Abstract
The name edge intelligence, also known as Edge AI, is a recent term used in the past few years to refer to the confluence of machine learning, or broadly speaking artificial intelligence, with edge computing. In this article, we revise the concepts regarding edge intelligence, such as cloud, edge, and fog computing, the motivation to use edge intelligence, and compare current approaches and analyze application scenarios. To provide a complete review of this technology, previous frameworks and platforms for edge computing have been discussed in this work to provide the general view of the basis for Edge AI. Similarly, the emerging techniques to deploy deep learning models at the network edge, as well as specialized platforms and frameworks to do so, are review in this article. These devices, techniques, and frameworks are analyzed based on relevant criteria at the network edge, such as latency, energy consumption, and accuracy of the models, to determine the current state of the art as well as current limitations of the proposed technologies. Because of this, it is possible to understand the current possibilities to efficiently deploy state-of-the-art deep learning models at the network edge based on technologies such as artificial intelligence accelerators, tensor processing units, and techniques that include federated learning and gossip training. Finally, the challenges of Edge AI are discussed in the work, as well as the future directions that can be extracted from the evolution of the edge computing and Internet of Things approaches.
- [1] . 2017. Characterization of 360-degree videos. In Proceedings of the Workshop on Virtual Reality and Augmented Reality Network. 1–6.Google Scholar
Digital Library
- [2] . 2016. A survey on mobile edge computing. In Proceedings of the 2016 10th International Conference on Intelligent Systems and Control (ISCO’16). 1–8. Google Scholar
Cross Ref
- [3] . 2020. Novel casestudy and benchmarking of AlexNet for edge AI: From CPU and GPU to FPGA. In Proceedings of the 2020 IEEE Canadian Conference on Electrical and Computer Engineering (CCECE’20). IEEE, Los Alamitos, CA, 1–4.Google Scholar
Cross Ref
- [4] . 2018. Cost efficient edge intelligence framework using Docker containers. In Proceedings of the 2018 IEEE 16th International Conference on Dependable, Autonomic, and Secure Computing, the 16th International Conference on Pervasive Intelligence and Computing, the 4th International Conference on Big Data Intelligence, and Computing and Cyber Science and Technology Congress (DASC/PiCom/DataCom/CyberSciTech’18). IEEE, Los Alamitos, CA, 800–807.Google Scholar
- [5] . 2011. Synthetic generation of high-dimensional datasets. IEEE Transactions on Visualization and Computer Graphics 17, 12 (2011), 2317–2324.Google Scholar
Digital Library
- [6] . 2019. Machine learning model updates in edge computing: An optimal stopping theory approach. In Proceedings of the 2019 18th International Symposium on Parallel and Distributed Computing (ISPDC’19). 1–8. Google Scholar
Cross Ref
- [7] . n.d. Benchmarking the Xnor AI2GO Platform on the Raspberry Pi. Accessed January 17, 2021 from https://www.globenewswire.com/en/news-release/2019/05/16/1826021/0/en/Xnor-releases-AI2GO-A-self-serve-edge-AI-platform-for-building-smart-on-device-solutions.html.Google Scholar
- [8] . n.d. Hands-On with the Smart-Edge Agile. Retrieved December 17, 2020 from https://blog.hackster.io/hands-on-with-the-smartedge-agile-b7b7f02b5d4b.Google Scholar
- [9] . n.d. AWS Greengrass. Retrieved December 17, 2020 from https://aws.amazon.com/greengrass/.Google Scholar
- [10] . 2017. Real-time video analytics: The killer app for edge computing. Computer 50, 10 (2017), 58–67.Google Scholar
Digital Library
- [11] . 2014. In-network cache simulations based on a YouTube traffic analysis at the edge network. In Proceedings of the 9th International Conference on Future Internet Technologies (CFI’14). ACM, New York, NY, Article
10 , 6 pages. Google ScholarDigital Library
- [12] . n.d. Retrieved December 11, 2020 from Apache Kafka. https://kafka.apache.org/.Google Scholar
- [13] . n.d. Apache MXNet. Retrieved December 3, 2020 from https://mxnet.apache.org/.Google Scholar
- [14] . n.d. Edgent. Retrieved December 17, 2020 from https://edgent.incubator.apache.org/.Google Scholar
- [15] . n.d. CoreML3 from Apple. Retrieved December 3, 2020 from https://developer.apple.com/machine-learning/core-ml/.Google Scholar
- [16] . 2017. Deep learning for Siri’s voice: On-device deep mixture density networks for hybrid unit selection synthesis. Online. Available at https://machinelearning.apple.com/2017/08/06/sirivoices.html.Google Scholar
- [17] . 2017. Hey Siri: An on-device DNN-powered voice trigger for Apple’s personal assistant. Available at https://machinelearning.apple.com/2017/10/01/heysiri.html.Google Scholar
- [18] . 2018. Artificial-Intelligence Hardware: New Opportunities for Semiconductor Companies. Technical Report. McKinsey & Company, New York, NY.Google Scholar
- [19] . 2016. Sparsification and separation of deep learning layers for constrained resource inference on wearables. In Proceedings of the 14th ACM Conference on Embedded Network Sensor Systems CD-ROM (SenSys’16).176–189. Google Scholar
Digital Library
- [20] . 2017. Cloud, fog and edge: Cooperation for the future? In Proceedings of the 2017 2nd International Conference on Fog and Mobile Edge Computing (FMEC’17). IEEE, Los Alamitos, CA, 62–67.Google Scholar
Cross Ref
- [21] . 2016. Gossip training for deep learning. arXiv preprint arXiv:1611.09726 (2016).Google Scholar
- [22] . 2019. Distributed optimization for deep learning with gossip exchange. Neurocomputing 330 (2019), 287–296.Google Scholar
Cross Ref
- [23] . 2018. FINN-R: An end-to-end deep-learning framework for fast exploration of quantized neural networks. ACM Transactions on Reconfigurable Technology and Systems 11, 3 (2018), 1–23.Google Scholar
Digital Library
- [24] . 2019. Towards federated learning at scale: System design. arXiv preprint arXiv:1902.01046 (2019).Google Scholar
- [25] . 2019. Spiking neural networks hardware implementations and challenges. ACM Journal on Emerging Technologies in Computing Systems 15, 2 (
July 2019), 1-35. Google ScholarDigital Library
- [26] . 2018. An edge based smart parking solution using camera networks and deep learning. In Proceedings of the 2018 IEEE International Conference on Cognitive Computing (ICCC’18). IEEE, Los Alamitos, CA, 17–24.Google Scholar
Cross Ref
- [27] . 2017. Edge computing architecture for applying AI to IoT. In Proceedings of the 2017 IEEE International Conference on Big Data (Big Data’17). IEEE, Los Alamitos, CA, 3012–3016.Google Scholar
Cross Ref
- [28] . 2015. FAST: A fog computing assisted distributed analytics system to monitor fall for stroke mitigation. In Proceedings of the 2015 IEEE International Conference on Networking, Architecture, and Storage (NAS’15).2–11.Google Scholar
Cross Ref
- [29] . 2017. Learning efficient object detection models with knowledge distillation. In Proceedings of the 31st International Conference on Neural Information Processing Systems. 742–751.Google Scholar
- [30] . 2019. Deep learning with edge computing: A review. Proceedings of the IEEE 107, 8 (
Aug. 2019), 1655–1674. Google ScholarCross Ref
- [31] . 2016. Smart urban surveillance using fog computing. In Proceedings of the 2016 IEEE/ACM Symposium on Edge Computing (SEC’16). IEEE, Los Alamitos, CA, 95–96.Google Scholar
Cross Ref
- [32] . 2019. F-cooper: Feature based cooperative perception for autonomous vehicle edge computing system using 3D point clouds. In Proceedings of the 4th ACM/IEEE Symposium on Edge Computing. 88–100.Google Scholar
Digital Library
- [33] . 2019. NeuroPilot: A cross-platform framework for edge-AI. In Proceedings of the 2019 IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS’19). IEEE, Los Alamitos, CA, 167–170.Google Scholar
Cross Ref
- [34] . 2019. Accelerating convolutional neural networks with dynamic channel pruning. In Proceedings of the 2019 Data Compression Conference (DCC’19). IEEE, Los Alamitos, CA, 563–563.Google Scholar
Cross Ref
- [35] . 2020. Prema: A predictive multi-task scheduling algorithm for preemptible neural processing units. In Proceedings of the 2020 IEEE International Symposium on High Performance Computer Architecture (HPCA’20). IEEE, Los Alamitos, CA, 220–233.Google Scholar
Cross Ref
- [36] . 2014. Neuromorphic hardware system for visual pattern recognition with memristor array and CMOS neuron. IEEE Transactions on Industrial Electronics 62, 4 (2014), 2410–2419.Google Scholar
Cross Ref
- [37] . n.d. Cisco Fog Director. Retrieved December 17, 2020 from https://www.cisco.com/c/en/us/products/cloud-systems-management/fog-director/index.html.Google Scholar
- [38] . 2003. A knowledge plane for the internet. In Proceedings of the 2003 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communications. 3–10.Google Scholar
Digital Library
- [39] . n.d. Edge Computing Reference Architecture 2.0. Retrieved February 16, 2022 from http://en.ecconsortium.net/Uploads/file/20180328/1522232376480704.pdf.Google Scholar
- [40] . 2019. Edge computing framework: Understanding the opportunity roadmap. http://docplayer.net/149826792-Edge-computing-framework-understanding-the-opportunity-roadmap-chetan-sharma.html.Google Scholar
- [41] . 1967. Nearest neighbor pattern classification. IEEE Transactions on Information Theory 13, 1 (1967), 21–27.Google Scholar
Digital Library
- [42] . n.d. CRESCO. Retrieved December 15, 2020 from http://cresco.io/.Google Scholar
- [43] . n.d. Crosser. Retrieved December 17, 2020 from https://crosser.io/.Google Scholar
- [44] . 2020. Offloading autonomous driving services via edge computing. IEEE Internet of Things Journal 7, 10 (2020), 10535–10547.Google Scholar
Cross Ref
- [45] . 2018. GossipGraD: Scalable deep learning using gossip communication based asynchronous gradient descent. arXiv preprint arXiv:1803.05880 (2018).Google Scholar
- [46] . 2005. Histograms of oriented gradients for human detection. In Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05). 996–893. Google Scholar
Digital Library
- [47] . 2019. A survey of deep learning and its applications: A new paradigm to machine learning. Archives of Computational Methods in Engineering 27 (2019), 1071–1092.Google Scholar
- [48] . 2020. A hybrid secure routing and monitoring mechanism in IoT-based wireless sensor networks. Ad Hoc Networks 97 (2020), 102022.Google Scholar
Digital Library
- [49] . 2020. Edge intelligence: The confluence of edge computing and artificial intelligence. IEEE Internet of Things Journal 7, 8 (2020), 7457–7469.Google Scholar
Cross Ref
- [50] . 2019. Deep learning on mobile devices—A review. arXiv:1904.09274 (2019). Google Scholar
Cross Ref
- [51] . 2011. An empirical evaluation of easily implemented, non parametric methods for generating synthetic datasets. Computational Statistics and Data Analysis 55, 12 (
Dec. 2011), 3232–3243. Google ScholarCross Ref
- [52] . 2017. A reconfigurable streaming deep convolutional neural network accelerator for Internet of Things. IEEE Transactions on Circuits and Systems I: Regular Papers 65, 1 (2017), 198–208.Google Scholar
Cross Ref
- [53] . 2017. UCI Machine Learning Repository. Retrieved February 16, 2022 from http://archive.ics.uci.edu/ml.Google Scholar
- [54] . 2021. Benchmarking quantized neural networks on FPGAs with FINN. arXiv preprint arXiv:2102.01341 (2021).Google Scholar
- [55] . 2019. Modeling and evaluating a complex edge computing based systems: An emergency management support system case study. Internet of Things 6 (2019), 100054. Google Scholar
Cross Ref
- [56] . 2018. White Paper on AI Chip Technologies (2018). Tsinghua University and Beijing Innovation Center for Future Chips.Google Scholar
- [57] . 2019. What’s Powering Artificial Intelligence? White Paper. ARM.Google Scholar
- [58] . n.d. Caffe2Go. Retrieved December 3, 2020 from https://pytorch.org/.Google Scholar
- [59] . 2017. FAR-EDGE project H2020. Online. Available at http://faredge.eu/#/.Google Scholar
- [60] . 2018. An edge computing architecture in the Internet of Things. In Proceedings of the 2018 IEEE 21st International Symposium on Real-Time Distributed Computing (ISORC’18). IEEE, Los Alamitos, CA, 99–102.Google Scholar
Cross Ref
- [61] . 2019. Towards the decentralised cloud: Survey on approaches and challenges for mobile, ad hoc, and edge computing. ACM Computing Surveys 51, 6 (2019), 1–36.Google Scholar
Digital Library
- [62] . n.d. Foghorn Framework. Retrieved December 17, 2020 from https://www.foghorn.io/.Google Scholar
- [63] . n.d. OGEMA. Retrieved December 3, 2020 from https://www.iis.fraunhofer.de/en/ff/lv/iot-system/tech/edge-computing/edge-framework.html.Google Scholar
- [64] . 2018. A multi-cloudlet infrastructure for future smart cities: An empirical study. In Proceedings of the 1st International Workshop on Edge Systems, Analytics, and Networking (EdgeSys’18). ACM, New York, NY, 19–24. Google Scholar
Digital Library
- [65] . 2017. An extensible edge computing architecture: Definition, requirements and enablers. Proceedings of the 11th International Conference on Mobile Ubiquitous Computing, Systems, Services, and Technologies (UBICOMM’17).Google Scholar
- [66] . 2018. Medical image forgery detection for smart healthcare. IEEE Communications Magazine 56, 4 (2018), 33–37.Google Scholar
Cross Ref
- [67] . 2016. On developing smart transportation applications in fog computing paradigm. In Proceedings of the 6th ACM Symposium on Development and Analysis of Intelligent Vehicular Networks and Applications (DIVANet’16). ACM, New York, NY, 91–98. Google Scholar
Digital Library
- [68] . 2016. Smart agents and fog computing for smart city applications. In Proceedings of the First International Conference on Smart Cities—Volume 9704 (Smart-CT 2016). Springer-Verlag, Berlin, Germany, 137–146. Google Scholar
Digital Library
- [69] . 2018. Multimodal deep neural networks using both engineered and learned representations for biodegradability prediction. arXiv preprint arXiv:1808.04456 (2018).Google Scholar
- [70] . n.d. Google Coral. Retrieved November 21, 2020 from https://coral.ai/.Google Scholar
- [71] . n.d. ML Kit. Retrieved December 3, 2020 from https://firebase.google.com/products/ml-kit.Google Scholar
- [72] . n.d. SparkFun Edge. Retrieved December 3, 2020 from https://www.sparkfun.com/products/15170.Google Scholar
- [73] . n.d. TensorFlow Lite. Retrieved November 21, 2020 from https://www.tensorflow.org/lite.Google Scholar
- [74] . n.d. TensorFlow Lite Performance. Retrieved November 21, 2020 from https://www.tensorflow.org/lite/performance/gpu_advanced.Google Scholar
- [75] . 2020. Blockchain powered secure range-free localization in wireless sensor networks. Arabian Journal for Science and Engineering 45, 8 (2020), 6139–6155.Google Scholar
Cross Ref
- [76] . 2019. Optimization in Edge Computing and Small-Cell Networks. Springer International, Cham, Switzerland, 17–31. Google Scholar
Cross Ref
- [77] . 2017. ProtoNN: Compressed and accurate kNN for resource-scarce devices. In Proceedings of the International Conference on Machine Learning. 1331–1340.Google Scholar
- [78] . 2014. Towards wearable cognitive assistance. In Proceedings of the 12th Annual International Conference on Mobile Systems, Applications, and Services. 68–81.Google Scholar
Digital Library
- [79] . 2013. Just-in-time provisioning for cyber foraging. In Proceeding of the 11th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys’13). ACM, New York, NY, 153–166. Google Scholar
Digital Library
- [80] . 2019. Smart-Edge-CoCaCo: AI-enabled smart edge with joint computation, caching, and communication in heterogeneous IoT. IEEE Network 33, 2 (2019), 58–64. Google Scholar
Cross Ref
- [81] . 2015. Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531 (
March 2015).Google Scholar - [82] . 2020. Meta-learning in neural networks: A survey. arXiv:2004.05439 (2020).Google Scholar
- [83] . 2018. Edge computing framework for enabling situation awareness in IoT based smart city. Journal of Parallel and Distributed Computing 122 (2018), 226–237.Google Scholar
Cross Ref
- [84] . 2018. Predictive view generation to enable mobile 360-degree and VR experiences. In Proceedings of the 2018 Morning Workshop on Virtual Reality and Augmented Reality Network. 20–26.Google Scholar
Digital Library
- [85] . 2019. Short-term traffic prediction by two-level data driven model in 5G-enabled edge computing networks. IEEE Access 7 (2019), 123981–123991.Google Scholar
Cross Ref
- [86] . 2018. Building edge intelligence for online activity recognition in service-oriented IoT systems. Future Generation Computer Systems 87 (2018), 557–567.Google Scholar
Digital Library
- [87] . 2018. VideoEdge: Processing camera streams using hierarchical clusters. In Proceedings of the 2018 IEEE/ACM Symposium on Edge Computing (SEC’18). IEEE, Los Alamitos, CA, 115–131.Google Scholar
Cross Ref
- [88] . 2017. DeepMon: Mobile GPU-based deep learning framework for continuous vision applications. In Proceedings of the 15th Annual International Conference on Mobile Systems, Applications, and Services. 82–95.Google Scholar
Digital Library
- [89] . n.d. Intel Movidius. Retrieved December 3, 2020 from https://www.movidius.com/.Google Scholar
- [90] . 2013. Increasing Mobile Operators’ Value Proposition with Edge Computing. Technical Brief. Intel.Google Scholar
- [91] . n.d. Kinect Dataset. Retrieved November 21, 2020 from https://rgbd-dataset.cs.washington.edu/publications.htmlGoogle Scholar
- [92] . n.d. Google TPU. Accessed January 24, 2021 from https://coral.ai/docs/edgetpu/faq/.Google Scholar
- [93] . 2017. Combining edge computing and blockchains for flexibility and performance in industrial automation. In Proceedings of the International Conference on Mobile Ubiquitous Computing, Systems, Services, and Technologies (UBICOMM’17).Google Scholar
- [94] . 2019. Taxonomy of Edge Computing: Challenges, Opportunities, and Data Reduction Methods. Springer International, Cham, Switzerland, 51–69. Google Scholar
Cross Ref
- [95] . 2019. Related’s Hudson Yards: Smart city or surveillance city? Online. Available at https://therealdeal.com/2019/03/15/hudsonyards-smart-city-or-surveillance-city/.Google Scholar
- [96] . 2018. Enabling full body AR with Mask R-CNN2Go. https://research.facebook.com/blog/2018/01/enabling-full-body-ar-with-mask-r-cnn2go/.Google Scholar
- [97] . 2017. Neurosurgeon: Collaborative intelligence between the cloud and mobile edge. ACM SIGARCH Computer Architecture News 45, 1 (2017), 615–629.Google Scholar
Digital Library
- [98] . 2019. Adaptable and data-driven softwarized networks: Review, opportunities, and challenges. Proceedings of the IEEE 107, 4 (2019), 711–731.Google Scholar
Cross Ref
- [99] . 2019. Edge computing: A survey. Future Generation Computer Systems 97 (2019), 219–235. Google Scholar
Digital Library
- [100] . 2017. Short paper: Towards low-cost indoor localization using edge computing resources. In Proceedings of the 2017 IEEE 20th International Symposium on Real-Time Distributed Computing (ISORC’17). IEEE, Los Alamitos, CA, 28–31.Google Scholar
Cross Ref
- [101] . 2011. Neural synaptic weighting with a pulse-based memristor circuit. IEEE Transactions on Circuits and Systems I: Regular Papers 59, 1 (2011), 148–158.Google Scholar
Cross Ref
- [102] . 2015. A shared parking model in vehicular network using fog and cloud environment. In Proceedings of the 2015 17th Asia-Pacific Network Operations and Management Symposium (APNOMS’15). IEEE, Los Alamitos, CA, 321–326.Google Scholar
Cross Ref
- [103] . 2021. Special session: Reliability analysis for ML/AI hardware. arXiv preprint arXiv:2103.12166 (2021).Google Scholar
- [104] . 2019. FastGRNN: A fast, accurate, stable and tiny kilobyte sized gated recurrent neural network. arXiv preprint arXiv:1901.02358 (2019).Google Scholar
- [105] . 2016. Deep learning on FPGAs: Past, present, and future. arXiv preprint arXiv:1602.04283 (2016).Google Scholar
- [106] . 2016. Neural architectures for named entity recognition. arXiv preprint arXiv:1603.01360 (2016).Google Scholar
- [107] . 2018. Technology trend of edge AI. In Proceedings of the 2018 International Symposium on VLSI Design, Automation, and Test (VLSI-DAT’18). IEEE, Los Alamitos, CA, 1–2.Google Scholar
Cross Ref
- [108] . 2017. PiCasso: A lightweight edge computing platform. In Proceedings of the 2017 IEEE 6th International Conference on Cloud Networking (CloudNet’17). IEEE, Los Alamitos, CA, 1–7.Google Scholar
Cross Ref
- [109] . 2016. DeepCham: Collaborative edge-mediated adaptive deep learning for mobile object recognition. In Proceedings of the 2016 IEEE/ACM Symposium on Edge Computing (SEC’16).64–76.Google Scholar
Cross Ref
- [110] . 2020. Edge AI: On-demand accelerating deep neural network inference via edge computing. IEEE Transactions on Wireless Communications 19, 1 (2020), 447–457. Google Scholar
Cross Ref
- [111] . 2018. JALAD: Joint accuracy-and latency-aware deep structure decoupling for edge-cloud execution. In Proceedings of the 2018 IEEE 24th International Conference on Parallel and Distributed Systems (ICPADS’18). IEEE, Los Alamitos, CA, 671–678.Google Scholar
Cross Ref
- [112] . 2016. Mobile edge computing: Progress and challenges. In Proceedings of the 2016 4th IEEE International Conference on Mobile Cloud Computing, Services, and Engineering (MobileCloud’16). IEEE, Los Alamitos, CA, 83–84.Google Scholar
Cross Ref
- [113] . 2010. Object bank: A high-level image representation for scene classification & semantic feature sparsification. In Advances in Neural Information Processing Systems 23 (NIPS’10).Google Scholar
- [114] . 2020. A survey of AI accelerators for edge environment. In World Conference on Information Systems and Technologies. Springer, 35–44.Google Scholar
Cross Ref
- [115] . 2017. Runtime neural pruning. In Proceedings of the 31st International Conference on Neural Information Processing Systems. 2178–2188.Google Scholar
Digital Library
- [116] . n.d. EdgeX. Retrieved December 15, 2020 from https://www.edgexfoundry.org/.Google Scholar
- [117] . 2018. Blockchain-enabled security in electric vehicles cloud and edge computing. IEEE Network 32, 3 (2018), 78–83.Google Scholar
- [118] . 2017. Edge computing framework for distributed smart applications. In Proceedings of the 2017 IEEE Conference on SmartWorld, Ubiquitous Intelligence and Computing, Advanced and Trusted Computed, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People, and Smart City Innovation (SmartWorld/SCALCOM/UIC/ ATC/CBDCom/IOP/SCI’17). IEEE, Los Alamitos, CA, 1–8.Google Scholar
- [119] . 2020. Vehicular edge computing and networking: A survey. arXiv:1980.06849 (2020).Google Scholar
- [120] . 2019. Edge assisted real-time object detection for mobile augmented reality. In Proceedings of the 25th Annual International Conference on Mobile Computing and Networking. 1–16.Google Scholar
Digital Library
- [121] . 2020. AdaDeep: A usage-driven, automated deep model compression framework for enabling ubiquitous intelligent mobiles. IEEE Transactions on Mobile Computing 20 (2020), 3282–3297.Google Scholar
- [122] . 2019. Edge computing for autonomous driving: Opportunities and challenges. Proceedings of the IEEE 107, 8 (2019), 1697–1716.Google Scholar
Cross Ref
- [123] . 2019. Intelligent edge computing for IoT-based energy management in smart cities. IEEE Network 33, 2 (2019), 111–117. Google Scholar
Cross Ref
- [124] . 2019. BeagleBone AI Makes a Sneak Preview. Retrieved December 17, 2020 from https://beagleboard.org/blog/2019-05-16-beaglebone-ai-preview.Google Scholar
- [125] . n.d. Macchina Platform for the Network Edge. Retrieved November 21, 2020 from https://macchina.io/.Google Scholar
- [126] . 2015. A survey on smart grid metering infrastructures: Threats and solutions. In Proceedings of the 2015 IEEE International Conference on Electro/Information Technology (EIT’15). IEEE, Los Alamitos, CA, 386–391.Google Scholar
Cross Ref
- [127] . 2010. Preceding vehicle detection using histograms of oriented gradients. In Proceedings of the 2010 International Conference on Communications, Circuits, and Systems (ICCCAS’10). IEEE, Los Alamitos, CA, 354–358.Google Scholar
Cross Ref
- [128] . 2015. Ad hoc cloud computing. In Proceedings of the 2015 IEEE 8th International Conference on Cloud Computing. IEEE, Los Alamitos, CA, 1063–1068.Google Scholar
Digital Library
- [129] . 2017. Communication-efficient learning of deep networks from decentralized data. In Proceedings of the 20th International Conference on Artificial Intelligence and Statistics. 1273–1282.Google Scholar
- [130] . 2019. Device-enhanced MEC: Multi-access edge computing (MEC) aided by end device computation and caching: A survey. IEEE Access 7 (2019), 166079–166108.Google Scholar
Cross Ref
- [131] . 2017. Knowledge-defined networking. ACM SIGCOMM Computer Communication Review 47, 3 (2017), 2–10.Google Scholar
Digital Library
- [132] . n.d. Azure IoT Edge. Retrieved December 17, 2020 from https://azure.microsoft.com/es-es/services/iot-edge/.Google Scholar
- [133] . n.d. MIT Pedestrian Database. Retrieved December 22, 2020 from http://cbcl.mit.edu/software-datasets/PedestrianData.html.Google Scholar
- [134] . 2018. Introduction to edge computing in IIoT. https://www.iiconsortium.org/pdf/Introduction_to_Edge_Computing_in_IIoT_2018-06-18.pdf.Google Scholar
- [135] . 2018. Deep learning for IoT big data and streaming analytics: A survey. IEEE Communications Surveys Tutorials 20, 4 (2018), 2923–2960. Google Scholar
Digital Library
- [136] . 2019. Edge intelligence-assisted smoke detection in foggy surveillance environments. IEEE Transactions on Industrial Informatics 16, 2 (2019), 1067–1075.Google Scholar
Cross Ref
- [137] . 2019. arXiv:1908.00080 (2019).Google Scholar
- [138] . 2018. Fog computing: Survey of trends, architectures, requirements, and research directions. IEEE Access 6 (2018), 47980–48009. Google Scholar
Cross Ref
- [139] . 2016. Pruning random forests for prediction on a budget. In Advances in Neural Information Processing Systems. 2334–2342.Google Scholar
- [140] . 2019. PipeDream: Generalized pipeline parallelism for DNN training. In Proceedings of the 27th ACM Symposium on Operating Systems Principles. 1–15.Google Scholar
Digital Library
- [141] . 2017. A machine learning approach to pedestrian detection for autonomous vehicles using high-definition 3D range data. Sensors 17, 1 (2017), 18.Google Scholar
- [142] . 2011. Multimodal deep learning. In Proceedings of the 28th International Conference in Machine Learning (ICML’11). 689–696.Google Scholar
- [143] . n.d. Dragon NaturallySpeaking. Retrieved January 24, 2021 from https://www.nuance.com/dragon.html.Google Scholar
- [144] . n.d. Nvidia Jetson Nano. Retrieved November 21, 2020 from https://www.nvidia.com/es-es/autonomous-machines/embedded-systems/jetson-nano/.Google Scholar
- [145] . 2013. Joint deep learning for pedestrian detection. In Proceedings of the IEEE International Conference on Computer Vision.2056–2063.Google Scholar
- [146] . 2009. A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering 22, 10 (2009), 1345–1359.Google Scholar
Digital Library
- [147] . 2015. A survey of cloudlet based mobile computing. In Proceedings of the 2015 International Conference on Cloud Computing and Big Data (CCBD’15). 268–275. Google Scholar
Digital Library
- [148] . 2013. A survey of methods for distributed machine learning. Progress in Artificial Intelligence 2, 1 (2013), 1–11.Google Scholar
- [149] . 2020. A survey of multi-access edge computing in 5G and beyond: Fundamentals, technology integration, and state-of-the-art. IEEE Access 8 (2020), 116974–117017.Google Scholar
Cross Ref
- [150] . 2018. Edge intelligence: Challenges and opportunities of near-sensor machine learning applications. In Proceedings of the 2018 IEEE 29th International Conference on Application-specific Systems, Architectures, and Processors (ASAP’18). 1–7. Google Scholar
Cross Ref
- [151] . 2016. Going deeper with embedded FPGA platform for convolutional neural network. In Proceedings of the 2016 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays.26–35. Google Scholar
Digital Library
- [152] . 2019. The role of mobile edge computing towards assisting IoT with distributed intelligence: A smartliving perspective. In Mobile Solutions and Their Usefulness in Everyday Life. Springer, 33–45.Google Scholar
Cross Ref
- [153] . 2018. Exploiting smart e-Health gateways at the edge of Healthcare Internet-of-Things: A fog computing approach. Future Generation Computer Systems 78 (2018), 641–658. Google Scholar
Digital Library
- [154] . 2019. Edge computing for Internet of Things: A survey, e-healthcare case study and future direction. Journal of Network and Computer Applications 140 (2019), 1–22. Google Scholar
Digital Library
- [155] . 2019. Edge computing for Internet of Things: A survey, e-healthcare case study and future direction. Journal of Network and Computer Applications 140 (2019), 1–22.Google Scholar
Digital Library
- [156] . 2016. Minerva: Enabling low-power, highly-accurate deep neural network accelerators. In Proceedings of the 2016 ACM/IEEE 43rd Annual International Symposium on Computer Architecture (ISCA’16). IEEE, Los Alamitos, CA, 267–278.Google Scholar
Digital Library
- [157] . 2019. AI at the Edge: A GigaOm Research Byte. ARM.Google Scholar
- [158] . 2017. Serving at the edge: A scalable IoT architecture based on transparent computing. IEEE Network 31, 5 (2017), 96–105.Google Scholar
Digital Library
- [159] . 2017. DeepSense: A unified deep learning framework for time-series mobile sensing data processing. In Proceedings of the 26th International Conference on World Wide Web.351–360.Google Scholar
- [160] . 2016. Mobile-edge computing architecture: The role of MEC in the Internet of Things. IEEE Consumer Electronics Magazine 5 (2016), 84–91.Google Scholar
Cross Ref
- [161] . 2019. Edge computing framework for wearable sensor-based human activity recognition. In Proceedings of the International Symposium on Innovative and Interdisciplinary Applications of Advanced Technologies. 376–387.Google Scholar
- [162] . 2019. Efficient deployment of predictive analytics in edge gateways: Fall detection scenario. In Proceedings of the 2019 IEEE 5th World Forum on Internet of Things (WF-IoT’19). IEEE, Los Alamitos, CA, 41–46.Google Scholar
Cross Ref
- [163] . 2020. A survey of edge computing-based designs for IoT security. Digital Communications and Networks 6, 2 (2020), 195–202.Google Scholar
Cross Ref
- [164] . 2021. Cloud-native network slicing using software defined networking based multi-access edge computing: A survey. IEEE Access 9 (2021), 10903–10924.Google Scholar
Cross Ref
- [165] . 2018. Ultra-low latency cloud-fog computing for Industrial Internet of Things. In Proceedings of the 2018 IEEE Wireless Communications and Networking Conference (WCNC’18). 1–6. Google Scholar
Digital Library
- [166] . 2016. Edge computing: Vision and challenges. IEEE Internet of Things Journal 3, 5 (2016), 637–646.Google Scholar
- [167] . 2016. Decision jungles: Compact and rich models for classification. In Proceedings of the 26th International Conference on Neural Information Processing Systems (NIPS’13). 234–242.Google Scholar
- [168] . 2019. A review of binarized neural networks. Electronics 8, 6 (
June 2019), 661. Google ScholarCross Ref
- [169] . 2019. A review of edge computing reference architectures and a new global edge proposal. Future Generation Computer Systems 99 (2019), 278–294.Google Scholar
Digital Library
- [170] . 2019. Edge computing, IoT and social computing in smart energy scenarios. Sensors 19, 15 (2019), 3353.Google Scholar
Cross Ref
- [171] . 2019. Artificial intelligence-driven mechanism for edge computing-based industrial applications. IEEE Transactions on Industrial Informatics 15, 7 (
July 2019), 4235–4243. Google ScholarCross Ref
- [172] . 2019. Machine learning models to tiny IoT devices. In Proceedings of the 40th ACM SIGPLAN Conference on Programming Language Design and Implementation.79–95.Google Scholar
- [173] . 2015. Smart items, fog and cloud computing as enablers of servitization in healthcare. Sensors & Transducers 185, 2 (2015), 121.Google Scholar
- [174] . 2019. Heterogeneous computing for edge AI. In Proceedings of the 2019 International Symposium on VLSI Design, Automation, and Test (VLSI-DAT’19). IEEE, Los Alamitos, CA, 1–2.Google Scholar
Cross Ref
- [175] . 2021. A novel hybrid optimization for cluster-based routing protocol in information-centric wireless sensor networks for IoT based mobile edge computing. Wireless Personal Communications 2021 (2021), 1–24.Google Scholar
Digital Library
- [176] . n.d. Neuroshield. Retrieved November 21, 2020 from https://www.general-vision.com/hardware/neuroshield/Google Scholar
- [177] . 2018. When edge meets learning: Adaptive control for resource-constrained distributed machine learning. In Proceedings of the 2018 IEEE INFOCOM Conference.Google Scholar
Digital Library
- [178] . 2020. Convergence of edge computing and deep learning: A comprehensive survey. IEEE Communications Surveys & Tutorials 22, 2 (2020), 869–904.Google Scholar
Cross Ref
- [179] . 2019. In-Edge AI: Intelligentizing mobile edge computing, caching and communication by federated learning. IEEE Network 33, 5 (2019), 156–165.Google Scholar
Digital Library
- [180] . 2015. Big Data: Principles and Best Practices of Scalable Realtime Data Systems. Simon & Schuster.Google Scholar
- [181] . 2017. Automated systolic array architecture synthesis for high throughput CNN inference on FPGAs. In Proceedings of the 54th Annual Design Automation Conference.1–6.Google Scholar
Digital Library
- [182] . 2019. Machine learning + Distributed IoT = Edge intelligence. In Proceedings of the 2019 IEEE 39th International Conference on Distributed Computing Systems (ICDCS’19). 1715–1719. Google Scholar
Cross Ref
- [183] . 2019. Machine learning at Facebook: Understanding inference at the edge. In Proceedings of the 2019 IEEE International Symposium on High Performance Computer Architecture (HPCA’19). IEEE, Los Alamitos, CA, 331–344.Google Scholar
Cross Ref
- [184] . 2016. Google’s neural machine translation system: Bridging the gap between human and machine translation. arXiv preprint arXiv:1609.08144 (2016).Google Scholar
- [185] . 2019. Demonstrating advantages of neuromorphic computation: A pilot study. Frontiers in Neuroscience 13 (2019), 260.Google Scholar
Cross Ref
- [186] . n.d. AI2GO. Retrieved December 3, 2020 from https://www.xnor.ai/.Google Scholar
- [187] . 2018. Systolic array based accelerator and algorithm mapping for deep learning algorithms. In IFIP International Conference on Network and Parallel Computing. Springer, 153–158.Google Scholar
Digital Library
- [188] . 2010. AmazingStore: Available, low-cost online storage service using cloudlets. In Proceedings of the 9th International Conference on Peer-to-Peer Systems (IPTPS’10). 2. http://dl.acm.org/citation.cfm?id=1863145.1863147.Google Scholar
- [189] . 2017. DeepIoT: Compressing deep neural network structures for sensing systems with a compressor-critic framework. In Proceedings of the 15th ACM Conference on Embedded Network Sensor Systems. 1–14.Google Scholar
Digital Library
- [190] . 2017. A 1.06-to-5.09 TOPS/W reconfigurable hybrid-neural-network processor for deep learning applications. In Proceedings of the 2017 Symposium on VLSI Circuits. IEEE, Los Alamitos, CA, C26–C27.Google Scholar
Cross Ref
- [191] . 2017. A high energy efficient reconfigurable hybrid neural network processor for deep learning applications. IEEE Journal of Solid-State Circuits 53, 4 (2017), 968–982.Google Scholar
Cross Ref
- [192] . 2018. Recent trends in deep learning based natural language processing. IEEE Computational Intelligence Magazine 13, 3 (2018), 55–75.Google Scholar
Cross Ref
- [193] . 2015. A deep learning prediction process accelerator based FPGA. In Proceedings of the 2015 15th IEEE/ACM International Symposium on Cluster, Cloud, and Grid Computing. IEEE, Los Alamitos, CA, 1159–1162.Google Scholar
Digital Library
- [194] . 2017. A survey on the edge computing for the Internet of Things. IEEE Access 6 (2017), 6900–6919.Google Scholar
Cross Ref
- [195] . 2014. Augmented brain computer interaction based on fog computing and linked data. In Proceedings of the 2014 International Conference on Intelligent Environments. 374–377. Google Scholar
Digital Library
- [196] . 2019. Artificial intelligence/machine learning in IoT for authentication and authorization of edge devices. In Proceedings of the 2019 International Conference on Applied and Engineering Mathematics (ICAEM’19). 220–224. Google Scholar
Cross Ref
- [197] . 2015. The design and implementation of a wireless video surveillance system. In Proceedings of the 21st Annual International Conference on Mobile Computing and Networking. 426–438.Google Scholar
Digital Library
- [198] . 2019. OpenEI: An open framework for edge intelligence. In Proceedings of the 2019 IEEE 39th International Conference on Distributed Computing Systems (ICDCS’19). IEEE, Los Alamitos, CA, 1840–1851.Google Scholar
Cross Ref
- [199] . 2018. pCAMP: Performance comparison of machine learning packages on the edges. In Proceedings of the USENIX Workshop on Hot Topics in Edge Computing (HotEdge’18).Google Scholar
- [200] . 2019. On-board artificial intelligence based on edge computing in optical transport networks. In Proceedings of the 2019 Optical Fiber Communications Conference and Exhibition (OFC’19). IEEE, Los Alamitos, CA, 1–3.Google Scholar
Cross Ref
- [201] . 2018. DeepThings: Distributed adaptive deep learning inference on resource-constrained iot edge clusters. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems 37, 11 (2018), 2348–2359.Google Scholar
Cross Ref
- [202] . 2019. Edge intelligence: Paving the last mile of artificial intelligence with edge computing. Proceedings of the IEEE 107, 8 (
Aug. 2019), 1738–1762. Google ScholarCross Ref
Index Terms
Edge Intelligence: Concepts, Architectures, Applications, and Future Directions
Recommendations
Edge Computing with Artificial Intelligence: A Machine Learning Perspective
Recent years have witnessed the widespread popularity of Internet of things (IoT). By providing sufficient data for model training and inference, IoT has promoted the development of artificial intelligence (AI) to a great extent. Under this background and ...
Roadmap for edge AI: a Dagstuhl perspective
Based on the collective input of Dagstuhl Seminar (21342), this paper presents a comprehensive discussion on AI methods and capabilities in the context of edge computing, referred as Edge AI. In a nutshell, we envision Edge AI to provide adaptation for ...
Edge Intelligence: On-Demand Deep Learning Model Co-Inference with Device-Edge Synergy
MECOMM'18: Proceedings of the 2018 Workshop on Mobile Edge CommunicationsAs the backbone technology of machine learning, deep neural networks (DNNs) have have quickly ascended to the spotlight. Running DNNs on resource-constrained mobile devices is, however, by no means trivial, since it incurs high performance and energy ...






Comments