Wireless Image Retrieval at the Edge
Open accessThe DOI matches its Crossref record; 42 of the 44 references it cites were found in a registry or an index, and 1 are not checked yet.
Abstract
We study the image retrieval problem at the wireless edge, where an edge device captures an image, which is then used to retrieve similar images from an edge server. These can be images of the same person or a vehicle taken from other cameras at different times and locations. Our goal is to maximize the accuracy of the retrieval task under power and bandwidth constraints over the wireless link. Due to the stringent delay constraint of the underlying application, sending the whole image at a sufficient quality is not possible. We propose two alternative schemes based on digital and analog communications, respectively. In the digital approach, we first propose a deep neural network (DNN) aided retrieval-oriented image compression scheme, whose output bit sequence is transmitted over the channel using conventional channel codes. In the analog joint source and channel coding (JSCC) approach, the feature vectors are directly mapped into channel symbols. We evaluate both schemes on image based re-identification (re-ID) tasks under different channel conditions, including both static and fading channels. We show that the JSCC scheme significantly increases the end-to-end accuracy, speeds up the encoding process, and provides graceful degradation with channel conditions. The proposed architecture is evaluated through extensive simulations on different datasets and channel conditions, as well as through ablation studies.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“The analog JSCC approach was presented in [234] to compress and then code the feature vectors, followed by leveraging the received perturbed signal directly for wireless image retrieval at the edge server via a fully-connected neural network.”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [234]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 1×
“Following the state-of-the-art retrieval methods [17], [21] we employ the ResNet-50 network [34], pretrained on Ima-”
MethodA Mathematical Theory of Communication
Shannon, C. E. (1948) · Bell System Technical Journal · doi:10.1002/j.1538-7305.1948.tb01338.x · Cited by 88,486
ImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380
Information Theory and Reliable Communication.
Lee, P. M., & Gallager, R. T. (1970) · Journal of the Royal Statistical Society. Series A (General) · doi:10.2307/2344305 · Cited by 5,399
Scalable Person Re-identification: A Benchmark
Zheng, L., Shen, L., Tian, L., Wang, S., Wang, J., & Tian, Q. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.133 · Cited by 4,506 · cited 1×
“Market-1501 [40] contains 32217 images of 1501 pedestrians taken from a total of six cameras in front of a”
BackgroundLearning internal representations by error propagation
RUMELHART, D. E., HINTON, G. E., & WILLIAMS, R. J. (1988) · Readings in Cognitive Science · doi:10.1016/b978-1-4832-1446-7.50035-2 · Cited by 3,002 · cited 1×
“More recently, JSCC schemes based on autoencoders [30], which are DNNs aimed at unsupervised data coding, have been introduced [2], [3], [31], [32], and are shown to provide comparable or better performance than state-of-the-art digital schemes.”
BackgroundIn Defense of the Triplet Loss for Person Re-Identification
Hermans, A., Beyer, L., & Leibe, B. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1703.07737 · Cited by 2,687
DeepReID: Deep Filter Pairing Neural Network for Person Re-identification
Li, W., Zhao, R., Xiao, T., & Wang, X. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.27 · Cited by 2,611 · cited 8×
“In order to measure the performance of the retrieval task, we employ three widely used datasets: CUHK03 [39] is a benchmark for person retrieval that contains 14096 images of 1467 identities taken from two different camera views.”
MethodInfluentialGray, R. M., & Neuhoff, D. L. (1998) · IEEE Transactions on Information Theory · doi:10.1109/18.720541 · Cited by 1,642
Deep Joint Source-Channel Coding for Wireless Image Transmission
Bourtsoulatze, E., Burth Kurka, D., & Gunduz, D. (2019) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2019.2919300 · Cited by 1,476
Cited by
Items 101–200 of 279
Interference Suppressed NOMA for Semantic-Aware Communication Networks
Zhang, Y., Zhong, R., Liu, Y., Xu, W., & Zhang, P. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3371775 · Cited by 17
“The other way is to directly transmit the semantic stream as analog signals, which is widely adopted in existing semantic-related works [19], [28].”
BackgroundAttention-Aware Semantic Communications for Collaborative Inference
Im, J., Kwon, N., Park, T., Woo, J., Lee, J., & Kim, Y. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3440313 · Cited by 27
“Due to the consistent layer dimensions of ViTs, conventional methods of collaborative inference [2], [3], [4], [5], [6], [19], which typically partition a single DNN model, are ineffective at reducing communication cost for ViT models.”
BackgroundMethodMulti-User Goal-Oriented Communications With Energy-Efficient Edge Resource Management
Binucci, F., Banelli, P., Lorenzo, P. D., & Barbarossa, S. (2023) · IEEE Transactions on Green Communications and Networking · doi:10.1109/tgcn.2023.3275199 · Cited by 19
“The authors of [20] presented a scheme for image retrieval where the extracted vector features are directly mapped to the channel input symbols, without resorting to any channel coding technique, and the server retrieves the most relevant images directly from the noisy channel output.”
BackgroundClassification-Oriented Distributed Semantic Communication for Multivariate Time Series
Zhao, B., Xing, H., Wang, X., Xiao, Z., & Xu, L. (2023) · IEEE Signal Processing Letters · doi:10.1109/lsp.2023.3265330 · Cited by 19
“Traditional encoder and decoder networks [6], [7], [8], like convolutional neural networks (CNNs) [11], [12] or Transformers [13], [14], [15], may not be able to capture multi-scale features by an identical neural architecture, possibly leading to imbalanced attention to different variables.”
BackgroundAn Asynchronous Multi-Task Semantic Communication Method
Tian, Z., Vo, H. K., Zhang, C., Min, G., & Yu, S. (2023) · IEEE Network · doi:10.1109/mnet.2023.3321547 · Cited by 18
“In practical, the DL-based semantic codecs can achieve both joint source and channel coding (JSCC) or serving as replacements of conventional encoder (decoder) to code (decode) the transmitted information [9].”
BackgroundMachine Learning Enabled Heterogeneous Semantic and Bit Communication
Zhang, M., Zhong, R., Mu, X., & Liu, Y. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3397635 · Cited by 17
“In several recent works, SemCom was proved to be able to enhance the transmission efficiency in object recognition, wireless image retrieval, and video conferencing [14], [15], [16].”
BackgroundAdversarial Attacks Against Shared Knowledge Interpretation in Semantic Communications
Hoang, V.-T., Nguyen, V.-L., Chang, R., Lin, P., Hwang, R., & Duong, T. Q. (2025) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2025.3528891 · Cited by 21
“The authors of [15], [16] developed a framework employing an edge server to categorize pictures and decrease the required transmission bandwidth.”
High Quality Compression and Transmission of Remote Sensing Images Based on Semantic Communication
Jiang, Y., Kun, X., Ouyang, Y., Wen, J., Zhang, G., Liang, W., & Feng, Q. (2025) · IEEE Transactions on Sustainable Computing · doi:10.1109/tsusc.2025.3544249 · Cited by 16
Dynamic Resource Allocation for Multi-User Goal-oriented Communications at the Wireless Edge
Binucci, F., Banelli, P., Lorenzo, P. D., & Barbarossa, S. (2022) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco55093.2022.9909863 · Cited by 17
“Also, goaloriented semantic communications were recently proposed in [9], [10].”
BackgroundDigital Semantic Device-Edge Co-Inference With Task-Oriented ARQ
Li, X., Bi, S., Wang, S., Li, X., & Zhang, Y. A. (2024) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2024.3390213 · Cited by 20
“Atypicalapproachinvolvesreplacingthequantizationprocesswith additiveuniformnoiseandusesreparameterizationmethodtoenable gradient-basedoptimizationduringbackpropagation[7],[8].”
Device-Edge Digital Semantic Communication with Trained Non-Linear Quantization
Guo, L., Chen, W., Sun, Y., & Ai, B. (2023) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2023-spring57618.2023.10200355 · Cited by 19
“[2] proposed a task-oriented DL-based communication system for image retrieval, which directly maps the feature vectors into channel symbols.”
BackgroundInfluentialChen, W., Shao, S., Yang, Q., Zhang, Z., & Zhang, P. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3608054 · Cited by 14
“ThisintegrationofDL enablesSemComsystemstoachieveexcellentperformance intransmittingvarioustypesofinformation,suchastext[7], [8],speech[8],[9],image[10],[11],[12],video[13],and multimodaldata[14].”
Distributed Deep Joint Source-Channel Coding with Decoder-Only Side Information
Yilmaz, S. F., Özyılkan, E., Gündüz, D., & Erkip, E. (2024) · IEEE International Conference on Machine Learning for Communication and Networking (ICMLCN) · doi:10.1109/icmlcn59089.2024.10625214 · Cited by 14
“This data-driven communication scheme has later been extended to different channel models [14], [15], different source signals [16], [17], inference problems [18], multi-user scenarios [15], [19] and as well as to perceptual quality-oriented image transmission [20], [21].”
BackgroundCompression Ratio Learning and Semantic Communications for Video Imaging
Zhang, B., Qin, Z., & Li, G. Y. (2024) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2024.3405853 · Cited by 13
“In earlier task-oriented communications [26], the RAN is not implemented and all source data is transmitted with the same communication costs.”
BackgroundResultAdaSem: Adaptive Goal-Oriented Semantic Communications for End-to-End Camera Relocalization
Liao, Q., & Tung, T.-Y. (2024) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom52122.2024.10621397 · Cited by 13
“…results, deep learning-based JSCC (DeepJSCC) has shown advantages over the standard communications protocols in terms of both lower distortion and communication overhead in many use cases, such as video transmission [6], image retrieval [7], person identification [8], and machine translation [9].”
BackgroundPrivate Collaborative Edge Inference via Over-the-Air Computation
Yilmaz, S. F., Hasırcıoglu, B., Qiao, L., & Gündüz, D. (2025) · IEEE Transactions on Machine Learning in Communications and Networking · doi:10.1109/tmlcn.2025.3526551 · Cited by 18
“While a growing body of work studies distributed training over wireless networks, the literature on distributed wireless inference, particularly using deep learning techniques, is relatively limited [8], [9], [10].”
Li, N., Iosifidis, A., & Zhang, Q. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3518399 · Cited by 14
“Similarly, Jankowski et al. [24] proposed DeepJSCC to compress the intermediate feature by using a CNN-based encoder.”
Hu, H., Song, K., Fan, R., Zhan, C., Xu, J., & Yang, J. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3539226 · Cited by 12
“This surge in visual information poses significant challenges to current communication systems in terms of storage and transmission overhead [1], [2].”
Cross-Domain Lossy Compression as Entropy Constrained Optimal Transport
Liu, H., Zhang, G., Chen, J., & Khisti, A. J. (2022) · IEEE Journal on Selected Areas in Information Theory · doi:10.1109/jsait.2022.3229670 · Cited by 21
“We note that deep learning based compression for classification tasks has been considered in [30], [31].”
BackgroundAn Image Adaptive Rate Mechanism in Semantic Communication for Image Endogenous Semantics
Shi, G., Li, H., Gao, D., Yang, M., & Dong, Y. (2024) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2024.3396426 · Cited by 13
“Such advancements have not only revolutionized communication systems but have also catalyzed enhancements in various downstream applications [7], [8], [9], [10], [11], [12], ranging from wireless networking to economic models of communication, indicating a promising trajectory toward…”
BackgroundSemantic Communications: A New Paradigm for Networked Intelligence
Zhang, Y., Wang, F., Xu, W., & Liu, C. (2022) · Machine learning for signal processing ... · doi:10.1109/mlsp55214.2022.9943480 · Cited by 11
“…semantic communications, this paradigm further reduces the data volume and is promising to serve diverse intelligent tasks, whose effectiveness has been demonstrated under single-user cases with different tasks, including speech recognition [22], image retrieval [23], and image classification [24].”
BackgroundSelf-Critical Alternate Learning-Based Semantic Broadcast Communication
Lu, Z., Li, R., Lei, M., Wang, C., Zhao, Z., & Zhang, H. (2024) · IEEE Transactions on Communications · doi:10.1109/tcomm.2024.3487513 · Cited by 13
“For example, Jankowski et al. [29], [30] target the identification of individuals or vehicles and design two alternative (digital and analog communications-based) schemes to enhance retrieval accuracy.”
BackgroundDeep Joint Source Channel Coding for Privacy-Aware End-to-End Image Transmission
Letafati, M., Kalkhoran, S. A. A., Erdemir, E., Khalaj, B. H., Behroozi, H., & Gündüz, D. (2025) · IEEE Transactions on Machine Learning in Communications and Networking · doi:10.1109/tmlcn.2025.3564907 · Cited by 10
“Recently, intelligent multimedia transmission is receiving much attention due to its various applications in XR, metaverse, and surveillance systems [3].”
Task-Oriented Low-Label Semantic Communication With Self-Supervised Learning
Gu, R., Xu, W., Yang, Z., Niyato, D. T., & Yener, A. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3574468 · Cited by 11
“Most existing task-oriented semantic communication studies utilize supervised learning algorithms for network training, implemented through single-phase [9], [18], [19] or multi-phase frameworks [20], [21].”
AirNet: Neural Network Transmission Over the Air
Jankowski, M., Gündüz, D., & Mikolajczyk, K. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3388329 · Cited by 15
“DeepJSCC has also been applied to other downstream tasks, such as remote classification [3], [4], retrieval [41], or anomaly detection [5] problems.”
MethodSelf-citationLiu, M., Wang, Y., & Wang, T. (2025) · IEEE Wireless Communications Letters · doi:10.1109/lwc.2025.3527145 · Cited by 10
“In [7], ten RC modules of VGG were retained in the SemEnc, with structured pruning to further reduce complexity.”
InfluentialAnalog Joint Source-Channel Coding for Distributed Functional Compression using Deep Neural Networks
Saidutta, Y. M., Abdi, A., & Fekri, F. (2021) · IEEE International Symposium on Information Theory (ISIT) · doi:10.1109/isit45174.2021.9517797 · Cited by 12
Distributed Hierarchical Deep Reinforcement Learning for Semantic-Aware Resource Allocation
Yu, K., He, Q., Yu, C., Yang, X., & Wu, G. (2025) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2025.3571485 · Cited by 12
“Furthermore, some researches studied task-oriented semantic communication systems such as recognition [30], classi-fication [31], [32], etc.”
Task-Oriented Edge Networks: Decentralized Learning Over Wireless Fronthaul
Lee, H., & Kim, S. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3347234 · Cited by 9
“For this reason, single EN systems in [15], [17], [19], and [20] cannot be straightforwardly applied to multi-EN task-oriented networks.”
BackgroundMethodLearning Precoding for Semantic Communications
Guo, J., & Yang, C. (2022) · IEEE International Conference on Communications workshops · doi:10.1109/iccworkshops53468.2022.9814464 · Cited by 9
Impact of Quantization Noise on CNN-based Joint Source-Channel Coding and Modulation
Matsumoto, K., Inoue, Y., Hara–Azumi, Y., Maruta, K., Nakayama, Y., & Hisano, D. (2023) · IEEE Consumer Communications and Networking Conference · doi:10.1109/ccnc51644.2023.10060488 · Cited by 9
“CNNbased methods have been actively studied in recent years [5]– [9].”
MethodDiao, Y., Zhang, Y., Martini, D. D., Zhao, P. G., & Li, L. (2025) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2025.3574605 · Cited by 13
“This architecture allows for initial data processing on the device, followed by more intensive computations at the edge, thus reducing latency and improving overall system efficiency [34], [35], [36], [37], [38], [39], [40].”
Wang, Y., Li, R., Wang, C., Ye, J., Feng, C., & Guo, S. (2025) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2025.3586462 · Cited by 10
Resilient Image Semantic Communication Based on Rate-Optimized Information Bottleneck Theory
Wang, W., Wang, C., Zhang, Z., Xu, W., Zhang, P., Jiang, F., Xu, J., Cai, Y., Pang, M., Xu, L., et al. (2025) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2025.3626331 · Cited by 8
“Jankowski et al. proposed a joint feature compression and transmission system to handle the limited computing resources on edge servers [31].”
Zero-Shot Semantic Communication With Multimodal Foundation Models
Hu, J., Wu, H., Zhang, W., Wang, F., Xu, W., Gao, H., & Gündüz, D. (2025) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2025.3632893 · Cited by 8
“Particularly,aDeepJSCCscheme isproposedin[7]tocommunicateretrieval-orientedsemanticfeatures.”
Task-Oriented Explainable Semantic Communications Based on Structured Scene Graphs
Sun, S., Qin, Z., Xie, H., & Tao, X. (2023) · Globecom · doi:10.1109/globecom54140.2023.10436793 · Cited by 10
“Jankowski et al. [7] and Xie et al. [8] proposed retrieval-oriented and visual question answering-oriented image compression schemes under constrained power and bandwidth.”
BackgroundSuper-Resolution Reconstruction for Cross-Modal Communications in Industrial Internet of Things
Liu, H., Wei, X., Gao, Y., Zhou, L., & Qian, Y. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3388260 · Cited by 8
“Drawing inspiration from the work [15], we implement a strategy where we individually train each component of the network models and then combine them for joint training.”
BackgroundMethodTinyAirNet: TinyML Model Transmission for Energy-Efficient Image Retrieval From IoT Devices
Shiraishi, J., Thorsager, M., Pandey, S. R., & Popovski, P. (2024) · IEEE Communications Letters · doi:10.1109/lcomm.2024.3436816 · Cited by 12
“The TinyAirNet can be applicable to a variety of IoT data collection scenarios to reduce wasteful energy consumption, including object detection and vehicles at the edge [11], AI empowered IoT sensing [4], and wild-life animal monitoring/tracking [12].”
BackgroundMulti-Level Reliability Interface for Semantic Communications Over Wireless Networks
Tung, T.-Y., Esfahanizadeh, H., Du, J., & Viswanathan, H. (2025) · IEEE Transactions on Communications · doi:10.1109/tcomm.2025.3535903 · Cited by 8
“This technique has been successfully applied and evaluated across various domains such as image [8], video [9], text [10], [11], generative applications [12], and image classification [13], and it has demonstrated security against eavesdroppers [14], [15], [16].”
Learning to Communicate with Intent: An Introduction
Gutierrez-Estevez, M. A., Wu, Y., & Zhou, C. (2023) · IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC) · doi:10.1109/pimrc56721.2023.10294013 · Cited by 8
“Another research direction is studying particular applications that benefit from GOCom, such as multi-agent RL (MARL) environments [8], remote image retrieval [9], or ar X iv :2 21 1.”
BackgroundRemote Anomaly Detection in Industry 4.0 Using Resource-Constrained Devices
Kalør, A. E., Michelsanti, D., Chiariotti, F., Tan, Z., & Popovski, P. (2021) · IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC) · doi:10.1109/spawc51858.2021.9593188 · Cited by 8
Personalizing Semantic Communication: A Foundation Model Approach
Chen, Z., Yang, H. H., Chong, K. F. E., & Quek, T. Q. S. (2024) · IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC) · doi:10.1109/spawc60668.2024.10694080 · Cited by 7
“Moreover, existing DL-based methods are only capable of providing generic semantic services (e.g., semantic interpretation and reconstruction), whether in single-task scenarios [6], [9], [10], or in multi-user semantic communications [11], [12].”
MethodTwo‐View Image Semantic Cooperative Nonorthogonal Transmission in Distributed Edge Networks
Wang, W., Cai, D., Dong, Z., Yu, L., Xu, Y., & Liu, Z. (2024) · International Journal of Intelligent Systems · doi:10.1155/int/5081017 · Cited by 7
Bandwidth-Adaptive Semantic Communication System with Non-ROI Image Compression
Wu, C., Wu, C., Wu, J., Lin, Y., Du, Z., Si, H., & Bao, W. (2024) · IEEE Intl Conf on Dependable, Autonomic and Secure Computing, Intl Conf on Pervasive Intelligence and Computing, Intl Conf on Cloud and Big Data Computing, Intl Conf on Cyber Science and Technology Congress (DASC/PiCom/CBDCom/CyberSciTech) · doi:10.1109/cyberscitech64112.2024.00045 · Cited by 7
“Jankowski et al. [7] proposed an image retrieval system based on distributed edge inference.”
Model-Enabled Task-Oriented Semantic Communications Through Knowledge Synchronization
Tian, Z., Wang, W., Zhang, C., & Yu, S. (2025) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2025.3531461 · Cited by 7
“Existing SemCom methods are predominantly task-oriented, where each single SemCom system is designed to perform a specific communication task, such as image retrieval [5] and image classification [6].”
Semantic Importance-Aware Image Transmission in V2X Networks
Cai, A., Wang, L., Lin, Y., Liu, C., & Qian, P. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3582442 · Cited by 9
“While image retrieval [24], image classification [25], and speech recognition [26] focus on transmitting task information without reconstructing the source image.”
Chen, Y., Hong, X., Cui, B., & Peng, R. (2023) · Journal of Marine Science and Engineering · doi:10.3390/jmse11122280 · Cited by 7
Synchronous Semantic Communications for Video and Speech
Tian, Y., Ying, J., Qin, Z., Jin, Y., & Tao, X. (2024) · International Conference on Communications · doi:10.1109/icc51166.2024.10622487 · Cited by 6
“For specific intelligent tasks such as image retrieval [3] and speech recognition [4], task-oriented semantic communication only transmits task-related information, which reduces the transmission overhead.”
BackgroundHarnessing the Power of Pre-Trained Models for Efficient Semantic Communication of Text and Images
Kutay, E., & Yener, A. (2025) · Entropy · doi:10.3390/e27080813 · Cited by 6
“This architecture adopts an autoencoder-based design with end-to-end training with task-specific loss functions [15–18].”
Frame-Based Zero-Shot Semantic Channel Equalization for AI-Native Communications
Fiorellino, S., Battiloro, C., Strinati, E. C., & Lorenzo, P. D. (2026) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2026.3658783 · Cited by 6
“Another related line is deep joint source–channel coding (JSCC) [32], where learned feature vectors are directly mapped to channel symbols.”
Pase, F., Gündüz, D., & Zorzi, M. (2022) · IEEE International Symposium on Information Theory (ISIT) · doi:10.1109/isit50566.2022.9834399 · Cited by 8
Learning Multi-Rate Vector Quantization for Remote Deep Inference
Malka, M., Ginzach, S., & Shlezinger, N. (2023) · IEEE International Conference on Acoustics, Speech, and Signal Processing Workshops (ICASSPW) · doi:10.1109/icasspw59220.2023.10193526 · Cited by 7
“An encoderdecoder architecture is designed for the sensing and inferring devices, while constraining the latent features to a compressed representation via uniform quantization [9,10], scalar quantization [11], and vector quantization [12, 13].”
BackgroundMethodEnd-to-End Learning-Based Wireless Image Recognition Using the PyramidNet in Edge Intelligence
Lee, K., & Yu, N. Y. (2023) · IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC) · doi:10.1109/pimrc56721.2023.10293934 · Cited by 7
“• JSCC AE [14]: The ResNet50 without network splitting is used for feature extraction.”
BackgroundMethodInfluentialFor an Energy Sober Network that Contributes to a Sustainable Society
Elayoubi, S. E., & Hardouin, E. (2023) · IEEE Future Networks World Forum (FNWF) · doi:10.1109/fnwf58287.2023.10520513 · Cited by 6
“Another stream of goal-oriented works is related to the joint optimization of the network and application (see for instance [15] for image retrieval at the edge with communication constraints, [8] for vehicle platooning controller design under channel imperfections, and the 6G paradigm of joint…”
BackgroundRemote Inference Over Dynamic Links via Adaptive Rate Deep Task-Oriented Vector Quantization
Ben-Knaan, E. F., Malka, M., Ginzach, S., & Shlezinger, N. (2025) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2025.3603943 · Cited by 8
“Such designs employ DNN-based encoder-decoder architectures, while constraining the latent features to a fixed bit representation via uniform quantization [19], [20], [21], scalar quantization [22], [23], [24], and vector quantization [25], [26].”
Deep Joint Source-Channel Coding Based on Semantics of Pixels for Wireless Image Transmission
Sun, Q., Guo, C., Yang, Y., Tang, R., & Liu, C. (2023) · IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC) · doi:10.1109/pimrc56721.2023.10293843 · Cited by 6
“For example, Gunduz et al. [2] demonstrated image retrieval tasks and showed that near noiseless performance can be achieved under very harsh signal-to-noise ratio (SNR) and bandwidth constraints.”
BackgroundMethodEnhancing Communication Reliability from the Semantic Level under Low SNR
Liu, Y., Zhang, Y., Luo, P., Jiang, S., Cao, K., Zhao, H., & Wei, J. (2022) · Electronics · doi:10.3390/electronics11091358 · Cited by 8
An Elastic Coding and Decoding Method for Satellite Remote Sensing Image Semantic Transmission
Zhang, Z., Zhang, S., Li, H., Shi, G., Xue, J., & Li, B. (2025) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2025.3632547 · Cited by 5
“Recently, many researchers have applied semantic communication method into the text and image transmission tasks [18], [19], [20], [21], [22], [23], [24], [25].”
Task-Generalizable Deep Joint Source-Channel Coding via Hierarchical Masked Autoencoders
Dai, S., Chen, X., Deng, X., & Lin, S. (2026) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2026.3691535 · Cited by 5
“Jankowski et al. [17] proposed a classification-oriented wireless image transmission framework for person re-identification task optimized via a cross-entropy loss.”
Similarity Retrieval Based on Image Background Analysis
Zhu, C., Jiang, W., Zhou, W., & Xiao, H. Y. (2022) · International Journal of Software Science and Computational Intelligence · doi:10.4018/ijssci.309426 · Cited by 5
Learning Multi-Rate Task-Oriented Communications Over Symmetric Discrete Memoryless Channels
Zhang, A., & Guo, S. (2024) · IEEE Communications Letters · doi:10.1109/lcomm.2024.3450598 · Cited by 5
“Several paradigms have been proposed to reduce the amount of required information to perform the task, such as information bottleneck [2], feature compression [4], semantics-aware coding [5].”
BackgroundFederated Knowledge Distillation Enabled Image Semantic Communication
Xu, X., Zheng, K., Dou, H., Chen, M., & Wang, L. (2024) · Globecom · doi:10.1109/globecom52923.2024.10901608 · Cited by 5
“The device-side image semantic feature extractor is shufflenet v2, the JSCC encoder and decoder reference the design in [17], and the image reconstructor consists of multiple 2D transposed convolutional layers.”
Lee, H., Ahn, H., & Park, Y. D. (2022) · ICT Express · doi:10.1016/j.icte.2022.04.001 · Cited by 6
Over-the-Air Inference Through Analog Computation Over Multi-Hop MIMO Networks
Bian, C., Hua, M., & Gündüz, D. (2025) · IEEE Wireless Communications Letters · doi:10.1109/lwc.2025.3602209 · Cited by 8
“T HERE has been a growing trend toward the integration of computation and wireless communications to enhance the performance of future networks [1], [2].”
SSFU: Selective Semantic Feature Unlearning for Federated Learning in 6G Internet of Things Systems
Daluwatta, W., Khalil, I., Edirimannage, S., Elvitigala, C., Ng, J. S., & Niyato, D. T. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3625756 · Cited by 4
“Beyond text, SemCom systems have been developed for image [16], [17] and speech modalities [18], showcasing their cross-domain applicability.”
Lin, C., Guo, Y., Hao, J., & Zhang, Z. (2024) · Internet of Things · doi:10.1016/j.iot.2024.101383 · Cited by 4
Multiuser Semantic Communication With Federated Learning for Intelligent Search Service
Sun, M., Wu, D., Zhang, P., & Wang, R. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3555437 · Cited by 4
“We consider two channel models, i.e., the AWGN and Rayleigh fading channel, which are modeled as nontrainable neural network layers, respectively, [18], [20].”
InfluentialAligning Task- and Reconstruction-Oriented Communications for Edge Intelligence
Diao, Y., Zhang, Y., She, C., Zhao, G., & Li, L. (2025) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2025.3559121 · Cited by 8
“Emergent methods based on JSCC have shown great potential to solve this problem [11], [12], [13].”
InfluentialRobust Semantic Communication via Adversarial Training
Wei, K., Xie, R., Xu, W., Lu, Z., & Xiao, H. (2025) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2025.3585536 · Cited by 4
“The combination of the SN and CN forms a Joint Source-Channel Coding (JSCC) system, which is an excellent semantic communication framework, outperforming traditional separate source coding and channel coding (SSCC) approaches [8], [9], [19], [20].”
Performance Optimization for Task-Oriented Communications
Liu, C., Guo, C., & Yang, Y. (2024) · International Conference on Communications · doi:10.1109/icc51166.2024.10622483 · Cited by 4
“Jankowski et al. [6] considered image based re-identification for persons or cars as the communication task, where two schemes were proposed to improve the retrieval accuracy.”
BackgroundPilot-Free Semantic Communication Over Multi-User Mimo Fading Channels
Chen, W., Yang, Q., Yang, Z., Duan, Y., & Zhang, Z. (2024) · Proceedings - International Conference on Image Processing · doi:10.1109/icip51287.2024.10648086 · Cited by 5
Generative AI Driven Task-Oriented Adaptive Semantic Communications
Fu, Y., Cheng, W., Wang, J., Yin, L., & Zhang, W. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3644355 · Cited by 9
“For example, the authors of [9] proposed a JSCC-based image retrieval scheme, where the decoded feature vectors are directly fed to the edge server for retrieval inference.”
Goal-oriented Compression with a Constrained Decoder
Kobus, S., Tung, T.-Y., & Gündüz, D. (2023) · IEEE International Symposium on Information Theory (ISIT) · doi:10.1109/isit54713.2023.10206740 · Cited by 4
Robust Task-Oriented Communication with Semantic-Aware Masking and Discrete Codebook
Li, Y., Jin, Z., Song, T., Song, X., & Hu, J. (2025) · IEEE Wireless Communications and Networking Conference · doi:10.1109/wcnc61545.2025.10978541 · Cited by 3
“In applications focused on specific tasks, Only semantic information related to the task is extracted and sent, which is then utilized for decision-making [9].”
Task-oriented semantic communication towards industrial anomaly detection
Lin, C., Zhang, P., Wang, R., & Yang, Z. (2025) · Journal of the Franklin Institute · doi:10.1016/j.jfranklin.2025.107886 · Cited by 3
Deep Joint Source and Channel Coding
Wu, H., Bian, C., Yulin, S., & Gündüz, D. (2025) · Wiley eBooks · doi:10.1002/9781394247912.ch4 · Cited by 3
Unveiling Covert Semantics: Joint Source-Channel Coding Under a Covertness Constraint
Bounhar, A., Sarkiss, M., & Wigger, M. (2024) · Globecom · doi:10.1109/globecom52923.2024.10901642 · Cited by 3
Deep Reinforcement Learning- and Information Bottleneck-Enabled Task-Oriented Semantic Communication
Zhao, T., Li, F., Du, H., & Sun, L. (2025) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2025.3623199 · Cited by 5
“Jankowski et al. [17] introduced a joint source-channel coding (JSCC) method for image retrieval tasks under the power and bandwidth constraints of wireless links.”
Learning to Interfere in Non-Orthogonal Multiple-Access Joint Source-Channel Coding
Yilmaz, S. F., Karamanlı, C., & Gündüz, D. (2026) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2026.3688951 · Cited by 3
“Restrictions apply. tasks [12], and perceptual quality-focused image transmission [13], [14].”
EDIR: Efficient Distributed Image Retrieval of Novel Objects in Mobile Networks
Felemban, N. M., Mehmeti, F., Porta, T. L., & Kwon, H. (2023) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2023.3257268 · Cited by 3
“In [8], the authors address the problem of re-identification, where the image of a person or an object captured by mobile cameras is matched to images stored in a database in a central server.”
BackgroundJoint Transmission and Understanding of Semantics with Edge Intelligence
He, Q., Zhang, Y., & Chen, Z. (2023) · International Conference on Communications · doi:10.1109/icc45041.2023.10279072 · Cited by 3
“Joint source-channel coding (JSCC) schemes empowered by DL have been proposed to obtain higher transmission effi-ciency and robustness on the time-varying channel, where the semantic features are extracted, transmitted, and used to retrieve the original multimedia data [5-6].”
BackgroundMethodAnalog versus Digital Pulse Amplitude Modulation for Goal-Oriented Wireless Communications
Binucci, F., Banelli, P., Lorenzo, P. D., & Barbarossa, S. (2023) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco58844.2023.10289715 · Cited by 3
“Thus, differently from DJSCC approaches in [14], [15], [17], and references therein, we focus on the optimal management of computation and transmission resources for analog -PAM GOC systems, with the aim of striking the best trade-off between energy, latency, and accuracy.”
BackgroundLi, P., Huang, J., Zhang, S., & Qi, C. (2024) · International Conference on Multimedia Retrieval · doi:10.1145/3652583.3657612 · Cited by 4
“Edge intelligence and multimedia retrieval[4, 5, 14] have been widely applied in areas such as smart surveillance, smart homes, and intelligent transportation.”
MethodFang, J., Mota, J. F. C., Lu, B., Zhang, W., & Hong, X. (2024) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2024.3411692 · Cited by 3
“For example, [21] designed retrieval-oriented image compression schemes, [19] used channel feedback to improve the quality of transmission, and [18] considered an adaptive bandwidth to transmit information progressively under an erasure channel.”
BackgroundMethodGoal-Oriented Communications for Distributed Sensing: A Joint Scheduling and Estimation Approach
Costa, M. F. D., Elayoubi, S. E., & Hajji, W. (2024) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2024-spring62846.2024.10683003 · Cited by 3
“Another context of application is the joint optimization of the network and application, which has been studied for image retrieval [6] or vehicle platooning controller design under channel imperfections [7].”
BackgroundQML-IB: Quantized Collaborative Intelligence between Multiple Devices and the Mobile Network
Peng, J., Ren, B., Yang, L., Peng, C., Niu, P., & Wu, H. (2024) · IEEE International Symposium on Information Theory (ISIT) · doi:10.1109/isit57864.2024.10619157 · Cited by 3
“A hot viewpoint recently is that only the task-related data should be transmitted in future mobile networks [7]–[9].”
BackgroundTourism Promotion Mechanism Based on Virtual Reality Technology for Real-life Interactive Experience
Nie, X., & Pu, J. (2025) · Informatica · doi:10.31449/inf.v49i7.6274 · Cited by 3
BER-Aware Dynamic Resource Management for Edge-Assisted Goal-Oriented Communications
Binucci, F., & Banelli, P. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10096653 · Cited by 3
“Recent works on GOCs highlighted the great potentials of system designs that depart from classical Shannon-based communications paradigms, exploiting direct analogue modulation, JSCC, or semantic approaches [4, 5, 6, 7, 8].”
BackgroundDeep Joint Source-Channel Coding for Person Re-Identification With Related Image Sources
Chu, Z., Chen, W. R., Wang, N., & Ai, B. (2023) · IEEE International Conference on Communication Technology (ICCT) · doi:10.1109/icct59356.2023.10419799 · Cited by 2
“Instead of training different JSCC encoders as in the literature [12], [13], our proposed network uses the completely same joint encoder and a low-complexity decoder network.”
BackgroundMethodMulti-Stage Semantic Communication for Low-Latency Edge Inference
Ding, G., Gao, H., Liu, S., & Yu, G. (2025) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2025.3558066 · Cited by 2
“In the analog transmission scheme [6], [12], [15], the semantic features z are directly transmitted through the physical channel without being quantized.”
InfluentialRemote DoA Estimation via Subspace-Oriented Deep-Learning-Aided Vector Quantization
Zohar, R., Ginzach, S., & Shlezinger, N. (2025) · IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC) · doi:10.1109/spawc66079.2025.11143308 · Cited by 2
“This is achieved by designing deep neural network (DNN)-based encoder-decoder architectures, while constraining the latent features to a fixed bit representation via uniform quantization [11], [12], scalar quantization [13], [14], and vector quantization [15], [16].”
Synesthesia of Machines (SoM)-Based Task-Driven MIMO System for Image Transmission
Li, S., Zhang, R., Cheng, X., & Tang, J. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3606237 · Cited by 2
“As for downstream perceptual tasks, [21], [22], [24] considered simple image classification and retrieval tasks, while [25] addressed the semantic segmentation task.”
Shi, Y., Yu, L., & Zhu, Q. (2025) · Journal on Wireless Communications and Networking · doi:10.1186/s13638-025-02547-9 · Cited by 3
“” The main idea is to front-load the semantic analysis module to the transmitter, and leverage deep neural networks for semantic compression, thereby reducing the transmission of redundant information and breaking through the band-width constraints imposed by Shannon’s limit [8].”
TAHENNI, A., Ouameur, M. A., Bagaa, M., Massicotte, D., salmi, S., Figueiredo, F. A. P. D., & Ksentini, A. (2026) · Computer Networks · doi:10.1016/j.comnet.2026.112143 · Cited by 2
When Semantic Communication Meets Queueing: Cross-Layer Latency and Task Fidelity Optimization
Sagduyu, Y. E., & Erpek, T. (2026) · IEEE International Symposium on Dynamic Spectrum Access Networks · doi:10.1109/dyspan69846.2026.11571100 · Cited by 2
Deep Semantic Coding for Wireless Image Retrieval
Wang, Y., & Qi, C. (2022) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2022-fall57202.2022.10012945 · Cited by 2
“Inspired by this work, a retrieval-oriented image compression scheme is proposed, which is the first work to study image retrieval over wireless channel [9].”
BackgroundMethodInfluentialSemantic Communications for Joint Image Recovery and Classification
Lyu, Z., Zhu, G., Xu, J., Ai, B., & Cui, S. R. (2023) · IEEE Globecom Workshops (GC Wkshps) · doi:10.1109/gcwkshps58843.2023.10464706 · Cited by 4
“In short, the proposed deep JSCC frameworks in existing works [4]–[6], [8], [9] were dedicatedly designed for handling only one task.”
BackgroundMethodInfluentialMulti-Task Generalizable Communication: Beyond the Information Bottleneck
Leung, C. T., Ghosh, R., & Motani, M. (2024) · International Conference on Communications · doi:10.1109/icc51166.2024.10622203 · Cited by 2
“Many works have applied machine learning (ML) algorithms to fulfill different tasks such as classification [12], image retrieval [7], and natural language processing [13].”
BackgroundDynamic Feature Transmission for Image Communication via Deep Joint Source-Channel Coding
Liu, Y., Qin, Z., Wang, S., & Wang, C. (2024) · International Conference on Communication and Network Security (ICCNS) · doi:10.1145/3711618.3711644 · Cited by 2
“Jankowski, Gündüz, and Mikolajczyk [12]developed a digital and analog DeepJSCC to directly perform human/vehicle recognition tasks.”
Wireless Semantic Communication Based on Probability Distribution: An Initial Work
Luo, Q., Guo, Y., Zheng, A., Ni, Z., Yu, F., & Leung, V. C. M. (2025) · IEEE Wireless Communications and Networking Conference · doi:10.1109/wcnc61545.2025.10978695 · Cited by 2
Metadata
- Authors
- Jankowski, M., Gunduz, D., & Mikolajczyk, K.
- Affiliations
- Imperial College London (GB)
- Venue
- IEEE Journal on Selected Areas in Communications
- Published
- 2021-01
- Volume
- 39
- Issue
- 1
- Pages
- 89-100
- ISSN
- 0733-8716 (ISSN-L)
- 1558-0008
- OpenAlex
- W3044258868
- Cited by
- 3
- References
- 44
- Citations by year
- 2025: 2
- FWCI
- 0.10
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE Journal on Selected Areas in Communications (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Spiral (Imperial College London) (submittedVersion)
- arXiv (Cornell University)
- Funders
- European Research Council (ERC) through Starting Grant BEACON (677854)
- U.K. Engineering and Physical Sciences Research Council (EP/N007743/1, EP/S032398/1, EP/T023600/1)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Data Compression Techniques
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Retrieval and Classification Techniques
- Keywords
- edge computing, joint source-channel coding, fading channels, delay-constrained communication
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.