English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

With the recent advancements in edge artificial intelligence (AI), futuresixth-generation (6G) networks need to support new AI tasks such asclassification and clustering apart from data recovery. Motivated by thesuccess of deep learning, the semantic-aware and task-oriented communicationswith deep joint source and channel coding (JSCC) have emerged as new paradigmshifts in 6G from the conventional data-oriented communications with separatesource and channel coding (SSCC). However, most existing works focused on thedeep JSCC designs for one task of data recovery or AI task executionindependently, which cannot be transferred to other unintended tasks.Differently, this paper investigates the JSCC semantic communications tosupport multi-task services, by performing the image data recovery andclassification task execution simultaneously. First, we propose a newend-to-end deep JSCC framework by unifying the coding rate reductionmaximization and the mean square error (MSE) minimization in the loss function.Here, the coding rate reduction maximization facilitates the learning ofdiscriminative features for enabling to perform classification tasks directlyin the feature space, and the MSE minimization helps the learning ofinformative features for high-quality image data recovery. Next, to furtherimprove the robustness against variational wireless channels, we propose a newgated deep JSCC design, in which a gated net is incorporated for adaptivelypruning the output features to adjust their dimensions based on channelconditions. Finally, we present extensive numerical experiments to validate theperformance of our proposed deep JSCC designs as compared to various benchmarkschemes.

Semantic Communications for Image Recovery and Classification via Deep Joint Source and Channel Coding