Deep Learning With Edge Computing: A Review
The DOI matches its Crossref record; 130 of the 141 references it cites were found in a registry or an index.
Abstract
Deep learning is currently widely used in a variety of applications, including computer vision and natural language processing. End devices, such as smartphones and Internet-of-Things sensors, are generating data that need to be analyzed in real time using deep learning or used to train deep learning models. However, deep learning inference and training require substantial computation resources to run quickly. Edge computing, where a fine mesh of compute nodes are placed close to end devices, is a viable way to meet the high computation and low-latency requirements of deep learning on edge devices and also provides additional benefits in terms of privacy, bandwidth efficiency, and scalability. This paper aims to provide a comprehensive review of the current state of the art at the intersection of deep learning and edge computing. Specifically, it will provide an overview of applications where deep learning is used at the network edge, discuss various approaches for quickly executing deep learning inference across a combination of end devices, edge servers, and the cloud, and describe the methods for training deep learning models across multiple edge devices. It will also discuss open challenges in terms of systems performance, network technologies and management, benchmarks, and privacy. The reader will take away the following concepts from this paper: understanding scenarios where deep learning at the network edge can be useful, understanding common techniques for speeding up deep learning inference and performing distributed training on edge devices, and understanding recent trends and opportunities.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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
“As mentioned in Section 2.1, the progress of DL is very fast but when it comes to application in Edge or IoT devices then a huge gap is noticeable [67].”
From A survey on deep transfer learning to edge computing for mitigating the COVID-19 pandemic (2020) · cites this paper as [67]
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
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1506.01497 · Cited by 70,812
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
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
(2016) · Cited by 31,590
Research works printed but not linked: 30
- Jeans, 2019, Related’s Hudson Yards: Smart City or Surveillance City?
- AT&T Multi-Access Edge Computing
- Blog, Edge Computing at Chick-Fil-A
- Goodfellow, 2016, Deep Learning, 1
- 10.1145/3299710.3211336 · DOI · Paper page
- Tensorflow
- Tensorflowlite
- Caffe2
- Caffe
- Pytorch
- Cuda
- Cudnn
- AWS Deeplens
- Deep Learning for Siri’s Voice: On-Device Deep Mixture Density Networks for Hybrid Unit Selection Synthesis
- Google Duplex: An AI System for Accomplishing Real-World Tasks Over the Phone
- Pagespeed Insights: Improve Server Response Time
- 2017, Hey Siri: An On-Device DNN-Powered Voice Trigger for Apple’s Personal Assistant
- Google’s Pixel Buds are no Match for Professional Interpreters
- Jindal, 2018, Enabling Full Body AR With Mask R-CNN2GO
- 10.1145/3081333.3081360 · DOI · Paper page
- Edge TPU.
- VPU.
- NVIDIA EGX Edge Computing Platform
- Qualcomm Neural Processing SDK for AI
- 10.1109/MNET.2018.1700202 · DOI · Paper page
- Arcore Overview
- Teerapittayanon, 2017, BranchyNet: Fast inference via early exiting from deep neural networks, CoRR, abs/1709.01686, 1 · arXiv
- 10.1145/3195970.3196023 · doi:10.1145/3195970.3196023 · Paper page
- Hsieh, Gaia: Geo-distributed machine learning approaching LAN speeds, 629
- 10.1109/NCA.2017.8171350 · DOI · Paper page
Cited by
Items 501–600 of 1,000
Flexible Training and Uploading Strategy for Asynchronous Federated Learning in Dynamic Environments
Wu, M., Boban, M., & Dressler, F. (2024) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2024.3418613 · Cited by 13
“direction to enable fast and ubiquitous computing [1], [2].”
BackgroundAravamuthan, S. R., Walleser, E., & Döpfer, D. (2024) · Preventive Veterinary Medicine · doi:10.1016/j.prevetmed.2024.106300 · Cited by 16
Islam, M. F., Manab, M. A., Mondal, J. J., Zabeen, S., Rahman, F. B., Hasan, M. Z., Sadeque, F., & Noor, J. (2024) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2024.109475 · Cited by 26
Yun, J. Y., Choi, E. J., Chung, M. H., Kim, T. W., & Moon, J. W. (2024) · Building and Environment · doi:10.1016/j.buildenv.2024.112217 · Cited by 12
Inverse design of octagonal plasmonic structure for switching using deep learning
Adibnia, E., Ghadrdan, M., & Mansouri-Birjandi, M. A. (2025) · Results in Physics · doi:10.1016/j.rinp.2025.108197 · Cited by 12
Guo, G., Qamar, F., Kazmi, S. H. A., & Rehman, M. H. U. (2025) · Internet of Things · doi:10.1016/j.iot.2025.101686 · Cited by 12
Wu, S., & Zhang, X. (2022) · Wireless Communications and Mobile Computing · doi:10.1155/2022/4326437 · Cited by 13
An Online Approach for DNN Model Caching and Processor Allocation in Edge Computing
Chen, Z., Zhang, S., Ma, Z., Zhang, S., Qian, Z., Xiao, M., Wu, J., & Lu, S. (2022) · International Workshop on Quality of Service · doi:10.1109/iwqos54832.2022.9812874 · Cited by 15
“(2) DNN inference requests from users often comprise local geographic characteristics, such as similar inference tasks [8], so deploying a portion of deep models which are hotspots on edge servers in different locations can efficiently reduce communication delay and avoid excessive model deployment cost on the edge se…”
BackgroundA New NN-Based Approach to In-Sensor PDM-to-PCM Conversion for Ultra TinyML KWS
Vitolo, P., Liguori, R., Benedetto, L. D., Rubino, A., Pau, D. P., & Licciardo, G. D. (2022) · IEEE Transactions on Circuits & Systems II Express Briefs · doi:10.1109/tcsii.2022.3224022 · Cited by 13
“…has largely improved the voice user interfaces [2], but it has also increased the computational complexity and memory requirements of devices, which often recur to cloud resources with consequent limitations in terms of scalability, privacy, communication latencies and bandwidth availability [3].”
BackgroundParthasarathy, A., & Krishnamachari, B. (2022) · International Telecommunication Networks and Applications Conference (ITNAC) · doi:10.1109/itnac55475.2022.9998427 · Cited by 21
“IoT applications for edge inference range from retail to wearable technology [4], [6].”
BackgroundExploring the Performance and Efficiency of Transformer Models for NLP on Mobile Devices
Panopoulos, I., Nikolaidis, S., Venieris, S. I., & Venieris, I. S. (2023) · IEEE Symposium on Computers and Communications (ISCC) · doi:10.1109/iscc58397.2023.10217850 · Cited by 17
“To achieve this, different execution options have been proposed, including running DNN inference in the cloud, at the edge, and on-device [2].”
MethodJSDoop and TensorFlow.js: Volunteer Distributed Web Browser-Based Neural Network Training
Morell, J. Á., Camero, A., & Alba, E. (2019) · IEEE Access · doi:10.1109/access.2019.2950287 · Cited by 14
“re their privacy (eg., personal photos or voice records) [40]. Therefore, researchers are exploring the possibility of moving the inference and training of NNs to web browsers [41] and mobile devices [42], and analysing the feasibility of deploying deep models through geographically distributed servers and simultaneou…”
BackgroundInfluentialTask Offloading for Automatic Speech Recognition in Edge-Cloud Computing Based Mobile Networks
Cheng, S., Xu, Z., Li, X., Wu, X., Fan, Q., Wang, X., & Leung, V. C. M. (2020) · IEEE Symposium on Computers and Communications (ISCC) · doi:10.1109/iscc50000.2020.9219579 · Cited by 14
“In [10], [11], the authors describe the content related to the ASR program running in mobile devices.”
BackgroundAutoGeoLabel: Automated Label Generation for Geospatial Machine Learning
Albrecht, C. M., Marianno, F. J., & Klein, L. J. (2021) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata52589.2021.9672060 · Cited by 16
“Internet-of-Things (IoT) [27] edge devices [28].”
SLED: A Speculative LLM Decoding Framework for Efficient Edge Serving
Li, X., Spatharakis, D., Ghafouri, S., Fan, J., Vandierendonck, H., John, D., Ji, B., & Nikolopoulos, D. S. (2025) · ACM/IEEE Symposium on Edge Computing (SEC) · doi:10.1145/3769102.3770608 · Cited by 11
“Never-theless, edge environments, characterized by limited memory, processing power, and energy budgets, present formidable obstacles to efficient LLM inference.”
Multi-Model Specifications and their Application to Classification Systems
Burns, A., & Baruah, S. (2023) · International Conference on Real-Time Networks and Systems (RTNS) · doi:10.1145/3575757.3575760 · Cited by 11
Li, W. F., Arya, G., Roques‐Carmes, C., Lin, Z., Johnson, S. G., & Soljacic, M. (2023) · Optics Express · doi:10.1364/oe.492553 · Cited by 13
“The addition of a post processing step does not present a significant hurdle to real-life applications, given the relative simplicity of our reconstruction algorithm (compared to state-of-the-art artificial intelligence tasks implemented on centralized or edge computing platforms [38,39]).”
BackgroundResMap: Exploiting Sparse Residual Feature Map for Accelerating Cross-Edge Video Analytics
Chen, N., Zhang, S., Zhang, S., Yan, Y., Chen, Y., & Lu, S. (2023) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom53939.2023.10228990 · Cited by 12
Neural network methods for radiation detectors and imaging
Lin, S., Ning, S., Zhu, H., Zhou, T., Morris, C. L., Clayton, S. M., Cherukara, M. J., Chen, R. T., & Wang, . (2024) · Frontiers in Physics · doi:10.3389/fphy.2024.1334298 · Cited by 20
A Deployable Edge Computing Solution for Machine Condition Monitoring
Zhao, X., & Wang, P. E. (2024) · IEEE International Instrumentation and Measurement Technology Conference · doi:10.1109/i2mtc60896.2024.10560795 · Cited by 16
“It only provides on-device inference abilities [26], implying that model training and exporting should be conducted off-board.”
BackgroundIoT traffic management using deep learning based on osmotic cloud to edge computing
Absardi, Z. N., & Javidan, R. (2024) · Telecommunication Systems · doi:10.1007/s11235-024-01185-8 · Cited by 11
“Another way to reduce the data size is to find and send only the Region Of Interest (ROI), which is a method to eliminate the additional and unusable parts of the images [38].”
Luo, Q., Luan, T. H., Shi, W., & Fan, P. (2024) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2024.3444815 · Cited by 12
“Good at abstracting many factors that affect the VEC environment to a mapping problem and learning optimal computation offloading and resource allocation strategy from the environment, deep reinforcement learning (DRL) is becoming a candidate solution to the data scheduling problem [10], [11].”
Real‐time augmentation of diagnostic nasal endoscopy video using AI‐enabled edge computing
Bidwell, J., Gyawali, D., Morse, J., Ganeshan, V., Nguyen, T. P., & McCoul, E. D. (2024) · International Forum of Allergy & Rhinology · doi:10.1002/alr.23458 · Cited by 11
An AIoT System for Earthquake Early Warning on Resource Constrained Devices
Esposito, M., Belli, A., Falaschetti, L., Palma, L., & Pierleoni, P. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3527750 · Cited by 11
“However, most DL-picking and detection algorithms have not originally been developed, benchmarked, or optimized for constrained IoT devices, and are therefore not suitable for the implementation in low-power, low-resources instrumentation [25].”
Dai, C., Bao, S., Chen, S., Garg, S., Kaddoum, G., & Hassan, M. M. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3557431 · Cited by 14
“Currently, there are two primary architectures for task offloading: 1) cloud computing and 2) mobile edge computing (MEC) [8].”
Pourramezani, K., Baghaee, H. R., & Gharehpetian, G. B. (2025) · Energy Reports · doi:10.1016/j.egyr.2025.08.049 · Cited by 13
Na, J., Zhang, H., Deng, X., Zhang, B., & Ye, Z. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-65310-1_6 · Cited by 14
Energy-efficient neural networks with near-threshold processors and hardware accelerators
Nunez-Yanez, J. L., & Howard, N. (2021) · Journal of Systems Architecture · doi:10.1016/j.sysarc.2021.102062 · Cited by 13
Towards On-Device Face Recognition in Body-worn Cameras
Almadan, A., & Rattani, A. (2021) · International Workshop on Biometrics and Forensics (IWBF) · doi:10.1109/iwbf50991.2021.9465079 · Cited by 11
Compact and Stable Memristive Visual Geometry Group Neural Network
Ran, H., Wen, S., Li, Q., Cao, Y., Shi, K., & Huang, T. (2021) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2021.3104860 · Cited by 12
“…trend of Internet of Things (IoT) and artifi-cial intelligence (AI) technology [4], the limitation of cloud computing, which transmits all data to the computing center for analysis and then feeds results back to the calculation performed by the terminal device, is becoming obvious [5], [6].”
BackgroundFederated Learning for Privacy Preserving On-Device Speaker Recognition
Woubie, A., & Bäckström, T. (2021) · doi:10.21437/spsc.2021-1 · Cited by 11
“The data samples are often collected on end-devices such as smartphones and the model is trained using a computationally powerful centralized server [4].”
MethodThe Cloud based Edge Computing with IoT Infrastructure and Security
Velmurugadass, P., Dhanasekaran, S., & Sasikala, S. (2021) · International Conference on Computational Performance Evaluation (ComPE) · doi:10.1109/compe53109.2021.9751942 · Cited by 11
A 5G-Based VR Application for Efficient Port Management
Han, Y., Wang, W.-H., Chen, N., Zhong, Y. P., Zhou, R., Yan, H., Wang, J., & Bai, Y. (2022) · World Electric Vehicle Journal · doi:10.3390/wevj13060101 · Cited by 12
When Deep Learning Meets Steganography: Protecting Inference Privacy in the Dark
Liu, Q., Yang, J., Jiang, H., Wu, J., Peng, T., Wang, T., & Wang, G. (2022) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom48880.2022.9796975 · Cited by 16
“By outsourcing the large portions of a DNN, we can achieve a better balance between computation and communication, while ensuring enhanced privacy [44], [45].”
MethodMo, H., Ding, S., Yang, S., Vasilakos, A. V., & Zheng, X. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3205644 · Cited by 17
“Edge computing is a feasible method for placing a fine-grained grid of computing nodes close to terminal devices [36].”
MethodTask Offloading and Resource Allocation in CPU-GPU Heterogeneous Networks
Gong, C., Ma, M., Wu, L., Liu, W., Zhou, Y., & Yang, Y. (2022) · Globecom · doi:10.1109/globecom48099.2022.10000976 · Cited by 11
“This work was supported in part by the National Key Research and Development Program of China under grant 2019YFB1803304 more powerful computing capabilities than local devices [4].”
BackgroundDeep Learning Inference at the Edge for Mobile and Aerial Robotics
Faniadis, E., & Amanatiadis, A. A. (2020) · IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR) · doi:10.1109/ssrr50563.2020.9292575 · Cited by 16
“In an attempt to meet and satisfy these requirements, edge computing provides the computation resources close to the sensors and the device [1].”
BackgroundMachine Learning at Resource Constraint Edge Device Using Bonsai Algorithm
Naveen, S., & Kounte, M. R. (2020) · International Conference on Advances in Electronics Computers and Communications · doi:10.1109/icaecc50550.2020.9339514 · Cited by 11
Domain-Aware Neural Architecture Search for Classifying Animals in Camera Trap Images
Jia, L., Tian, Y., & Zhang, J. (2022) · Animals · doi:10.3390/ani12040437 · Cited by 15
“Edge computing [16] was ideally developed for such cases, i.”
BackgroundPervasive AI for IoT Applications: Resource-efficient Distributed Artificial Intelligence.Preprint
Baccour, E., Mhaisen, N., Abdellatif, A. A., Erbad, A., Mohamed, A. M. S., Hamdi, M., & Guizani, M. (2021) · arXiv (Cornell University) · Cited by 11
Review of large YOLOv8 and RT-DETR energy efficiency on edge devices for real-time detection
Suchý, I., & Turčaník, M. (2026) · Scientific Reports · doi:10.1038/s41598-026-46453-6 · Cited by 14
Cloud-Edge System for Scheduling Unpredictable LLM Requests With Combinatorial Bandit
Li, Y., Guo, J., Tang, Z., Ding, X., Wang, J., Wang, T., & Weijia Jia, B. (2025) · IEEE Transactions on Services Computing · doi:10.1109/tsc.2025.3611379 · Cited by 13
“Edge computing offers the advantage of low-latency processing due to user proximity but is often limited in computa-tionalpowerandmemory[1],[2].”
Peng, Y., Hu, H., Li, F., Jiang, Y., Tang, J., & Liu, Y. (2025) · Computer Networks · doi:10.1016/j.comnet.2025.111748 · Cited by 12
Deep Feature Learning From Electromyographic Signals for Gesture Recognition Systems
Zhong, W., Jiang, X., Szymaniak, K., Jabbari, M., Ma, C., & Nazarpour, K. (2025) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2025.3635419 · Cited by 14
“Although deep learning algorithms typically incur greater computational costs compared to traditional machine learning techniques, recent research has identified specific structural optimizations within deep learning to decrease inference latency [172], [173].”
Expanding the Edge: Enabling Efficient Winograd CNN Inference With Deep Reuse on Edge Device
Zhang, F., Wu, R., Guan, J., Zheng, Z., Guo, X., Zhang, X., Du, X., & Shen, X. (2023) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2023.3269017 · Cited by 10
“D EEP learning has shown successes and gained popularity in data science applications [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], especially with the booming of IoT devices.”
Leng, J., Guo, J. F., Wang, D., Zhong, Y., Xu, K., Huang, S., Liu, J., Yu, C., Ye, Z., & Liu, Q. (2024) · IEEE Transactions on Computational Social Systems · doi:10.1109/tcss.2024.3395467 · Cited by 10
“BoT middleware platform places a fi ne-grained grid of computing server nodes close to the equipment on the edge, which achieves the federated training and prediction of machine learning models on the edge server, ensuring privacy, bandwidth ef fi ciency, and scalability [42].”
Towards robustness evaluation of backdoor defense on quantized deep learning models
Zhu, Y., Peng, H., Fu, A., Yang, W. Y., Ma, H., Al-Sarawi, S. F., Abbott, D., & Gao, Y. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124599 · Cited by 15
A Comparative Analysis of Compression and Transfer Learning Techniques in DeepFake Detection Models
Karathanasis, A., Violos, J., & Kompatsiaris, I. Y. (2025) · Mathematics · doi:10.3390/math13050887 · Cited by 19
Aldegheri, S., Bombieri, N., Fummi, F., Girardi, S., Muradore, R., & Piccinelli, N. (2020) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac18072.2020.9218659 · Cited by 10
“To address all these requirements, robotic SW applications have to be configured to be run across heterogeneous Cloud-Server-Edge computing platforms [2].”
MethodEnergy-efficient Collaborative Sensing: Learning the Latent Correlations of Heterogeneous Sensors
Liang, Y., Wang, X., Yu, Z., Guo, B., Zheng, X., & Samtani, S. (2021) · ACM Transactions on Sensor Networks · doi:10.1145/3448416 · Cited by 11
“In addition, edge computing is shedding lights on mobile devices oriented deep learning [8, 70].”
BackgroundA Runtime Resource Management and Provisioning Middleware for Fog Computing Infrastructures
Miele, A., Ceballos, H. Z., Cassano, L., Bolchini, C., & Ortiz, J. E. (2022) · ACM Transactions on Internet of Things · doi:10.1145/3506718 · Cited by 11
“In fact, they are mainly based on image/video/sensor processing and machine learning algorithms [Chen and Ran 2019] that follow the discussed template.”
BackgroundMethodRecognition of coal and gangue based on multi-dimensional gray gradient feature fusion
Luo, Q., Wang, S., Li, X., & He, L. (2022) · Energy Sources Part A Recovery Utilization and Environmental Effects · doi:10.1080/15567036.2022.2119309 · Cited by 11
Trajectory prediction dimensionality reduction for low-cost connected automated vehicle systems
Yao, H., Li, Q., & Li, X. (2022) · Transportation Research Part D Transport and Environment · doi:10.1016/j.trd.2022.103439 · Cited by 10
Rate-Constrained Remote Contextual Bandits
Pase, F., Gündüz, D., & Zorzi, M. (2022) · IEEE Journal on Selected Areas in Information Theory · doi:10.1109/jsait.2022.3231459 · Cited by 12
“ML techniques, mainly driven by deep neural networks (NNs), achieve stateof-the-art performance in many practical applications such as computer vision, speech recognition and automatic control [2].”
BackgroundAlmadan, A., & Rattani, A. (2023) · IEEE Access · doi:10.1109/access.2023.3265357 · Cited by 13
BIRP: Batch-aware Inference Workload Redistribution and Parallel Scheme for Edge Collaboration
Sun, H., Chen, X., Qian, Z., Li, Z., Chen, N., Cao, T., Xu, S., & Zhou, Y.-T. (2023) · Proceedings of the International Conference on Parallel Processing · doi:10.1145/3605573.3605615 · Cited by 10
“The utilization of DNN inference for extracting information features from various data types such as images, sounds, and sequential data [20], as well as leveraging the proximity of edges [4], has led to extensive applications of edge intelligence in diverse domains, including the Industrial Internet of Things(IIoT) […”
MethodResearch and Application of Edge Computing Based on Deep Learning
Cui, W. (2020) · Journal of Physics Conference Series · doi:10.1088/1742-6596/1646/1/012016 · Cited by 11
“For example: the edge device collects data, the edge server serves as the first exit point, and the cloud data center serves as the second exit point.[6] Edge caching: Optimize latency by caching DNN inference results.”
MethodAn Energy-Efficient Silicon Photonic-Assisted Deep Learning Accelerator for Big Data
Li, M., & Wang, Y. (2020) · Wireless Communications and Mobile Computing · doi:10.1155/2020/6661022 · Cited by 10
A Survey on Edge Computing in Bioinformatics and Health Informatics
Tsai, Y.-T., & Lin, Z. Y. (2020) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm49941.2020.9313103 · Cited by 12
“Most researches focus on the edge computing environment with deep learning technologies [17].”
BackgroundMethodECML: Improving Efficiency of Machine Learning in Edge Clouds
Dhakal, A., Kulkarni, S. G., & Ramakrishnan, K. K. (2020) · IEEE International Conference on Cloud Networking · doi:10.1109/cloudnet51028.2020.9335804 · Cited by 12
“Edge clouds are often used to offload processing from end-devices [6]–[8].”
Improving the Accuracy of Early Exits in Multi-Exit Architectures via Curriculum Learning
Bakhtiarnia, A., Zhang, Q., & Iosifidis, A. (2021) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn52387.2021.9533875 · Cited by 14
“Yet, deep learning has many use cases in such settings, including but not limited to video surveillance, voice assistants, network intrusion detection and augmented reality [6].”
Meyer, J., Frank, A., Schlebusch, T., & Kasneci, E. (2022) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3530884 · Cited by 19
“by adaption of bottleneck layers as proposed by [27] or by model pruning or other model optimization techniques are still possible [7] to reduce model size and thus computational complexity.”
MethodAll-Analog Silicon Integration of Image Sensor and Neural Computing Engine for Image Classification
Zambrano, B., Strangio, S., Rizzo, T., Garzón, E., Lanuzza, M., & Iannaccone, G. (2022) · IEEE Access · doi:10.1109/access.2022.3203394 · Cited by 11
Foreground segmentation and location of coal and gangue under complex similar background
Luo, Q., Wang, S., Guo, Y., He, L., Cheng, G., & Li, X. (2022) · Measurement Science and Technology · doi:10.1088/1361-6501/aca991 · Cited by 10
A Pre-Trained Deep Learning Model for Fast Online Prediction of Structural Seismic Responses
Tang, W.-J., Wang, D., Huang, H., Dai, J.-C., & Shi, F. (2023) · International Journal of Structural Stability and Dynamics · doi:10.1142/s0219455424501608 · Cited by 12
On the Pivotal Role of Artificial Intelligence Toward the Evolution of Smart Grids
Massaoudi, M., Refaat, S. S., & Abu‐Rub, H. (2021) · Wiley eBooks · doi:10.1002/9781119422464.ch15 · Cited by 12
Sparse and dense matrix multiplication hardware for heterogeneous multi-precision neural networks
Nunez-Yanez, J. L., & Hosseinabady, M. (2021) · Array · doi:10.1016/j.array.2021.100101 · Cited by 14
Federated Learning from Heterogeneous Data via Controlled Air Aggregation with Bayesian Estimation
Gafni, T., Cohen, K., & Eldar, Y. C. (2024) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2024.3351469 · Cited by 15
“The common strategy is to gather these samples at a computationally powerful server, which uses them to train its model [3].”
BackgrounddeepPGSegNet: MRI-based pituitary gland segmentation using deep learning
Choi, U.-S., Sung, Y., & Ogawa, S. (2024) · Frontiers in Endocrinology · doi:10.3389/fendo.2024.1338743 · Cited by 13
Benmerar, T. Z., Θεοδωρόπουλος, Θ., Fevereiro, D., Rosa, L., Rodrigues, J., Taleb, T., Barone, P., Giuliani, G., Tserpes, K., & Cordeiro, L. (2024) · Cluster Computing · doi:10.1007/s10586-024-04413-7 · Cited by 10
“On top of that, training models on edge devices [64] reduces latency and allows real-time decision-making.”
A Digitized You in My Eye: A Perceptually Driven Spatial Communication Prototype for XR
Hou, X., Guan, Y., Han, T., Wu, P., Lu, N., Dajer, M. A., & Peng, L. (2024) · IEEE Wireless Communications · doi:10.1109/mwc.011.2300344 · Cited by 9
The Analysis and Optimization of Volatile Clients in Over-the-Air Federated Learning
Shi, F., Lin, W., Wang, X., Li, K., & Zomaya, A. Y. (2024) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2024.3427709 · Cited by 12
“To address this challenge, emerging edge learning technologies within B5G Networks leverage distributed signal processing [2], encompassing advancements like edge computing [3], semantic communication [4], and wireless sensing network [5].”
BackgroundKim, J.-W., & Lee, H.-S. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3439852 · Cited by 10
“This mitigates security and privacy problems, lowers transmission time, and conserves network bandwidth [17], but executing the DNN model at the edge may result in larger execution time compared to the cloud due to its limited computing resources [18].”
BackgroundBojesomo, A., Liatsis, P., & Al-Marzouqi, H. (2024) · IEEE Signal Processing Magazine · doi:10.1109/msp.2024.3381808 · Cited by 12
“For example, efficient deep learning networks are important in the context of addressing the computational limitations of edge devices to support the increased adaptation of deep learning to devices limited by resources [3].”
BackgroundHao, J., Sun, J., Zhu, Z., Chen, Z., & Yan, Y. (2024) · Alexandria Engineering Journal · doi:10.1016/j.aej.2024.08.031 · Cited by 10
Salcedo, E. (2024) · Journal of Imaging · doi:10.3390/jimaging10120326 · Cited by 15
Resource Allocation Optimization Model for Computing Continuum
Mihaiu, M., Mocanu, B.-C., Negru, C., Petrescu-Niţă, A., & Pop, F. (2025) · Mathematics · doi:10.3390/math13030431 · Cited by 9
Zhang, Y., Liang, W., Xu, Z., Jia, X., & Yang, Y. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3540017 · Cited by 12
“With the advance of AI and machine learning, intelligent service provisioning in MEC networks now becomes a main research topic [4], [13], [14], [39], [44].”
Jamali, M., Hajialigol, N., & Fattahi, A. (2025) · Materials Today Sustainability · doi:10.1016/j.mtsust.2025.101098 · Cited by 14
GreenScale: Carbon Optimization for Edge Computing
Son, Y., Gupta, U., McCrabb, A., Kim, Y. G., Bertacco, V., Brooks, D., & Wu, C.-J. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3555153 · Cited by 64
“1, different components can be used by each user depending on which computing component is selected as the execution (or offloading) target [37], [38], [39].”
Afaq, Y., & Akram, S. V. (2025) · Sustainable Futures · doi:10.1016/j.sftr.2025.100761 · Cited by 14
Hybrid Adaptive MPC with Edge AI for 6-DoF Industrial Robotic Manipulators
Urrea, C. (2025) · Mathematics · doi:10.3390/math13193066 · Cited by 9
Emulating human-like adaptive vision for efficient and flexible machine visual perception
Wang, Y., Yang, Y., Yue, Y., Wang, H., Jiang, H., Han, Y., Ni, Z., Pu, Y., Shi, M., Lu, R., et al. (2025) · Nature Machine Intelligence · doi:10.1038/s42256-025-01130-7 · Cited by 38
Artificial Intelligence-Enhanced Flexible Sensors for Human Motion and Posture Sensing
Jiang, Y., & He, T. (2026) · Sensors · doi:10.3390/s26051562 · Cited by 10
“The choice between these architectures depends on the specific trade-offs between computational complexity, power consumption, and the need for immediate feedback in practical wearable sensing scenarios [100,101].”
Transistor-type optoelectronic sensors from light intensity detection to multifunctional perception
Fu, J., Jiang, H., Liu, X., Zhu, X., Kah-Wee, A., Gao, W., & Qiu, C. (2026) · Nature Sensors · doi:10.1038/s44460-026-00035-1 · Cited by 14
Dynamic Markov Model: Password Guessing Using Probability Adjustment Method
Guo, X., Liu, Y., Tan, K., Mao, W., Jin, M., & Lu, H. (2021) · Applied Sciences · doi:10.3390/app11104607 · Cited by 15
An Energy-Efficient SNN Processor Design based on Sparse Direct Feedback and Spike Prediction
Bang, S., Lew, D., Choi, S., & Park, J. (2021) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn52387.2021.9534107 · Cited by 9
“Although many personal mobile edge devices are equipped with extensive applications of artificial neural networks (ANNs) [1] these days, running complex ANN applications on resource constrained edge devices is still a challenging job due to large power consumption.”
BackgroundPreservation of Higher Accuracy Computing in Resource-Constrained Devices Using Deep Neural Approach
Manikandan, R., Mathumathi, T., Ramesh, C., Arun, S., Krishnamoorthy, R., & Padmapriya, S. (2022) · International Conference on Artificial Intelligence and Smart Energy (ICAIS) · doi:10.1109/icais53314.2022.9742923 · Cited by 9
Streaming Overlay Architecture for Lightweight LSTM Computation on FPGA SoCs
Ioannou, L., & Fahmy, S. A. (2022) · ACM Transactions on Reconfigurable Technology and Systems · doi:10.1145/3543069 · Cited by 9
SyReNN: A tool for analyzing deep neural networks
Sotoudeh, M., Tao, Z., & Thakur, A. (2023) · International Journal on Software Tools for Technology Transfer · doi:10.1007/s10009-023-00695-1 · Cited by 11
“on hardware with limited resources [32], and finally small models could be trained and deployed directly on edge devices [48,9,23,35,36].”
BackgroundIndoor Scene Recognition Mechanism Based on Direction-Driven Convolutional Neural Networks
Daou, A., Pothin, J., Honeiné, P., & Bensrhair, A. (2023) · Sensors · doi:10.3390/s23125672 · Cited by 10
“Several works have combined on-device computing with edge computing and/or cloud computing, resulting in hybrid architectures [20].”
InfluentialConvolutional neural networks combined with feature selection for radio‐frequency fingerprinting
Baldini, G., Amerini, I., Dimc, F., & Bonavitacola, F. (2023) · Computational Intelligence · doi:10.1111/coin.12592 · Cited by 9
“23 DL usually consists of a collection of processing layers that can automatically learn features from data through multiple levels of abstraction.”
BackgroundDivide-and-conquer the NAS puzzle in resource-constrained federated learning systems
Venkatesha, Y., Kim, Y., Park, H., & Panda, P. (2023) · Neural Networks · doi:10.1016/j.neunet.2023.10.006 · Cited by 11
Visibility Graph-Based Wireless Anomaly Detection for Digital Twin Edge Networks
Bertalanič, B., Hribar, J., & Fortuna, C. (2024) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2024.3393853 · Cited by 10
“To show the efficiency of our method, we measured inference times for each model for one prediction on a Raspberry Pi 5 embedded device, 6 which represents a mobile computing device [53].”
MethodComparative Analysis of Pre-trained Deep Neural Networks for Plant Disease Classification
George, R., Thuseethan, S., & Ragel, R. (2024) · International Joint Conference on Computer Science and Software Engineering · doi:10.1109/jcsse61278.2024.10613633 · Cited by 9
“However, this process often requires powerful computer devices [7] and extensive training time [8].”
BackgroundLi, L. P., Han, J.-Y., Zheng, W.-Z., Huang, R.-J., & Lai, Y.-H. (2021) · Journal of Medical Internet Research · doi:10.2196/25460 · Cited by 9
DeepFreeze: Cold Boot Attacks and High Fidelity Model Recovery on Commercial EdgeML Device
Won, Y.-S., Chatterjee, S., Jap, D., Basu, A., & Bhasin, S. (2021) · International Conference on Computer Aided Design · doi:10.1109/iccad51958.2021.9643512 · Cited by 12
Convolutional feature extraction for process monitoring using ultrasonic sensors
Bowler, A. L., Pound, M. P., & Watson, N. J. (2021) · Computers & Chemical Engineering · doi:10.1016/j.compchemeng.2021.107508 · Cited by 10
Simulating multi-agent-based computation offloading for autonomous cars
Ouarnoughi, H., Strugeon, E. G.-L., & Niar, S. (2021) · Cluster Computing · doi:10.1007/s10586-021-03440-y · Cited by 10
“Among the greedy embedded applications, AI applications for instance, and especially those based on deep learning models, can be executed: 1) on edge devices, 2) on fog or cloud servers, or 3) across these different levels [18].”
BackgroundMadokoro, H., Nix, S., Woo, H., & Sato, K. (2021) · Applied Sciences · doi:10.3390/app112411807 · Cited by 11
Metadata
- Authors
- Chen, J., & Ran, X.
- Affiliations
- University of California, Riverside (US)
- ORCID
- Chen, Jiasi
- Venue
- Proceedings of the IEEE
- Published
- 2019-08
- Volume
- 107
- Issue
- 8
- Pages
- 1655-1674
- ISSN
- 0018-9219 (ISSN-L)
- 1558-2256
- OpenAlex
- W2960833983
- Cited by
- 1,540
- References
- 141
- Citations by year
- 2026: 205 · 2025: 250 · 2024: 270 · 2023: 245 · 2022: 232 · 2021: 218 · 2020: 114 · 2019: 6
- FWCI
- 77.78
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- Proceedings of the IEEE (publishedVersion)
- Funders
- National Science Foundation (CNS-1817216)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Information Systems › Internet of Things and AI
- Keywords
- edge computing, deep learning, edge inference, distributed deep learning training, edge servers, cloud-edge collaboration, Internet of Things, low latency, privacy, spectral efficiency
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.