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
Beyond Transmitting Bits: Context, Semantics, and Task-Oriented Communications
Gunduz, D., Qin, Z., Aguerri, I. E., Dhillon, H. S., Yang, Z., Yener, A., Wong, K. K., & Chae, C.-B. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3223408 · Cited by 776
“A semantic communication system for image retrieval over a wireless channel is considered in [189], which aims to identify the top-k similar images to a query image in a large dataset of images.”
InfluentialEdge artificial intelligence for 6g: vision, enabling technologies, and applications
Letaief, K. B., Shi, Y., Lu, J., & Lu, J. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3126076 · Cited by 829
“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.”
MethodXu, W., Yang, Z., Ng, D. W. K., Levorato, M., Eldar, Y. C., & Debbah, M. (2023) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2023.3239189 · Cited by 660
“For these computing tasks, it has been shown in [222], [223], [224], [225], [226] that the transmission rate can be further reduced below the entropy of the source and joint source-channel coding can perform better than a conventional separate coding design in terms of computing accuracy and speed.”
Semantic Communication Systems for Speech Transmission
Weng, Z., & Qin, Z. (2021) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3087240 · Cited by 663
“Inspired by this, different types of sources have been considered in recent investigation on E2E semantic communication systems, which mainly focus on the image and text transmission [23]–[31].”
BackgroundMethodDistributed Learning in Wireless Networks: Recent Progress and Future Challenges
Chen, M., Gunduz, D., Huang, K., Saad, W., Bennis, M., Feljan, A. V., & Poor, H. V. (2021) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3118346 · Cited by 548
Task-Oriented Multi-User Semantic Communications
Xie, H., Qin, Z., Tao, X., & Letaief, K. B. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3191326 · Cited by 458
“[16] considered image based re-identification for person or cars as the communication task, in which two schemes (digital and analog) are proposed to improve the retrieval accuracy.”
BackgroundNonlinear Transform Source-Channel Coding for Semantic Communications
Dai, J., Wang, S., Tan, K., Si, Z., Qin, X., Niu, K., & Zhang, P. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3180802 · Cited by 364
Learning Task-Oriented Communication for Edge Inference: An Information Bottleneck Approach
Shao, J., Mao, Y., & Zhang, J. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3126087 · Cited by 338
Deep Learning Enabled Semantic Communications With Speech Recognition and Synthesis
Weng, Z., Qin, Z., Tao, X., Pan, C., Liu, G., & Li, G. Y. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3240969 · Cited by 286
“has been investigated in [19], which imposes IoT devices to compress images to the low computational complexity and to reduce the required transmission bandwidth.”
BackgroundWireless Deep Video Semantic Transmission
Wang, S., Dai, J., Liang, Z., Niu, K., Si, Z., Dong, C., Qin, X., & Zhang, P. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3221977 · Cited by 263
“As one modern version, recent deep learning methods for realizing JSCC have stimulated significant interest in both artificial intelligence (AI) and wireless communication communities [15]–[19].”
BackgroundMethodWhat is Semantic Communication? A View on Conveying Meaning in the Era of Machine Intelligence
Lan, Q., Wen, D., Zhang, Z., Zeng, Q., Chen, X., Popovski, P., & Huang, K. (2021) · Journal of Communications and Information Networks · doi:10.23919/jcin.2021.9663101 · Cited by 255
Robust Semantic Communications With Masked VQ-VAE Enabled Codebook
Hu, Q., Zhang, G., Qin, Z., Cai, Y., Yu, G., & Li, G. Y. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3265201 · Cited by 282
“The existing works on semantic communications can be divided into two categories: data reconstruction [10], [11], [12], [13], [14], [15], [16] and task execution [17], [18], [19], [20], [21].”
BackgroundTask-Oriented Multi-User Semantic Communications for VQA
Xie, H., Qin, Z., & Li, G. Y. (2022) · IEEE Wireless Communications Letters · doi:10.1109/lwc.2021.3136045 · Cited by 243
“There exist works that only focus on the single-modal data, i.e., image classification [8] and image retrieval [9].”
BackgroundA Unified Multi-Task Semantic Communication System for Multimodal Data
Zhang, G., Hu, Q., Qin, Z., Cai, Y., Yu, G., & Tao, X. (2024) · IEEE Transactions on Communications · doi:10.1109/tcomm.2024.3364990 · Cited by 219
“For the task execution applications, only the task-specific semantic information is extracted and encoded at the transmitter [21]–[25].”
BackgroundInfluentialZhu, G., Lyu, Z., Jiao, X., Liu, P., Chen, M., Xu, J., Cui, S., & Zhang, P. (2023) · Science China Information Sciences · doi:10.1007/s11432-022-3652-2 · Cited by 239
“For example, for classification tasks, the authors in [86] and [87] proposed a retrieval-oriented wireless image transmission framework to maximize the classification accuracy, where the JSCC framework is trained by the cross-entropy between the predictions and the ground-truth labels.”
BackgroundWITT: A Wireless Image Transmission Transformer for Semantic Communications
Yang, K., Wang, S., Dai, J., Tan, K., Niu, K., & Zhang, P. (2023) · ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp49357.2023.10094735 · Cited by 218
“In particular, for image transmission tasks, deep JSCC [3] and its variants [4,8–10] have competitive performance and much lower complexity compared to advanced image codec (JPEG/JPEG2000/BPG) followed by capacity-approaching channel code family (such as low-density parity-check (LDPC) coding [11]).”
BackgroundGenerative Joint Source-Channel Coding for Semantic Image Transmission
Erdemir, E., Tung, T.-Y., Dragotti, P. L., & Gündüz, D. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3288243 · Cited by 190
“Instead, the receiver may be interested in some downstream task, such as classification [12], [13], or retrieval [14], in which case it would be sufficient to only convey the relevant features of the source signal for the prescribed task.”
BackgroundTask-Oriented Communication for Multidevice Cooperative Edge Inference
Shao, J., Mao, Y., & Zhang, J. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3191118 · Cited by 209
Bandwidth-Agile Image Transmission With Deep Joint Source-Channel Coding
Kurka, D. B., & Gunduz, D. (2021) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2021.3090048 · Cited by 188
Semantics-Empowered Communications: A Tutorial-Cum-Survey
Lu, Z., Li, R., Lu, K., Chen, X., Hossain, E., Zhao, Z., & Zhang, H. (2023) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2023.3333342 · Cited by 213
“For the single-modal EL-SemCom frameworks, Jankowski et al. [145], [213] study the image re-identification task for edge devices, by remotely retrieving similar images from an edge server, such as persons or cars captured by other cameras.”
MethodAdaptable Semantic Compression and Resource Allocation for Task-Oriented Communications
Liu, C., Guo, C., Yang, Y., & Jiang, N. (2023) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2023.3346481 · Cited by 192
“Recently, deep learning (DL) has emerged as a popular solution for signal processing tasks and semantic communications due to its powerful feature extraction capability [13], [14], [15], [16], [17], [18], [19], [20], [21], [22].”
BackgroundDeep Joint Source-Channel Coding for Semantic Communications
Xu, J., Tung, T.-Y., Ai, B., Chen, W., Sun, Y., & Gündüz, D. (2023) · IEEE Communications Magazine · doi:10.1109/mcom.004.2200819 · Cited by 207
“The results in [13] show that DeepJSCC can outperform retrieval-oriented compression followed by capacity-achieving channel codes, even though extremely short block lengths are considered.”
BackgroundLyu, Z., Zhu, G., Xu, J., Ai, B., & Cui, S. R. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3349330 · Cited by 165
“For example, the authors in [19] proposed an image classification oriented wireless transmission framework for wireless person re-identification based on the cross-entropy loss.”
BackgroundMethodNext Generation Advanced Transceiver Technologies for 6G and Beyond
You, C., Cai, Y., Liu, Y., Renzo, M. D., Duman, T. M., Yener, A., & Swindlehurst, A. L. (2025) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2025.3536557 · Cited by 202
“Numerous studies have emerged in various domains to implement this approach for significantly reducing transmission overhead [245]–[248], [250], [274].”
A Perspective on Time Toward Wireless 6G
Popovski, P., Chiariotti, F., Huang, K., Kalør, A. E., Kountouris, M., Παππάς, Ν., & Soret, B. (2022) · Proceedings of the IEEE · doi:10.1109/jproc.2022.3190205 · Cited by 171
Robust Information Bottleneck for Task-Oriented Communication With Digital Modulation
Xie, S., Ma, S., Ding, M., Shi, Y., Tang, M., & Wu, Y. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3288252 · Cited by 170
“task [23] by transmitting the feature vectors compressed by the DNN-based joint source-channel coding scheme.”
BackgroundMethodTung, T.-Y., Kobus, S., Roig, J. S. P., & Gündüz, D. (2021) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3087248 · Cited by 142
Robust Semantic Communications Against Semantic Noise
Hu, Q., Zhang, G., Qin, Z., Cai, Y., Yu, G., & Li, G. Y. (2022) · 2022 IEEE 96th Vehicular Technology Conference (VTC2022-Fall) · doi:10.1109/vtc2022-fall57202.2022.10012843 · Cited by 97
“The existing works on semantic communications can be categorized into: (i) data reconstruction, where the global semantic information behind data is extracted and the data is reconstructed based on the received semantic information [3]– [5]; (ii) task execution, where only the task-related semantic information is enco…”
BackgroundQoE-Aware Resource Allocation for Semantic Communication Networks
Yan, L., Qin, Z., Zhang, R., Li, Y., & Li, G. Y. (2022) · Globecom · doi:10.1109/globecom48099.2022.10001594 · Cited by 114
“, text transmission [2] and image retrieval [3], and multimodal tasks, e.”
BackgroundYang, Y., Guo, C., Liu, F., Sun, L., Liu, C., & Sun, Q. (2022) · IEEE Industrial Electronics Magazine · doi:10.1109/mie.2022.3174331 · Cited by 106
“6 IEEE INDUSTRIAL ELECTRONICS MAGAZINE ■ SEPTEMBER 2023 [11] compressed transmitted information by reducing the Shannon entropy of the extracted representations.”
BackgroundMeng, S., Wu, S., Zhang, J., Cheng, J., Zhou, H., & Zhang, Q. (2024) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2024.3416309 · Cited by 110
“The authors in [202] consider image retrieval issue at the edge, where one edge device captures and transmits images to the edge server, and the server conducts retrieval task to predict the identification of the images.”
BackgroundMultimodal semantic communication accelerated bidirectional caching for 6G MEC
Wang, C., Yu, X., Xu, L., Wang, Z., & Wang, W. (2022) · Future Generation Computer Systems · doi:10.1016/j.future.2022.10.036 · Cited by 94
Semantic Communication With Memory
Xie, H., Qin, Z., & Li, G. Y. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3288236 · Cited by 96
“[7] have developed digital and analog deep joint source-channel coding (JSCC) to perform the person/car re-identification task directly, which improves the image retrieval accuracy effec-”
MethodChannel-Adaptive Wireless Image Transmission With OFDM
Wu, H., Shao, Y., Mikolajczyk, K., & Gündüz, D. (2022) · IEEE Wireless Communications Letters · doi:10.1109/lwc.2022.3204837 · Cited by 83
“However, an increasing number of wireless applications, such as Internet-of-Things and edge intelligence [2], [3], [4], require the efficient transmission of large volumes of data under strict delay constraints, resulting in an increasing interest in joint source channel coding (JSCC) in recent years.”
BackgroundSelf-citationJoint Source–Channel Coding: Fundamentals and Recent Progress in Practical Designs
Gündüz, D., Wigger, M., Tung, T.-Y., Zhang, P., & Xiao, Y. L. (2024) · Proceedings of the IEEE · doi:10.1109/jproc.2024.3477331 · Cited by 126
“…underlying communication channel is noisy, and we want to communicate under extreme bandwidth and/or power limitations, semantic and goal-oriented communications will significantly benefit from JSCC approaches to push the limits of communications for relevant applications [11], [15], [18], [31].”
Privacy-Preserving Task-Oriented Semantic Communications Against Model Inversion Attacks
Wang, Y., Guo, S., Deng, Y., Zhang, H., & Fang, Y. M. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3369170 · Cited by 89
“This new paradigm has received intensive attention lately, leading to the implementations of wireless image retrieval at the edge [11], [12], one-device edge inference [13], [14], and multi-device cooperative edge inference [7].”
BackgroundIntegrated Sensing-Communication-Computation for Over-the-Air Edge AI Inference
Zhuang, Z., Wen, D., Shi, Y., Zhu, G., Wu, S., & Niyato, D. T. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3306465 · Cited by 83
“A joint source and channel coding (JSCC) approach was developed in [25] to map feature vectors into channel symbols.”
MethodDeepMA: End-to-End Deep Multiple Access for Wireless Image Transmission in Semantic Communication
Zhang, W., Bai, K., Zeadally, S., Zhang, H., Shao, H., Ma, H., & Leung, V. C. M. (2023) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2023.3326302 · Cited by 96
“These tasks include answering visual questions [15], image retrieval [16], speech recognition [6], and video classi-fication [17].”
BackgroundProgressive Feature Transmission for Split Classification at the Wireless Edge
Lan, Q., Zeng, Q., Popovski, P., Gündüz, D., & Huang, K. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3221778 · Cited by 81
“On the other hand, the joint sourceand-channel coding schemes addressing fading channels are applicable to analog transmission only [9], [10].”
BackgroundMethodInfluentialDai, J., Wang, S., Yang, K., Tan, K., Qin, X., Si, Z., Niu, K., & Zhang, P. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3288246 · Cited by 77
“realize JSCC [14], [15], [16], [17], [18], [19], [20], [21], which can be optimized for specific end-to-end transmission objectives.”
BackgroundA Unified Multi-Task Semantic Communication System with Domain Adaptation
Zhang, G., Hu, Q., Qin, Z., Cai, Y., & Yu, G. (2022) · Globecom · doi:10.1109/globecom48099.2022.10000850 · Cited by 75
“For the task execution applications, only the task-specific semantic information is extracted and encoded at the transmitter [8]–[12].”
BackgroundInfluentialTransformer-Aided Wireless Image Transmission With Channel Feedback
Wu, H., Shao, Y., Özfatura, E., Mikolajczyk, K., & Gündüz, D. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3386052 · Cited by 72
“…Perceptual quality evaluation We can observe that the reconstruction performance is significantly improved after each interaction, To conduct a more comprehensive assessment of the model’s perceptual qualities, we extend our evaluation with LPIPS measurements applied to both the Cifar10 and CelebA…”
BackgroundMethodSelf-citationMachine Learning and Wireless Communications
(2022) · Cambridge University Press eBooks · doi:10.1017/9781108966559 · Cited by 68
Collaborative Semantic Communication for Edge Inference
Lo, W. F., Mital, N., Wu, H., & Gündüz, D. (2023) · IEEE Wireless Communications Letters · doi:10.1109/lwc.2023.3256006 · Cited by 63
“edge inference, we first consider an orthogonal multiple access (OMA) scheme employing time division multiple access (TDMA) with distributed JSCC, and show that it outperforms the schemes in [5], as well as a conventional separate”
BackgroundMethodInfluentialGenerative Model based Highly Efficient Semantic Communication Approach for Image Transmission
Han, T., Tang, J., Yang, Q., Duan, Y., Zhang, Z., & Shi, Z. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10096372 · Cited by 62
“A task-specific semantic communication system was proposed in [3] for image transmission targeting person re-identification, which achieves a compression ratio of 1/128.”
BackgroundQoE-Based Semantic-Aware Resource Allocation for Multi-Task Networks
Yan, L., Qin, Z., Li, C., Zhang, R., Li, Y., & Tao, X. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3386807 · Cited by 62
“In addition, intelligent tasks to be executed at the receiver have been paid attention as well, such as image retrieval [14], image classification [15], and speech recognition [16].”
BackgroundCollaborative Inference for AI-Empowered IoT Devices
Shlezinger, N., & Bajić, I. V. (2022) · IEEE Internet of Things Magazine · doi:10.1109/iotm.001.2200152 · Cited by 60
“For the latter one, there has been recent work on decoder-side error mitigation [7], as well as joint source-channel coding [8].”
BackgroundTask-Oriented Communication for Edge Video Analytics
Shao, J., Zhang, X., & Zhang, J. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3314888 · Cited by 72
“The same idea was utilized in other intelligence edge applications including image retrieval [34], point cloud processing [35], and semantic communication [27].”
MethodFeature Importance-Aware Task-Oriented Semantic Transmission and Optimization
Wang, Y., Han, S., Xu, X., Liang, H., Meng, R., Dong, C., & Zhang, P. (2024) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2024.3375496 · Cited by 73
“Therefore, researches [17], [18], [19], [20] further optimize the semantic communication system from the perspective of intelligent tasks while considering the requirements of both transmission and task performance.”
BackgroundWireless Transmission of Images with the Assistance of Multi-level Semantic Information
Zhang, Z., Yang, Q., He, S., Sun, M., & Chen, J. (2022) · International Symposium on Wireless Communication Systems · doi:10.1109/iswcs56560.2022.9940401 · Cited by 58
“[17] present a joint feature compression and transmission system to deal with the limited computational resources at the edge server, named DeepJSCC, which not only improves end-to-end reliability but also reduces computation complexity.”
BackgroundMulti-User Semantic Communications for Cooperative Object Identification
Zhang, Y., Xu, W., Gao, H., & Wang, F. (2022) · IEEE International Conference on Communications workshops · doi:10.1109/iccworkshops53468.2022.9814491 · Cited by 55
Distributed satellite information networks: architecture, enabling technologies, and trends
Zhang, Q., Xu, L., Huang, J., Yang, T., Jiao, J., Wang, Y., Shi, Y., Zhang, C., Zhang, X., Zhang, K., et al. (2025) · Science China Information Sciences · doi:10.1007/s11432-024-4408-1 · Cited by 66
Semantic Importance-Aware Communications With Semantic Correction Using Large Language Models
Guo, S., Wang, Y., Ye, J., Zhang, A., Zhang, P., & Xu, K. (2025) · IEEE Transactions on Machine Learning in Communications and Networking · doi:10.1109/tmlcn.2025.3530875 · Cited by 64
“This approach aids in tasks such as retrieval, classification [9], and reconstruction [10], , as exemplified in studies focused on image reconstruction [11] and image retrieval [12], which demonstrate enhanced performance compared to traditional communication systems.”
Semantic Communications With Variable-Length Coding for Extended Reality
Zhang, B., Qin, Z., & Li, G. Y. (2023) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2023.3300509 · Cited by 58
“Some representative works consider wireless image retrieval [13], [9], machine translation [14], visual question answering [15] and so on.”
BackgroundSemantic Communications Based on Adaptive Generative Models and Information Bottleneck
Barbarossa, S., Comminiello, D., Grassucci, E., Pezone, F., Sardellitti, S., & Lorenzo, P. D. (2023) · IEEE Communications Magazine · doi:10.1109/mcom.005.2200829 · Cited by 67
Islam, N., & Shin, S. (2024) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2024.3425314 · Cited by 64
“For instance, the image classification-oriented SemCom system described in [96] enhances recognition accuracy; similarly, the system in [97] aimed at image reidentification of a person improves retrieval accuracy.”
BackgroundInnovative adaptive edge detection for noisy images using wavelet and Gaussian method
Li, H., & Xu, K. (2025) · Scientific Reports · doi:10.1038/s41598-025-86860-9 · Cited by 48
Toward Natively Intelligent Semantic Communications and Networking
Trevlakis, S. E., Παππάς, Ν., & Boulogeorgos, A. A. (2024) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2024.3371871 · Cited by 50
“Task execution [16], [60], [61] and data reconstruction [62], [63], [64], [65] are the two primary applications of the current efforts on SemCom.”
BackgroundReasoning Over the Air: A Reasoning- Based Implicit Semantic-Aware Communication Framework
Xiao, Y. L., Liao, Y., Li, Y., Shi, G., Poor, H. V., Saad, W., Debbah, M. A., & Bennis, M. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3312115 · Cited by 45
“In [14], the authors investigated the image retrieval problem in a wireless edge network with power and bandwidth constraints.”
BackgroundTimely and Massive Communication in 6G: Pragmatics, Learning, and Inference
Gündüz, D., Chiariotti, F., Huang, K., Kalør, A. E., Kobus, S., & Popovski, P. (2023) · IEEE BITS the Information Theory Magazine · doi:10.1109/mbits.2023.3322667 · Cited by 55
Guo, L., Chen, W., Sun, Y., & Ai, B. (2024) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2024.3510586 · Cited by 45
“The studies discussed above consider analog communication schemes, where the feature vectors are directly mapped into analog symbols for transmission without being converted to bits at any stage [10].”
On privacy, security, and trustworthiness in distributed wireless large AI models
Yang, Z., Xu, W., Liang, L., Cui, Y., Qin, Z., & Debbah, M. A. (2025) · Science China Information Sciences · doi:10.1007/s11432-024-4465-3 · Cited by 45
“These learning strategies, including end-to-end learning [32] and multi-stage learning [33], ensure effective semantic alignment for specific communication tasks.”
Over-the-Air Split Machine Learning in Wireless MIMO Networks
Yang, Y., Zhang, Z., Tian, Y., Yang, Z., Huang, C., Zhong, C., & Wong, K. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3242701 · Cited by 49
Adaptive Semantic-Bit Communication for Extended Reality Interactions
Wang, C., Li, Y., Gao, F., Deng, D., Xu, J., Liu, Y., & Wang, W. (2023) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2023.3310654 · Cited by 34
“…transmittingtheencodeddata,andthendecodingandrestoring sourcedataaccordingtotheencodeddata,semanticcommunica-tionsignificantlyreducesthedatatraffic,whichdemonstratesthe greatpotentialtoaddresswirelesscommunicationbottlenecks inXR[26],[27],[28],[29],[30],[31].”
Over-the-Air Ensemble Inference with Model Privacy
Yilmaz, S. F., Hasırcıoglu, B., & Gündüz, D. (2022) · IEEE International Symposium on Information Theory (ISIT) · doi:10.1109/isit50566.2022.9834591 · Cited by 33
Xu, X., Xu, B., Han, S., Dong, C., Xiong, H., Meng, R., & Zhang, P. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3305011 · Cited by 32
“In particular, a variety of intelligent tasks, such as image classification task [16], image retrieval task [17], visual question answering (VQA) task [18], and so on, are implemented at the receiver.”
BackgroundMethodToward Communication-Efficient Digital Twin via AI-Powered Transmission and Reconstruction
Li, M., Chen, C., Yang, X., Zhou, J. T., Zhang, T., & Li, Y. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3310089 · Cited by 33
“These approaches suffer from the cliff effect and only achieve desired performance with accurate target channel quality estimation, making them unsuitable for low-latency digital twin applications [13].”
BackgroundMethodInfluentialFeatures Disentangled Semantic Broadcast Communication Networks
Ma, S., Zhang, Z., Wu, Y., Li, H., Shi, G., Gao, D., Shi, Y., Li, S., & Al‐Dhahir, N. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3334225 · Cited by 33
“Specifically, semantic communications have attracted intensive research efforts in text [10], [13], [15], [16], [17], [18], speech/audio [19], [20], [21], video [22], [23], and image transmission [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37].”
MethodVideoQA-SC: Adaptive Semantic Communication for Video Question Answering
Guo, J., Chen, W. R., Sun, Y., Xu, J., & Ai, B. (2025) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2025.3559160 · Cited by 31
“W ITH the development of artificial intelligence (AI) technology, many edge machines deploy AI models to process information for intelligent tasks [1], [2], which support AI-empowered applications such as remote healthcare, autonomous driving, and the Internet of Things (IoT).”
Multi-Device Task-Oriented Communication via Maximal Coding Rate Reduction
Cai, C., Yuan, X., & Zhang, Y. A. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3461336 · Cited by 33
“For image retrieval, the authors in [10], [11] proposed a deep joint source and channel coding (JSCC) scheme based on the autoencoder architecture, where the wireless channel is incorporated in the autoencoder as a non-trainable layer.”
Decentralized Low-Latency Collaborative Inference via Ensembles on the Edge
Malka, M., Farhan, E., Morgenstern, H., & Shlezinger, N. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3497167 · Cited by 37
Zhang, M., Abdi, M., Dasari, V. R., & Restuccia, F. (2025) · Computer Networks · doi:10.1016/j.comnet.2025.111531 · Cited by 45
Semantic Data Sourcing for 6G Edge Intelligence
Huang, K., Lan, Q., Liu, Z., & Yang, L. (2023) · IEEE Communications Magazine · doi:10.1109/mcom.001.2200962 · Cited by 33
“To overcome bottleneck, we propose in this article a novel framework of SEMantic DAta Sourcing (SEMDAS) locating semantically-matched data sources to 5G’s key innovation lies in providing an infrastructure requirements in rate, reliability, and latency, three types of connectivity have been defi…”
MethodJoint URLLC Traffic Scheduling and Resource Allocation for Semantic Communication Systems
Ding, G., Liu, S., Yuan, J., & Yu, G. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3339239 · Cited by 29
“In addition to the data reconstruction task, the semantic communication systems for more general tasks have also been investigated [13], [14], [15], [16].”
MethodResource-Efficient Ubiquitous Sensor Networks for Smart Agriculture: A Survey
Arif, M., Maya, J. A., Anandan, N., Pérez, D. A., Tonello, A. M., Zangl, H., & Rinner, B. (2024) · IEEE Access · doi:10.1109/access.2024.3516814 · Cited by 34
Synchronous Multi-Modal Semantic Communication System With Packet-Level Coding
Tian, Y., Ying, J., Qin, Z., Jin, Y., & Tao, X. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3534995 · Cited by 29
“For single-modal sources or tasks such as text [4], [5], [6], speech [7], image [8], [9], [10], and video [11], semantic communication systems extract and transmit tailored semantic features related to tasks at the receiver side.”
Information Compression in the AI Era: Recent Advances and Future Challenges
Chen, J., Fang, Y., Khisti, A. J., Özgür, A., & Shlezinger, N. (2025) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2025.3560359 · Cited by 42
“A candidate approach to do so is to model its operation as additive noise during training [85], [86], [87].”
Goal-Oriented Communications for the IoT: System Design and Adaptive Resource Optimization
Lorenzo, P. D., Merluzzi, M., Binucci, F., Battiloro, C., Banelli, P., Strinati, E. C., & Barbarossa, S. (2023) · IEEE Internet of Things Magazine · doi:10.1109/iotm.001.2300163 · Cited by 29
“Differently from several other papers on GO and semantic communications, which mainly focus on the design of proper data compression/coding schemes, e.g., [3]–[5], [7], [9]–[11], our paper looks at the problem from a wider and cross-layer perspective.”
BackgroundA Multi-Task Semantic Communication System for Natural Language Processing
Sheng, Y., Li, F., Liang, L., & Jin, S. (2022) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2022-fall57202.2022.10013075 · Cited by 31
“Furthermore, all the systems above focus on the single-modal data, such as image [11], while the multimodal data is usually required in practical communication scenarios.”
BackgroundsDAC—Semantic Digital Analog Converter for Semantic Communications
Bao, Z., Liang, H., Liu, X., Li, C., Dong, C., Xu, X., Guo, C., Chen, H., & Zhang, P. (2025) · IEEE Transactions on Communications · doi:10.1109/tcomm.2025.3566976 · Cited by 26
“Works in [17], [18], and [19] study the multi-task architecture to further improve semantic communication efficiency.”
Deep Joint Source-Channel Coding for Wireless Image Transmission with Semantic Importance
Sun, Q., Guo, C., Yang, Y., Chen, J., Tang, R., & Liu, C. (2022) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2022-fall57202.2022.10012892 · Cited by 28
“[4] extracted semantic information for image retrieval tasks.”
BackgroundFederated KD-Assisted Image Semantic Communication in IoT Edge Learning
Xu, X., Xu, Y., Dou, H., Chen, M., & Wang, L. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3429391 · Cited by 30
“The device-side image semantic feature extractor is Shufflenet v2, the JSCC encoder and decoder reference the design in [29], as shown in Fig.”
MethodDeep Semantic Communication for Knowledge Sharing in Internet of Vehicles
Wang, Z., Leng, S., Zhang, H., & Yuen, C. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3531910 · Cited by 24
“With the advancement of deep learning, deep semantic communication has been widely studied in text [6], [7], image [23], [24], and speech transmission systems [25], to enhance transmission robustness against noise and distortions.”
Letafati, M., Behroozi, H., Khalaj, B. H., & Jorswieck, E. A. (2021) · IEEE Transactions on Communications · doi:10.1109/tcomm.2021.3128423 · Cited by 26
Distributed Inference in Resource-Constrained IoT for Real-Time Video Surveillance
Khan, M. A., Hamila, R., Erbad, A., & Gabbouj, M. (2022) · IEEE Systems Journal · doi:10.1109/jsyst.2022.3198711 · Cited by 29
“Although, local computing of DL models may not be efficient due to limited computational resources, edge-based processing may be inevitable in some applications, e.g., in image retrieval problem in which the data resides at the edge server [4].”
BackgroundHow to Evaluate Semantic Communications for Images With ViTScore Metric?
Zhu, T., Peng, B., Liang, J., Han, T., Wan, H., Fu, J., & Chen, J. (2024) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2024.3392803 · Cited by 24
“SC have attracted growing interest from both academia and industry (see [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24] and the references therein).”
Explainable Semantic Communication for Text Tasks
Liu, C., Guo, C., Yang, Y., Ni, W., Zhou, Y., Li, L., & Quek, T. Q. S. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3450902 · Cited by 28
“Focusing on specific mission, task-oriented semantic communication has been proposed in [13]–[17].”
Content-Aware Semantic Communication for Goal-Oriented Wireless Communications
Fu, Y., Cheng, W., & Zhang, W. (2023) · IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS) · doi:10.1109/infocomwkshps57453.2023.10226153 · Cited by 27
“For example, the authors of [9] proposed the Joint Source and Channel Coding (JSCC) based retrieval-oriented image compression scheme, which can be adapted to the image query task for edge server performed over noisy channels.”
BackgroundSemantic communication for cooperative perception based on importance map
Sheng, Y., Ye, H., Liang, L., Jin, S., & Li, G. Y. (2024) · Journal of the Franklin Institute · doi:10.1016/j.jfranklin.2024.106739 · Cited by 31
Tackling Distribution Shifts in Task-Oriented Communication With Information Bottleneck
Li, H., Shao, J., He, H., Song, S., Zhang, J., & Letaief, K. B. (2025) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2025.3559116 · Cited by 27
“…information transmission as a data reconstruction task and utilizing an autoencoder architecture [15], deep learning (DL) has been seamlessly integrated into the communication systems, leading to the development of DL-based joint source-channel coding (DeepJSCC) [6], [16] for image transmission.”
Cross-modal communication technology: A survey
Wei, X., Wu, D., Zhou, L., & Guizani, M. (2023) · Fundamental Research · doi:10.1016/j.fmre.2023.08.002 · Cited by 22
Secure Deep-JSCC Against Multiple Eavesdroppers
Kalkhoran, S. A. A., Letafati, M., Erdemir, E., Khalaj, B. H., Behroozi, H., & Gündüz, D. (2023) · Globecom · doi:10.1109/globecom54140.2023.10436928 · Cited by 22
“Driven by the growing interest in semantic communication systems [1], intelligent transmission of multimedia content has received much attention because of its various applications in augmented/virtual reality (AR/VR), Metaverse [2]–[4], and surveillance systems [5], [6].”
BackgroundTask-Oriented Scene Graph-Based Semantic Communications With Adaptive Channel Coding
Sun, S., Qin, Z., Xie, H., & Tao, X. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3450697 · Cited by 25
Liu, F., Sun, Z., Yang, Y., Guo, C., & Zhao, S. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3350656 · Cited by 21
“To build a semantic communication system for intelligent tasks, Jankowski et al. [19] proposed a semantic communication system oriented to image retrieval task.”
MethodJoint Device Scheduling and Resource Allocation for ISCC-Based Multiview–Multitask Inference
Wang, D., Wen, D., He, Y., Chen, Q., Zhu, G., & Yu, G. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3456569 · Cited by 22
SCSC: A Novel Standards-Compatible Semantic Communication Framework for Image Transmission
Han, X., Wu, Y., Gao, Z., Feng, B., Shi, Y., Gündüz, D., & Zhang, W. (2025) · IEEE Transactions on Communications · doi:10.1109/tcomm.2025.3529221 · Cited by 21
“Jankowski et al. [5] extended DeepJSCC to task-oriented applications.”
AirNet: Neural Network Transmission over the Air
Jankowski, M., Gündüz, D., & Mikolajczyk, K. (2022) · IEEE International Symposium on Information Theory (ISIT) · doi:10.1109/isit50566.2022.9834372 · Cited by 21
Self-citationSemantic Feature Division Multiple Access for Multi-User Digital Interference Networks
Ma, S., Zhang, C., Shen, B., Wu, Y., Li, H., Li, S., Shi, G., & Al‐Dhahir, N. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3427675 · Cited by 19
“• For task-oriented communications, based on a retrieval-oriented deep image compression scheme,the authors in [8] proposed both digital and analog JSCC schemes for wireless image retrieval at the wireless edge.”
BackgroundMethodAdaptive resource optimization for edge inference with goal-oriented communications
Binucci, F., Banelli, P., Lorenzo, P. D., & Barbarossa, S. (2022) · EURASIP Journal on Advances in Signal Processing · doi:10.1186/s13634-022-00958-0 · Cited by 24
“The single DNN architecture that these GOC schemes have to train is (typically) very complex because it has to work well for a variety of state channels, noise levels, and required task performance, which the GOC may have to face.”
BackgroundMethodInfluentialOver-the-Air Edge Inference via End-to-End Metasurfaces-Integrated Artificial Neural Networks
Stylianopoulos, K., Lorenzo, P. D., & Alexandropoulos, G. C. (2026) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2026.3674047 · Cited by 28
“Some of its layers reside at the TX, the rest are collocated with the RX. Low-dimensional feature vectors, that are outputs of intermediate TX layers, are transmitted over the channel [6], [7], [8], with the channel introducing potentially unfavorable transformations.”
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.