Infineon Technologies AG, Germany
Dept. Computer Science and Artificial Intelligence, University of Granada (Spain), Spain
Dept. of Electronics and Computer Technology, University of Granada (Spain), Spain
Edge Intelligence
的名称，也称为
Edge AI
，是最近几年使用的一个术语，指的是机器学习或广义的人工智能与边缘计算的融合。在这篇手稿中，我们修改了有关边缘智能的概念，例如云计算、边缘计算和雾计算，以及使用边缘智能的动机，并比较了当前的方法并分析了应用场景。为了提供对这项技术的完整回顾，本手稿中讨论了以前的边缘计算框架和平台，以提供边缘人工智能基础的一般视图。同样，在本手稿中回顾了在网络边缘部署深度学习 (DL) 模型的新兴技术，以及这样做的专用平台和框架。这些设备，基于网络边缘的相关标准（例如模型的延迟、能耗和准确性）分析技术和框架，以确定当前技术水平以及所提出技术的当前局限性。因此，有可能了解当前基于人工智能加速器、张量处理单元和包括联邦学习和八卦训练在内的技术在网络边缘有效部署最先进的深度学习模型的可能性是什么. 最后，手稿中讨论了边缘 AI 的挑战以及可以从边缘计算和物联网 (IoT) 方法的演变中提取的未来方向。模型的能耗和准确性，以确定当前的技术水平以及所提出技术的当前限制。因此，有可能了解当前基于人工智能加速器、张量处理单元和包括联邦学习和八卦训练在内的技术在网络边缘有效部署最先进的深度学习模型的可能性是什么. 最后，手稿中讨论了边缘 AI 的挑战以及可以从边缘计算和物联网 (IoT) 方法的演变中提取的未来方向。模型的能耗和准确性，以确定当前的技术水平以及所提出技术的当前限制。因此，有可能了解当前基于人工智能加速器、张量处理单元和包括联邦学习和八卦训练在内的技术在网络边缘有效部署最先进的深度学习模型的可能性是什么. 最后，手稿中讨论了边缘 AI 的挑战以及可以从边缘计算和物联网 (IoT) 方法的演变中提取的未来方向。基于人工智能加速器、张量处理单元和包括联邦学习和八卦训练在内的技术，可以了解当前在网络边缘有效部署最先进的深度学习模型的可能性是什么。最后，手稿中讨论了边缘 AI 的挑战以及可以从边缘计算和物联网 (IoT) 方法的演变中提取的未来方向。基于人工智能加速器、张量处理单元和包括联邦学习和八卦训练在内的技术，可以了解当前在网络边缘有效部署最先进的深度学习模型的可能性是什么。最后，手稿中讨论了边缘 AI 的挑战以及可以从边缘计算和物联网 (IoT) 方法的演变中提取的未来方向。
The name
Edge Intelligence
, also known as
Edge AI
, is a recent term used in the last few years to refer to the confluence of Machine Learning, or broadly speaking Artificial Intelligence, with Edge Computing. In this manuscript, 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 discusses in this manuscript in order to provide the general view of the basis for Edge AI. Similarly, the emerging techniques to deploy Deep Learning (DL) models at the network edge, as well as specialized platforms and frameworks to do so, are review in this manuscript. 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 what are the current possibilities to efficiently deploy state-of-the-art DL models at the network edge based on technologies such as AI accelerators, Tensor Processing Units and techniques that include Federated Learning and Gossip Training. Finally, the challenges of Edge AI are discusses in the manuscript as well as the Future directions that can be extracted from the evolution of the Edge Computing and Internet of Things (IoT) approaches.
