Task-Scalable Image Semantic Communication via Conditional Affine Transforms and Pixel-Wise Quality Control

Deep autoencoder-based joint source-channel coding (JSCC) has gained significant attention for end-to-end image semantic communication systems. However, existing methods typically optimize a uniform bandwidth-distortion trade-off over the entire image, potentially leading to the loss of crucial deta...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE Wireless Communications and Networking Conference : [proceedings] : WCNC s. 1 - 6
Hlavní autoři: Wang, Jun, Yao, Shengshi, Wang, Sixian, Si, Zhongwei, Wang, Fengyu, Liu, Zhenyu, Dai, Jincheng
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 24.03.2025
Témata:
ISSN:1558-2612
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract Deep autoencoder-based joint source-channel coding (JSCC) has gained significant attention for end-to-end image semantic communication systems. However, existing methods typically optimize a uniform bandwidth-distortion trade-off over the entire image, potentially leading to the loss of crucial details and inconsistent content for tasks with diverse regions of interest. In this paper, we propose a flexible fine-grained bandwidth allocation method for deep JSCC that enables highly efficient, task-scalable image transmission across various semantic communication scenarios using a single codec. Our method optimizes the bandwidth-distortion trade-off by constraining image distortion through a 2D pixel-wise quality map. Guided by the pixel-wise quality map, we introduce a novel conditional affine transformation that generates dedicated semantic feature maps tailored to specific tasks. Additionally, we introduce a semantic guidance network to automatically generate task-aware quality maps via backpropagation without additional retraining. This approach leverages a pretrained variable-length neural JSCC codec and adjusts the transmission quality on a fine-grained level, eliminating the need to train separate models for different tasks. Experimental results demonstrate the effectiveness of our bandwidth allocation method, enhancing task-specific performance in various goal-oriented image communication scenarios without additional training.
AbstractList Deep autoencoder-based joint source-channel coding (JSCC) has gained significant attention for end-to-end image semantic communication systems. However, existing methods typically optimize a uniform bandwidth-distortion trade-off over the entire image, potentially leading to the loss of crucial details and inconsistent content for tasks with diverse regions of interest. In this paper, we propose a flexible fine-grained bandwidth allocation method for deep JSCC that enables highly efficient, task-scalable image transmission across various semantic communication scenarios using a single codec. Our method optimizes the bandwidth-distortion trade-off by constraining image distortion through a 2D pixel-wise quality map. Guided by the pixel-wise quality map, we introduce a novel conditional affine transformation that generates dedicated semantic feature maps tailored to specific tasks. Additionally, we introduce a semantic guidance network to automatically generate task-aware quality maps via backpropagation without additional retraining. This approach leverages a pretrained variable-length neural JSCC codec and adjusts the transmission quality on a fine-grained level, eliminating the need to train separate models for different tasks. Experimental results demonstrate the effectiveness of our bandwidth allocation method, enhancing task-specific performance in various goal-oriented image communication scenarios without additional training.
Author Wang, Sixian
Yao, Shengshi
Wang, Fengyu
Si, Zhongwei
Wang, Jun
Liu, Zhenyu
Dai, Jincheng
Author_xml – sequence: 1
  givenname: Jun
  surname: Wang
  fullname: Wang, Jun
  organization: Beijing University of Posts and Telecommunications,Key Laboratory of Universal Wireless Communications, Ministry of Education,Beijing,China
– sequence: 2
  givenname: Shengshi
  surname: Yao
  fullname: Yao, Shengshi
  organization: Beijing University of Posts and Telecommunications,Key Laboratory of Universal Wireless Communications, Ministry of Education,Beijing,China
– sequence: 3
  givenname: Sixian
  surname: Wang
  fullname: Wang, Sixian
  organization: Beijing University of Posts and Telecommunications,Key Laboratory of Universal Wireless Communications, Ministry of Education,Beijing,China
– sequence: 4
  givenname: Zhongwei
  surname: Si
  fullname: Si, Zhongwei
  email: sizhongwei@bupt.edu.cn
  organization: Beijing University of Posts and Telecommunications,Key Laboratory of Universal Wireless Communications, Ministry of Education,Beijing,China
– sequence: 5
  givenname: Fengyu
  surname: Wang
  fullname: Wang, Fengyu
  organization: Beijing University of Posts and Telecommunications,Key Laboratory of Universal Wireless Communications, Ministry of Education,Beijing,China
– sequence: 6
  givenname: Zhenyu
  surname: Liu
  fullname: Liu, Zhenyu
  organization: Institute for Communication Systems, University of Surrey,United Kingdom
– sequence: 7
  givenname: Jincheng
  surname: Dai
  fullname: Dai, Jincheng
  email: daijincheng@bupt.edu.cn
  organization: Beijing University of Posts and Telecommunications,Key Laboratory of Universal Wireless Communications, Ministry of Education,Beijing,China
BookMark eNo1kN1KwzAcxaMouM29gWBeoDUfTZtcjqJzMPxglV2Of9NE4tJUmk7c29uhXh3OD86Bc6boInTBIHRLSUopUXfb8qnMqchEyggT6YgKKbg8Q3NVKMkF4bnkuTpHEyqETFhO2RWaxvhBCCMiyyZoX0HcJxsNHmpv8KqFd4M3poUwOI3Lrm0PwWkYXBfwl4ORhMadHHi8sNYFg6seQrRd30YMocEv7tv4ZOuiwa8H8G44nkJD3_lrdGnBRzP_0xl6e7ivysdk_bxclYt14mghh0TpWnMpKalJXihjFTAQvKG1VbTJLAc2jrVAdS0azcwIWUMVaJoVWosM-Azd_PY6Y8zus3ct9Mfd_zn8B-SAXSE
ContentType Conference Proceeding
DBID 6IE
6IL
CBEJK
RIE
RIL
DOI 10.1109/WCNC61545.2025.10978538
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP All) 1998-Present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISBN 9798350368369
EISSN 1558-2612
EndPage 6
ExternalDocumentID 10978538
Genre orig-research
GroupedDBID 29I
6IE
6IF
6IH
6IK
6IL
6IN
AAJGR
AAWTH
ABLEC
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IEGSK
IPLJI
M43
OCL
RIE
RIL
ID FETCH-LOGICAL-i178t-9cbc38810b0679ef9a2a53d1bf91d4f3a2202fa1cb5dc2e91d2d19ac147cc54a3
IEDL.DBID RIE
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001514465200417&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Wed Nov 05 07:05:02 EST 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i178t-9cbc38810b0679ef9a2a53d1bf91d4f3a2202fa1cb5dc2e91d2d19ac147cc54a3
PageCount 6
ParticipantIDs ieee_primary_10978538
PublicationCentury 2000
PublicationDate 2025-March-24
PublicationDateYYYYMMDD 2025-03-24
PublicationDate_xml – month: 03
  year: 2025
  text: 2025-March-24
  day: 24
PublicationDecade 2020
PublicationTitle IEEE Wireless Communications and Networking Conference : [proceedings] : WCNC
PublicationTitleAbbrev WCNC
PublicationYear 2025
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0020544
Score 2.2864277
Snippet Deep autoencoder-based joint source-channel coding (JSCC) has gained significant attention for end-to-end image semantic communication systems. However,...
SourceID ieee
SourceType Publisher
StartPage 1
SubjectTerms Backpropagation
Channel allocation
Codecs
Image coding
Image communication
Image quality
Image semantic communications
joint source and channel coding
Quality control
Semantic communication
task-scalable
Training
Transforms
Title Task-Scalable Image Semantic Communication via Conditional Affine Transforms and Pixel-Wise Quality Control
URI https://ieeexplore.ieee.org/document/10978538
WOSCitedRecordID wos001514465200417&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
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT8MwDI7YxAEuvIZ4Kweu3dY0bZojmphAQtOkDbbblKSOVI11aN0m-Pc42QN24MCtsppWch527O-zCbnPhGQqkSKwTKuAG9yKUmpcy4qn6I2YpvEo37cX0emkw6HsrsnqngsDAB58BnX36HP52dQsXKis4bKlaF7SCqkIkazIWtvbFfoefA3gwvcag1anlTj_AO-ALK5vhu40UfE2pH30z78fk9oPG492t3bmhOxBcUoOfxUSPCPjvirHQQ_17ZhQ9HmCpwTtwQTVlhu6QwKhy1yhxGWqfRSQPliLH6L9jQdbUlVktJt_wnswyEugqzIbX26QQ7XXyGv7sd96CtZtFII8FOk8kEabKE3DpnZBI7BSMRVHWaitDDNuI8VQQ1aFRseZYYBCloVSmZALY2KuonNSLaYFXBCaJMpaGUMSuU7VKpIACprokehYx4ngl6Tm9Db6WFXKGG1UdvWH_JocuNlxmC7Gb0h1PlvALdk3y3lezu78_H4DWD6n-g
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3dT8IwEG8UTdQXvzB-2wdfB1vXbuujIRKIuJAwhTfSdtdkQYZhQPS_tx2g8uCDb8tl3ZLrx13vfr87hO7TkBMR8NDRRAqHKrMVOZdmLQsaGW9EuapE-b52wjiOBgPeXZHVSy4MAJTgM6jZxzKXn07U3IbK6jZbasxLtI12GKXEXdK1vu9XxvugKwiXebPeb8SNwHoI5hZIWG09eKONSmlFmof__P8Rqv7w8XD329Icoy3IT9DBr1KCp2iUiGLk9IzGLRcKt8fmnMA9GBvFZQpv0EDwIhNGYnPVZRwQP2htPoSTtQ9bYJGnuJt9wJvTzwrAy0Ibn3aQxbVX0UvzMWm0nFUjBSfzwmjmcCWVH0WeK23YCDQXRDA_9aTmXkq1L4jRkBaekixVBIyQpB4XyqOhUowK_wxV8kkO5wgHgdCaMwh826ta-BxAgGt8EskkC0J6gapWb8P3Za2M4Vpll3_I79BeK3nuDDvt-OkK7duZsggvQq9RZTadww3aVYtZVkxvy7n-AkC3q0E
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%3Abook&rft.genre=proceeding&rft.title=IEEE+Wireless+Communications+and+Networking+Conference+%3A+%5Bproceedings%5D+%3A+WCNC&rft.atitle=Task-Scalable+Image+Semantic+Communication+via+Conditional+Affine+Transforms+and+Pixel-Wise+Quality+Control&rft.au=Wang%2C+Jun&rft.au=Yao%2C+Shengshi&rft.au=Wang%2C+Sixian&rft.au=Si%2C+Zhongwei&rft.date=2025-03-24&rft.pub=IEEE&rft.eissn=1558-2612&rft.spage=1&rft.epage=6&rft_id=info:doi/10.1109%2FWCNC61545.2025.10978538&rft.externalDocID=10978538