYHaSTJIj3YKAblcrQStuozwcaaV2Bm0Hqv57nMyAtoceevLIijSW8_BnO3YI-ei4CdxUwJQWhokYI9NSSoZgoBo7jRghZ3S_T9Vspm9uzHwoVs-1MACQL5_BUfrMufzQ-rsUKjtO2VIjBJ646xJJ0ZdrPYVU0DVXuLcfS2MKc3z9dTqZoRNYyiMuK2HSo7Er5id36f_rWZVsVc5e_ac8W-TlAB_pST_fr8kzaLbJ5kpTwTdk_hngll62i6ajVzk0T1MJQQM_6KRNlooiTqXz5eLe-j_s5LddAj1tAutahoRe_MQThmYThksgxdJ2yLez0-vJORveTWAenZmOOcchlJZ7CUE6q2MUggdeWe_HGjlCOamcRuclVghXSh2cVYg8xgEtl3Ql3yVrTdvAHqEQnYhgrSq8EzjelVpXUFpXKcW58iNy-KjO-rZvj1Fnt6IwdVZ-nZRfD8ofkU9J408jU2vrzECt1sNOqR1CGoFe4Vg7i1gNSSwgFNHF0kvFzYjspJlY-V0_CW__wT8gG-coSj29mH15R14kefoQynuy1i3v4AN57u-7xa_lfl5NDxb_xz0
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=Deep+Joint+Source+Channel+Coding+for+Privacy-Aware+End-to-End+Image+Transmission&rft.jtitle=IEEE+transactions+on+machine+learning+in+communications+and+networking&rft.au=Letafati%2C+Mehdi&rft.au=Amirhossein+Ameli+Kalkhoran%2C+Seyyed&rft.au=Erdemir%2C+Ecenaz&rft.au=Hossein+Khalaj%2C+Babak&rft.date=2025&rft.issn=2831-316X&rft.eissn=2831-316X&rft.volume=3&rft.spage=568&rft.epage=584&rft_id=info:doi/10.1109%2FTMLCN.2025.3564907&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TMLCN_2025_3564907
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2831-316X&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2831-316X&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2831-316X&client=summon