Nguyen, Long V., Nguyen, Tuan T. ORCID: https://orcid.org/0000-0003-0055-8218, Dobre, Octavia A. and Duong, Trung Q. (2025) Digital semantic communication with neural image compression. In: IEEE INFOCOM 2025 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS). IEEE Xplore . Institute of Electrical and Electronics Engineers, Inc. (IEEE), Piscataway, New Jersey, pp. 1-2. ISBN 979-8331543709 ISSN 2159-4228 (Print), 2833-0587 (Online) (doi:10.1109/INFOCOMWKSHPS65812.2025)

Preview

PDF (Author's Accepted Manuscript) 51857 NGUYEN_Quantum_Machine_Learning_For_Drug_Discovery_(AAM)_2025.pdf - Accepted Version Download (6MB) | Preview

Abstract

Although analog semantic communication systems have attracted significant attention recently, there has been relatively less focus on digital semantic communication systems. In this work, we introduce a neural image compression-enabled semantic communication system to enhance the efficiency of digital image transmission, named NCSC. By employing an accurate and adaptable entropy model, NCSC obtains the efficiently compressed bitstreams, which are compatible with digital communication systems. Incorporating with the well-established digital components, our system trained on the MS-SSIM metric can achieve a significant bandwidth compression ratio of 0.002 at low SNR, reducing remarkably transmission overhead. Extensive simulations show that our approach outperforms traditional digital communication systems in terms of perceptual quality and bandwidth efficiency under challenging channel conditions.

Item Type:	Conference Proceedings
Title of Proceedings:	IEEE INFOCOM 2025 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS)
Uncontrolled Keywords:	adaptation models, image coding, accuracy, spectral efficiency, digital images, Bandwidth, semantic communication, entropy, image reconstruction, signal to noise ratio
Subjects:	Q Science > Q Science (General) Q Science > QA Mathematics Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Faculty / School / Research Centre / Research Group:	Faculty of Engineering & Science Faculty of Engineering & Science > School of Computing & Mathematical Sciences (CMS)
Last Modified:	01 Dec 2025 14:07
URI:	https://gala.gre.ac.uk/id/eprint/51857

Actions (login required)

View Item

Downloads