Pith. sign in

REVIEW 4 major objections 5 minor 54 references

SCSC: A Novel Standards-Compatible Semantic Communication Framework for Image Transmission

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read SCSC wraps the standard BPG/LDPC/QPSK/SVD digital chain in two learned modules, trains them through a proxy network, and reports over 29% channel-bandwidth savings at equal PSNR while improving semantic segmentation.

desk verdict Solid standards-compatible semantic image transmission integration; the load-bearing question is the unmeasured fidelity of its proxy network. read the letter →

arxiv 2501.02738 v1 pith:ELQ7E2FS submitted 2025-01-06 cs.IT math.IT

Xue Han , Yongpeng Wu , Zhen Gao , Biqian Feng , Yuxuan Shi , Deniz Gündüz , Wenjun Zhang This is my paper · ORCID
classification cs.ITmath.IT
keywords semanticcommunicationjointsource-channelcodingimagetransmissionMIMOstandard-compatibledeeplearningsegmentationproxynetwork
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

SCSC is a semantic communication framework that keeps the conventional digital transmission chain—BPG compression, LDPC channel coding, QPSK modulation, SVD precoding—unchanged and adds two trainable modules around it: PPEN, which preprocesses the image before compression so that task-relevant semantics survive, and PCEN, which refines precoding and combining for finite-alphabet MIMO transmission. Because the digital chain is non-differentiable, the authors train the modules end-to-end through a proxy neural network that mimics the chain's input-output behavior during backpropagation. The paper reports that this standard-compatible system outperforms both separation-based digital baselines and an end-to-end learned JSCC baseline, saving more than 29% of channel bandwidth at the same PSNR and substantially improving semantic segmentation mIoU. The practical significance is that semantic communication could be deployed on existing systems by adding two modules rather than replacing codecs and standards.

What carries the argument

Three components carry the argument. PPEN (preprocessing-empowered network) is a distortion-aware image transform built from deformable convolution, strip pooling, and quantization-adaptive layers; it selects and reshapes information so the standard codec preserves what the downstream task needs. PCEN (precoder & combiner-enhanced network) is a deep-unfolded implementation of ADMM iterations for finite-alphabet precoding, with trainable step sizes and a learned linear combiner, which refines the transmitted and received signals under the constellation constraint. The proxy network is a JSCC autoencoder trained to imitate the full digital chain; it supplies gradients through otherwise discrete blocks while the true chain is used for forward propagation.

What would settle it

Run the trained SCSC modules with the true BPG+LDPC+QPSK+SVD chain (no proxy) at the training SNRs and CBRs; if the reported bandwidth savings and mIoU gains vanish, or if the proxy's reconstructions at those operating points differ markedly from the real codec's, the proxy-training claim is falsified.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that the advantage of learned JSCC over conventional digital transmission does not require abandoning the digital chain. It shows that a preprocessing network and a precoder/combiner-enhancement network, trained jointly through a differentiable proxy of the non-differentiable codec chain, can make the standard chain preserve task-relevant semantics and transmit them efficiently over MIMO channels. The central numerical claim is a 29.46% reduction in channel bandwidth at equal PSNR relative to the BPG+LDPC+QPSK baseline, a 1.28 dB PSNR gain at a fixed bandwidth ratio, and a semantic-segmentation mIoU that exceeds both the digital baseline and the learned JSCC baselines on panoramic images.

Load-bearing premise

The load-bearing premise is that the proxy network is a faithful stand-in for the real digital chain, so that gradients computed through the proxy improve the true system; the paper does not measure the proxy-to-real discrepancy or check gradient alignment.

Editorial extensions

If this is right

  • Legacy deployment requires adding only two modules: PPEN before the source encoder and PCEN around the MIMO precoder/combiner; the underlying codecs remain untouched.
  • At equal PSNR the system saves 29.46% channel bandwidth, and at a fixed bandwidth ratio of 1/12 it gains 1.28 dB in PSNR over the standard digital chain.
  • On the panoramic segmentation dataset the system raises mIoU from 30.62% (digital baseline) to 56.37%, and matches or exceeds learned JSCC baselines in most class-wise categories.
  • The gains transfer to JPEG and JPEG2000 source codecs, to the CLIC2021 dataset, to object detection and classification tasks, and to practical 5G channel models with imperfect CSI.
  • The deep-unfolded PCEN keeps the added complexity small enough that SCSC's throughput is about 19 times the digital baseline's in the reported setup.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the proxy-fidelity assumption holds, the same wrapper strategy could be applied to other non-differentiable transmission chains, such as polar-coded or OFDM systems, without redesigning the semantic modules.
  • PPEN can be read as a learned rate-distortion shaping operator that spends bits on task-relevant structures; connecting it to rate-distortion theory could produce predictable trade-offs between reconstruction fidelity and task accuracy.
  • Because the transmitted signal remains a finite-alphabet digital constellation, SCSC may keep the security and multi-hop robustness advantages of digital transmission over analog JSCC, a benefit the paper motivates but does not quantify.
  • A direct next experiment would be to measure the proxy-codec reconstruction gap and compare proxy gradients with finite-difference gradients of the true chain at every training SNR and CBR, giving a deployment rule for when the proxy must be retrained.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes SCSC, a standards-compatible semantic communication framework for image transmission over MIMO channels. The system wraps conventional digital codecs (BPG/JPEG/JPEG2000 source coding, LDPC channel coding, QPSK modulation, SVD precoding) with two learnable modules: PPEN, a preprocessing network that filters the source image before compression, and PCEN, a deep-unfolded precoder/combiner enhancement network. Because the standard codecs are non-differentiable, training is performed through a proxy JSCC autoencoder that mimics the whole digital chain. The authors report PSNR, MS-SSIM, and semantic-segmentation mIoU results on Cityscapes, CVRG-Pano, and CLIC2021, claiming over 29% channel-bandwidth savings at equal PSNR (Table II) and gains in mIoU over several baselines, alongside robustness experiments across codecs, modulation orders, datasets, tasks, and imperfect CSI.

Significance. If the results hold, the framework is practically significant: it offers a path to semantic-communication gains while retaining deployment compatibility with existing digital standards, which is a genuine concern for end-to-end JSCC schemes. The PCEN design is a well-motivated ADMM-unfolded finite-alphabet precoder with a clean MSE formulation (Eq. (10)-(17)), and the quantization-adaptive PPEN architecture is a sensible way to handle multiple compression levels. The experimental breadth—multiple datasets, tasks, codecs, and channel models—is a strength. However, the central claim depends on an unvalidated proxy-network assumption, and the numerical evidence currently rests on single-run comparisons without error bars or code, which weakens confidence in the reported margins.

major comments (4)
  1. [Section III.C and Algorithm 2] The entire end-to-end training pipeline replaces the non-differentiable digital chain with a JSCC autoencoder proxy and backpropagates through it, yet the paper never quantifies how faithfully the proxy represents the real BPG+LDPC+QPSK+SVD system. Section III.C acknowledges that 'if the discrepancy is too significant, the proxy network cannot substitute for the conventional codecs,' but no measurement of this discrepancy is provided for any (Q, LDPC rate, SNR) combination, and no gradient-alignment check is reported. Since PPEN and PCEN are optimized exclusively through proxy gradients (Algorithm 2), a biased proxy could yield modules that perform well only in simulation with the proxy and poorly when deployed on the actual standards-based chain. The authors should add a quantitative proxy-fidelity evaluation (e.g., output MSE/PSNR between proxy and real chain across the operating range) and ideally a gradient-alignment test or a comparison of SCSC fine-tuned on the real system vs. proxy-trained only.
  2. [Section IV.A.4 and Table II] All performance claims, including the headline 29.46% BD-CBR savings, are based on single-run experiments with no reported error bars, random seeds, or statistical significance. The margins over some baselines are modest (e.g., SCSC 56.37 vs. DeepJSCC-MIMO 54.24 mIoU in Table I; BD-PSNR 1.28 vs. 1.08 dB in Table II), so it is unclear whether the differences are meaningful. Furthermore, the training details for the baseline models (ADJSCC, DeepJSCC-MIMO, DSSCC) are incomplete—epochs, learning-rate schedules, SNR sampling, and segmentation-backbone training are not specified—making the comparisons difficult to reproduce. Please provide multi-seed results with variance or significance tests, and release code or detailed training configurations.
  3. [Section IV.B.3 and Table I] The text states that SCSC 'slightly underperforms compared to the DeepJSCC-MIMO method' in the mIoU comparison of Fig. 8, yet Table I reports SCSC mIoU 56.37 vs. DeepJSCC-MIMO 54.24 at R=1/6. These statements conflict. Please clarify the operating point used for Table I and reconcile it with the curves in Fig. 8; if Table I is at a different CBR or SNR, that must be stated explicitly.
  4. [Section II.D and Algorithm 2] The proxy network is pre-trained to mimic the standard codec before end-to-end training, but PPEN changes the input distribution of the proxy during SCSC training. The paper does not state whether the proxy is re-finetuned or updated within the K end-to-end epochs. If the proxy remains frozen while PPEN evolves, its fidelity may degrade, potentially making the training gradients increasingly stale. Please specify the proxy update schedule and report the proxy loss (Eq. (18)) during end-to-end training as a sanity check.
minor comments (5)
  1. [Figure 10 caption] There is a typo: 'Gound Truth' should be 'Ground Truth'.
  2. [Eq. (8) and surrounding text] The notation is inconsistent: the equation defines ̄r_{i,c} and ^r_{j,c}, while the text refers to '¯ri,c and ^rj,c'. Please unify the notation.
  3. [Section IV.B.3] The sentence beginning 'While our approach slightly underperforms compared to the DeepJSCC-MIMO method...' is a grammatical fragment; it should be merged with the previous sentence or rephrased.
  4. [References [15] and [48]] References [15] and [48] appear to be the same work (arXiv preprint and IEEE IoT Journal version). Please cite only one, or clearly indicate that they are different versions of the same paper.
  5. [Fig. 6 and Table II] The annotations like '(41,1/6)' in the figures are not explained in the captions. Please add a note that these denote (BPG quantization parameter Q, LDPC code rate) combinations. Additionally, in Table II, the sign convention of BD-CBR (negative meaning bandwidth savings) should be stated explicitly.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the claimed bandwidth savings and task gains are measured against external digital and JSCC baselines, and the proxy network is a surrogate fitted to the real codec rather than the source of the claimed result.

full rationale

The central claims—29.46% BD-CBR savings and improved mIoU/PSNR—are established by comparing SCSC against external baselines: the standard digital BPG+LDPC scheme (Table II, Figs. 6-8), ADJSCC [47], and DeepJSCC-MIMO [9]. These comparisons use the real, non-differentiable codec in the forward path (Algorithm 2), so the headline gains are measured, not constructed from the training surrogate. The proxy network is a differentiable stand-in for the standard codec, fine-tuned to match real codec outputs via the MSE loss Lpro = d(\hat{x}, \hat{x}^a) in Eq. (18), and it is used only for gradient backpropagation; the paper does not claim the proxy itself predicts the final performance. No equation defines the reported gain in terms of the proxy or of any fitted parameter. The self-citations (e.g., the JSCC AE architecture from [5], which includes a co-author of this paper) are not load-bearing: [5] supplies only an architectural choice, and the proxy is subsequently re-fitted to the actual digital scheme. The paper itself notes a genuine unverified condition in Sec. III.C: "If the discrepancy is too significant, the proxy network cannot substitute for the conventional codecs, indicating a failure in training." This is a correctness/robustness limitation—proxy-gradient alignment is not measured—but it is not circularity, because the final evaluation and the reported bandwidth savings come from the true codec path, not from the proxy. Overall, the derivation chain is self-contained against external benchmarks, and no prediction reduces by construction to an input or to a self-citation chain.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The central design rests on the standard digital codecs being fixed and non-differentiable, perfect CSI (except one imperfect-CSI experiment), and the validity of proxy-gradient training. No fundamentally new physical entity is introduced; the learnable modules are computational artifacts. The main hand-chosen hyperparameters (loss weights, SNR range, quantization set, momentum) influence the reported gains.

free parameters (7)
  • Loss weighting coefficients λ1, λ2, λ3 = λ1=λ3=0.1, λ2=0.5
    Weights of PCEN loss, segmentation loss, and reconstruction MSE in Eq. (19); chosen by hand without sensitivity analysis.
  • BPG quantization parameter set Q = {28,31,34,37,41}
    Discrete compression qualities available to SCSC; the best combination per test scenario is selected, effectively fitting the codec operating point to the test condition.
  • Initial PCEN momentum α = 0.95
    Momentum parameter in Algorithm 1; fixed by hand.
  • Training SNR sampling range = [2,12] dB
    Uniform sampling range for channel SNR during training; chosen to cover the test range.
  • Initial learning rate β0 and schedule power p = 5e-4, p=0.9
    AdamW learning rate schedule; chosen by hand.
  • Channel estimation error variance (imperfect CSI) = CN(0,0.1)
    Used in the imperfect-CSI experiment only; chosen to model estimation error.
  • PCEN unfolding iterations T = not specified
    Number of deep-unfolding layers in PEN is not stated, which affects both performance and complexity.
assumptions (6)
  • domain assumption MIMO channel is block-fading with perfect CSI at both Tx and Rx (except in the imperfect-CSI experiment).
    Section II.B; all training and testing assume known H, which is optimistic and simplifies precoder design.
  • domain assumption BPG, JPEG, JPEG2000, LDPC and the modulator are treated as fixed, non-differentiable standard modules.
    Section II.A and II.D; the whole premise of standards compatibility.
  • domain assumption The downstream task model ERF-PSPNet is pretrained and its weights are frozen during SCSC training.
    Section IV.A.2; the segmentation performance depends on this fixed network.
  • ad hoc to paper A JSCC autoencoder proxy can imitate the input-output behavior of the standard digital chain well enough for gradient transfer.
    Section II.D and Algorithm 2; central enabling assumption, not quantitatively validated.
  • standard math Finite-alphabet precoding consensus ADMM converges and its unfolded iterations with trainable parameters remain effective.
    Section III.B; relies on prior work [40] for the algorithm and on the hypothesis that learned parameters improve it.
  • domain assumption The preprocessed image x is a valid input to BPG and the chosen Q values correspond to the reported compression ratios.
    Section II.A and IV.A.4; the mapping between Q and bitrate is not reported, but the CBR calculation depends on it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SCSC: A Novel Standards-Compatible Semantic Communication Framework for Image Transmission." pith.science (2026). https://pith.science/paper/ELQ7E2FS

@misc{pith2026250102738,
  author       = {Pith},
  title        = {Pith review of: SCSC: A Novel Standards-Compatible Semantic Communication Framework for Image Transmission},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ELQ7E2FS}},
  note         = {Machine review of arXiv:2501.02738}
}
read the original abstract

Joint source-channel coding (JSCC) is a promising paradigm for next-generation communication systems, particularly in challenging transmission environments. In this paper, we propose a novel standard-compatible JSCC framework for the transmission of images over multiple-input multiple-output (MIMO) channels. Different from the existing end-to-end AI-based DeepJSCC schemes, our framework consists of learnable modules that enable communication using conventional separate source and channel codes (SSCC), which makes it amenable for easy deployment on legacy systems. Specifically, the learnable modules involve a preprocessing-empowered network (PPEN) for preserving essential semantic information, and a precoder \& combiner-enhanced network (PCEN) for efficient transmission over a resource-constrained MIMO channel. We treat existing compression and channel coding modules as non-trainable blocks. Since the parameters of these modules are non-differentiable, we employ a proxy network that mimics their operations when training the learnable modules. Numerical results demonstrate that our scheme can save more than 29\% of the channel bandwidth, and requires lower complexity compared to the constrained baselines. We also show its generalization capability to unseen datasets and tasks through extensive experiments.

Figures

Figures reproduced from arXiv: 2501.02738 by the authors.

Figure 1
Figure 1. (a) Conventional digital communication system. (b) Overview of the proposed SCSC framework for semantic communications. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The detailed framework of our PPEN module. The DAC module is used in the last stage of the distortion compensation layer. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Illustration of the DAC semantic feature extraction block and the [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (12 more)
Figure 4
Figure 4. Figure 4: The diagram of the PEN and CEN models with parameters [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: The structure of the proxy network. we can obtain the following iterative process: W = [diag((U2H) HU2H)]−1 (U2H) H, r t = z t d + γ tW(xp − U2Hzt d ), z t+1 = ΠM(r t ), z t+1 d = α t z t d + (1 − α t )z t+1 , (17) where θ = {γ t , αt} are the learnable parameters in P…
Figure 6
Figure 6. Figure 6: Comparison of the PSNR metric for our proposed SCSC framework and the other baselines over a [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Comparison of MS-SSIM performance in a 2 × 2 MIMO fading channel; (a) shows the MS-SSIM performance versus channel SNR, where the average CBR is set to R = 1/12. (b) shows the MS-SSIM performance versus CBR over the fading channel at SNR = 6 dB. 2 4 6 8 10 12 SNR (dB) …
Figure 8
Figure 8. Figure 8: Performance of semantic segmentation on CVRG-Pano dataset over a [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: Visual comparison of reconstructed images. The first, second, and third rows correspond to SNR = 2, 6, and 10 dB, respectively. The first column [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 10
Figure 10. Figure 10: Visualized comparisons of the semantic segmentation task. Red circles highlight the examples of errors that do not occur in the proposed SCSC [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 11
Figure 11. Figure 11: Visualization of resource allocation map for feature vector over different channel bandwidth ratios. The resolution of the left original image is [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: Validation of robustness and generalization. (a) mIoU versus SNR. Comparison of our SCSC (JPEG) and SCSC (JPEG2000) with JPEG and [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]
Figure 13
Figure 13. Figure 13: Visualization results of the PPEN. The first column is the original [PITH_FULL_IMAGE:figures/full_fig_p013_13.png]
Figure 15
Figure 15. Figure 15: Comparisons of the task performance over a practical fading channel [PITH_FULL_IMAGE:figures/full_fig_p014_15.png]
Figure 14
Figure 14. Figure 14: Performance of object detection and image classification over MIMO [PITH_FULL_IMAGE:figures/full_fig_p014_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 51 canonical work pages

  1. [1]

    Semantic communications for future internet: Fundamentals, applications, and challenges,

    W. Yang, H. Du, Z. Q. Liew, W. Y . B. Lim, Z. Xiong, D. Niyato, X. Chi, X. Shen, and C. Miao, “Semantic communications for future internet: Fundamentals, applications, and challenges,” IEEE Commun. Surv. Tutor., vol. 25, no.1, pp. 213–250, Nov. 2022

    work page 2022

  2. [2]

    Toward wisdom-evolutionary and primitive- concise 6G: A new paradigm of semantic communication networks,

    P. Zhang, W. Xu, H. Gao, K. Niu, X. Xu, X. Qin, C. Yuan, Z. Qin, H. Zhao, J. Wei, and F. Zhang, “Toward wisdom-evolutionary and primitive- concise 6G: A new paradigm of semantic communication networks,” Engineering, vol. 8, pp. 60–73, Jan. 2022

    work page 2022

  3. [3]

    Beyond transmitting bits: Context, semantics, and task-oriented communications,

    D. G ¨und¨uz, Z. Qin, I. E. Aguerri, H. S. Dhillon, Z. Yang, A. Yener, and C. B. Chae, “Beyond transmitting bits: Context, semantics, and task-oriented communications,” IEEE J. Sel. Areas Commun. , vol. 41, no. 1, pp. 5–41, Jan. 2022

    work page 2022

  4. [4]

    Deep joint source- channel coding for wireless image transmission,

    E. Bourtsoulatze, D. B. Kurka, and D. G ¨und¨uz, “Deep joint source- channel coding for wireless image transmission,” in Proc. IEEE Int. Conf. Acoust. Speech Signal Process. (ICASSP) , Brighton, UK, May 2019, pp. 4774–4778

    work page 2019

  5. [5]

    Wireless image retrieval at the edge,

    M. Jankowski, D. G ¨und¨uz and K. Mikolajczyk, “Wireless image retrieval at the edge,” IEEE J. Sel. Areas Commun. , vol. 39, no. 1, pp. 89–100, Jan. 2021

    work page 2021

  6. [6]

    Deep learning enabled semantic communications with speech recognition and synthesis,

    Z. Weng, Z. Qin, X. Tao, C. Pan, G. Liu, and G. Y . Li, “Deep learning enabled semantic communications with speech recognition and synthesis,” IEEE Trans. Wireless Commun., vol. 22, no. 9, pp. 6227–6240, Sept. 2023

    work page 2023

  7. [7]

    Nonlinear transform source-channel coding for semantic communications,

    J. Dai, S. Wang, K. Tan, Z. Si, X. Qin, K. Niu, and P. Zhang, “Nonlinear transform source-channel coding for semantic communications,” IEEE J. Sel. Areas Commun. , vol. 40, no. 8, pp. 2300–2316, Jun. 2022

    2022

  8. [8]

    DeepJSCC-f: Deep joint source-channel coding of images with feedback,

    D. Burth Kurka and D. G ¨und¨uz, “DeepJSCC-f: Deep joint source-channel coding of images with feedback,” IEEE J. Sel. Areas Info. Theory , vol. 1, no. 1, pp. 178–193, May 2020

    work page 2020

Show all 54 references
  1. [9]

    Deep joint source-channel coding for adaptive image transmission over MIMO channels,

    H. Wu, Y . Shao, C. Bian, K. Mikolajczyk, and D. G ¨und¨uz, “Deep joint source-channel coding for adaptive image transmission over MIMO channels,” IEEE Trans. Wireless Commun. , vol. 23, no. 10, pp. 15002– 15017, Oct. 2024

    2024

  2. [10]

    Autoencoder-based semantic communication systems with relay channels,

    X. Luo, B. Yin, Z. Chen, B. Xia, and J. Wang, “Autoencoder-based semantic communication systems with relay channels,” in Proc. IEEE Int. Conf. Commun. Workshops (ICC Workshops) , Seoul, South Korea, May 2022, pp. 711–716

    2022

  3. [11]

    Deep joint source-channel coding over cooperative relay networks,

    C. Bian, Y . Shao, H. Wu, and D. G ¨und¨uz, “Deep joint source-channel coding over cooperative relay networks,” arXiv:2211.06705, Nov. 2022. [Online]. Available: https://arxiv.org/abs/2211.06705

    2022 arXiv

  4. [12]

    Robust semantic communications with masked VQ-V AE enabled codebook,

    Q. Hu, G. Zhang, Z. Qin, Y . Cai, G. Yu, and G. Y . Li, “Robust semantic communications with masked VQ-V AE enabled codebook,” IEEE Trans. Wireless Commun., early access, Apr. 2023

    2023

  5. [13]

    Deep source-channel coding for sentence semantic transmission with HARQ,

    P. Jiang, C.-K. Wen, S. Jin, and G. Y . Li, “Deep source-channel coding for sentence semantic transmission with HARQ,” IEEE Trans. Commun., vol. 70, no. 8, pp. 5225–5240, Aug. 2022

    2022

  6. [14]

    DeepJSCC- Q: constellation constrained deep joint source-channel coding,

    T. -Y . Tung, D. B. Kurka, M. Jankowski, and D. G ¨und¨uz, “DeepJSCC- Q: constellation constrained deep joint source-channel coding,” IEEE J. Sel. Areas Inf. Theory , vol. 3, no. 4, pp. 720–731, Dec. 2022

    2022

  7. [15]

    Joint task and data oriented semantic communications: A deep separate source- channel coding scheme,

    J. Huang, D. Li, C. Huang, X. Qin, and W. Zhang, “Joint task and data oriented semantic communications: A deep separate source- channel coding scheme,” arxiv:2302.13580, Jun. 2023. [Online]. Avail- able: https://arxiv.org/abs/2302.13580

    2023 arXiv

  8. [16]

    Versatile semantic coded transmission over MIMO fading channels,

    S. Yao, S. Wang, J. Dai, K. Niu, and P. Zhang, “Versatile semantic coded transmission over MIMO fading channels,” arxiv:2210.16741, Oct. 2022. [Online]. Available: https://arxiv.org/abs/2210.16741

    2022 arXiv

  9. [17]

    Deep joint source-channel and encryption coding: Secure semantic communications,

    T. -Y . Tung and D. G ¨und¨uz, “Deep joint source-channel and encryption coding: Secure semantic communications,” in IEEE Int. Conf. Commun. (ICC), Rome, Italy, May 2023, pp. 5620–5625

    2023

  10. [18]

    Cross-layer wireless multimedia transmission: challenges, principles, and new paradigms,

    M. van Der Schaar and Sai Shankar N, “Cross-layer wireless multimedia transmission: challenges, principles, and new paradigms,” IEEE Wireless Commun., vol. 12, no. 4, pp. 50–58, Aug. 2005

    2005

  11. [19]

    Joint source-channel video transmis- sion,

    F. Zhai and A. K. Katsaggelos, “Joint source-channel video transmis- sion,” in Synthesis Lectures on Image, Video, and Multimedia Processing Series, Series Editor: Al Bovik, Morgan & Claypool Publishers, Sept. 2007

    2007

  12. [20]

    Semantic communication meets edge intelligence,

    W. Yang, Z. Q. Liew, W. Y . B. Lim, Z. Xiong, D. Niyato, X. Chi, X. Cao, and K. B. Letaief, “Semantic communication meets edge intelligence,” IEEE Wireless Commun., vol. 29, no. 5, pp. 28–35, Oct. 2022

    2022

  13. [21]

    Preprocessing enhanced image compression for machine vision,

    G. Lu, X. Ge, T. Zhong, J. Geng, and Q. Hu, “Preprocessing enhanced image compression for machine vision,” arxiv:2206.05650, Jun. 2022. [Online]. Available: https://arxiv.org/abs/2206.05650

    2022 arXiv

  14. [22]

    Semantic preprocessor for image compression for machines,

    M. Yang, L. Herranz, F. Yang, L. Murn, M. G. Blanch, and S. Wan, “Semantic preprocessor for image compression for machines,” in Proc. IEEE Int. Conf. Acoust. Speech Signal Process. (ICASSP), Rhodes Island, Greece, Jun. 2023, pp. 1–5. 16

    2023

  15. [23]

    Transformer-empowered 6G intelligent networks: from massive MIMO processing to semantic communication,

    Y . Wang, Z. Gao, D. Zheng, S. Chen, D. G ¨und¨uz and H. V . Poor, “Transformer-empowered 6G intelligent networks: from massive MIMO processing to semantic communication,” IEEE Wireless Commun. , early access, Feb. 2022

    2022

  16. [24]

    Personalized saliency in task-oriented semantic communications: Image transmission and performance analysis,

    J. Kang, H. Du, Z. Li, Z. Xiong, S. Ma, D. Niyato, and Y . Li, “Personalized saliency in task-oriented semantic communications: Image transmission and performance analysis,” IEEE J. Sel. Areas Commun. , vol. 41, no. 1, pp.186–201, Nov. 2022

    2022

  17. [25]

    BPG image format,

    F. Bellard, “BPG image format,” Apr. 2018. [Online]. Available: https://bellard.org/bpg/

    2018

  18. [26]

    The JPEG still picture compression standard,

    G. K. Wallace, “The JPEG still picture compression standard,” IEEE Trans. Consumer Electron., vol. 38, no. 1, pp.18–34, Feb. 1992

    1992

  19. [27]

    JPEG2000: Standard for interactive imaging,

    D. Taubman and M. Marcellin, “JPEG2000: Standard for interactive imaging,” in Proc. IEEE, vol. 90, no. 8, pp.1336–1357, Aug. 2002

    2002

  20. [28]

    Design of low-density parity check codes for 5G new radio,

    T. Richardson and S. Kudekar, “Design of low-density parity check codes for 5G new radio,” IEEE Commun. Mag., vol. 56, no. 3, pp. 28–34, Mar. 2018

    2018

  21. [29]

    Channel polarization: A method for constructing capacity achieving codes for symmetric binary-input memoryless channels,

    E. Arıkan, “Channel polarization: A method for constructing capacity achieving codes for symmetric binary-input memoryless channels,” IEEE Trans. Inf. Theory, vol. 55, no. 7, pp. 3051–3073, Jul. 2009

    2009

  22. [30]

    DeepJSCC-f: Deep joint source-channel coding of images with feedback,

    D. B. Kurka and D. G ¨und¨uz, “DeepJSCC-f: Deep joint source-channel coding of images with feedback,” IEEE J. Sel. Areas Commun. , vol. 1, no. 1, pp. 178–193, Apr. 2020

    2020

  23. [31]

    Capacity of a mobile multiple antenna communication link in Rayleigh flat fading,

    T. L. Marzetta and B. M. Hochwald, “Capacity of a mobile multiple antenna communication link in Rayleigh flat fading,” IEEE Trans. Inf. Theory, vol. 45, no. 1, pp. 139–157, Jan. 1999

    1999

  24. [32]

    Massive MIMO detection techniques: A survey,

    M. A. Albreem, M. Juntti, and S. Shahabuddin, “Massive MIMO detection techniques: A survey,” IEEE Commun. Surveys Tuts. , vol. 21, no. 4, pp. 3109–3132, 1st Quart., 2019

    2019

  25. [33]

    Sparse-to-dense: Depth prediction from sparse depth samples and a single image,

    F. Ma and S. Karaman, “Sparse-to-dense: Depth prediction from sparse depth samples and a single image,” in Proc. Int. Conf. Robot. Automat. (ICRA), Brisbane, QLD, Australia, May 2018, pp. 4796–4803

    2018

  26. [34]

    Deformable convolutional networks,

    J. Dai, H. Qi, Y . Xiong, Y . Li, G. Zhang, H. Hu, and Y . Wei, “Deformable convolutional networks,” in Proc. IEEE Int. Conf. Comput. Vis. (ICCV) , Venice, Italy, Oct. 2017, pp. 764–773

    2017

  27. [35]

    Strip pooling: Rethinking spatial pooling for scene parsing,

    Q. Hou, L. Zhang, M.-M. Cheng, and J. Feng, “Strip pooling: Rethinking spatial pooling for scene parsing,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Seattle, W A, USA, Jun. 2020, pp. 4003–4012

    2020

  28. [36]

    The MIMO iterative waterfilling algorithm,

    G. Scutari, D. P. Palomar, and S. Barbarossa, “The MIMO iterative waterfilling algorithm,” IEEE Trans. Signal Process. , vol. 57, no. 5, pp. 1917–1935, May 2009

    1917

  29. [37]

    Globally optimal linear precoders for finite alphabet signals over complex vector Gaussian channels,

    C. Xiao, Y . R. Zheng, and Z. Ding, “Globally optimal linear precoders for finite alphabet signals over complex vector Gaussian channels,” IEEE Trans. Signal Process., vol. 59, no. 7, pp. 3301–3314, Jul. 2011

    2011

  30. [38]

    An overview of 5G advanced evolution in 3GPP release 18,

    X. Lin, “An overview of 5G advanced evolution in 3GPP release 18,” in IEEE Commun. Stand. , vol. 6, no. 3, pp. 77–83, Sept. 2022

    2022

  31. [39]

    Deep-unfolding for next-generation transceivers,

    Q. Hu, Y . Cai, G. Zhang, G. Yu, and G. Y . Li, “Deep-unfolding for next-generation transceivers,” arxiv:2305.08303, May 2023. [Online]. Available: https://arxiv.org/abs/2305.08303

    2023 arXiv

  32. [40]

    Finite-alphabet precoding for massive MU-MIMO with low-resolution DACs,

    C. -J. Wang, C. -K. Wen, S. Jin and S. -H. Tsai, “Finite-alphabet precoding for massive MU-MIMO with low-resolution DACs,” IEEE Trans. Wireless Commun., vol. 17, no. 7, pp. 4706–4720, Jul. 2018

    2018

  33. [41]

    Model-Driven Deep Learning for MIMO Detection,

    H. He, C. -K. Wen, S. Jin and G. Y . Li, “Model-Driven Deep Learning for MIMO Detection,” IEEE Trans. Signal Process., vol. 68, pp. 1702– 1715, Feb. 2020

    2020

  34. [42]

    The cityscapes dataset for semantic urban scene understanding,

    M. Cordts, M. Omran, S. Ramos, T. Rehfeld, M. Enzweiler, R. Be- nenson, U. Franke, S. Roth, and B. Schiele, “The cityscapes dataset for semantic urban scene understanding,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR) , Las Vegas, NV , USA, Jun. 2016, pp. 3213– 3223

    2016

  35. [43]

    Semantic segmentation of outdoor panoramic images,

    S. Orhan and Y . Bastanlar, “Semantic segmentation of outdoor panoramic images,” Signal Image Video Process. , vol. 16, pp. 643–650, Aug. 2021

    2021

  36. [44]

    Unifying terrain awareness through real-time semantic segmentation,

    K. Yang, L. M. Bergasa, E. Romera, R. Cheng, T. Chen, and K. Wang, “Unifying terrain awareness through real-time semantic segmentation,” in Proc. IEEE Intell. Vehicles Symp. (IV) , Changshu, China, Jun. 2018, pp. 1033–1038

    2018

  37. [45]

    Adam: A method for stochastic optimiza- tion,

    D. P. Kingma and J. L. Ba, “Adam: A method for stochastic optimiza- tion,” in Proc. Int. Conf. Learn. Rep. (ICLR), San Diego, CA, USA, 2015, pp. 1–15

    2015

  38. [46]

    A general 3-D non-stationary 5G wireless channel model

    S. Wu, C. Wang, E. Aggoune, M. Alwakeel, and X.You, “A general 3-D non-stationary 5G wireless channel model”, IEEE Trans. Commun. , vol. 66, no. 7, pp. 3065–3078, Jul. 2018

    2018

  39. [47]

    Wire- less image transmission using deep source channel coding with at- tention modules,

    J. Xu, B. Ai, W. Chen, A. Yang, P. Sun, and M. Rodrigues, “Wire- less image transmission using deep source channel coding with at- tention modules,” arxiv:2012.00533, Apr. 2022. [Online]. Available: https://arxiv.org/abs/2012.00533

    2012 arXiv

  40. [48]

    Joint task and data oriented semantic communications: A deep separate source-channel coding scheme,

    J. Huang, D. Li, C. Huang, X. Qin, and W. Zhang, “Joint task and data oriented semantic communications: A deep separate source-channel coding scheme,” IEEE Internet Things J. , vol. 11, no. 2, pp. 2255–2272, Jan. 2024

    2024

  41. [49]

    Calculation of average PSNR differences between RD- curves,

    G. Bjontegaard, “Calculation of average PSNR differences between RD- curves,” VCEG-M33, Austin, TX, USA, Apr. 2001, pp. 2–4

    2001

  42. [50]

    CLIC 2021: Challenge on learned image compression,

    “CLIC 2021: Challenge on learned image compression,” [Online]. Available: http://compression.cc, 2021

    2021

  43. [51]

    Grad-cam: Visual explanations from deep networks via gradient- based localization,

    R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Grad-cam: Visual explanations from deep networks via gradient- based localization,” in Proc. IEEE Int. Conf. Comput. Vis. (ICCV), Venice, Italy, Oct. 2017, pp. 618–626

    2017

  44. [52]

    Full resolution image compression with recurrent neural networks,

    G. Toderici, D. Vincent, N. Johnston, S. J. Hwang, D. Minnen, J. Shor, and M. Covell, “Full resolution image compression with recurrent neural networks,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Honolulu, Hawaii, USA, Jul. 2017, pp. 5435–5443

    2017

  45. [53]

    EfficientDet: Scalable and efficient ob- ject detection,

    M. Tan, R. Pang, and Q. V . Le, “EfficientDet: Scalable and efficient ob- ject detection,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Seattle, W A, USA, Jun. 2020, pp. 10781–10790

    2020

  46. [54]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Las Vegas, NV , USA, Jun. 2016, pp. 770–778

    2016

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.