OpenEI: An Open Framework for Edge Intelligence

  title={OpenEI: An Open Framework for Edge Intelligence},
  author={Xingzhou Zhang and Yifan Wang and Sidi Lu and Liangkai Liu and Lanyu Xu and Weisong Shi},
  journal={2019 IEEE 39th International Conference on Distributed Computing Systems (ICDCS)},
In the last five years, edge computing has attracted tremendous attention from industry and academia due to its promise to reduce latency, save bandwidth, improve availability, and protect data privacy to keep data secure. At the same time, we have witnessed the proliferation of AI algorithms and models which accelerate the successful deployment of intelligence mainly in cloud services. These two trends, combined together, have created a new horizon: Edge Intelligence (EI). The development of… 

Figures and Tables from this paper

Leveraging the Potentials of Federated AI Ecosystems
Light is shed on this research gap and a research agenda is proposed to foster the potentials of value cocreation within federated AI ecosystems.
Design Methodology for Deep Out-of-Distribution Detectors in Real-Time Cyber-Physical Systems
A design methodology is proposed to tune deep OOD detectors to meet the accuracy and response time requirements of embedded applications and it is shown that this design methodology can lead to a drastic reduction in response time in relation to an unoptimized OOD detector while maintaining comparable accuracy.
Digital Twin-Enabled AI Enhancement in Smart Critical Infrastructures for 5G
This work proposes a Digital Twin-Enabled Edge AI (DTE2AI), supported by the Energy-aware High Accuracy Strategy (EAHAS), which focuses on optimizing the training accuracy of AI tasks under the limits of training time and energy consumption.
LaBaNI: Layer-based Noise Injection Attack on Convolutional Neural Networks
This paper proposes a hardware-intrinsic attack (HIA), Layer-based Noise Injection (LaBaNI), which successfully performs misclassification without knowing the initial and final layers of the CNN.
Enabling Deep Learning for All-in EDGE paradigm
The key performance metrics for Deep Learning at the All-in EDGE paradigm are presented to evaluate various deep learning techniques and choose a suitable design to overcome difficulties due to other requirements such as high computation, high latency, and high bandwidth caused by Deep Learning applications in real-world scenarios.
A Systematic Literature Review on Distributed Machine Learning in Edge Computing
The challenges of running ML/DL on edge devices in a distributed way are investigated, paying special attention to how techniques are adapted or designed to execute on these restricted devices.
New Opportunities, Challenges, and Applications of Edge-AI for Connected Healthcare in Internet of Medical Things for Smart Cities
It has been recommended that the different models that have been proposed in several studies must be validated further and implemented in different domains, to validate the effectiveness of these models and to ensure that these models can be implemented in several regions effectively.
Edge Intelligence: Concepts, architectures, applications and future directions
The concepts regarding Edge Intelligence are revised, such as Cloud, Edge and Fog Computing, the motivation to use Edge Intelligence, and the challenges of Edge AI are discussed as well as the Future directions that can be extracted from the evolution of the Edge Computing and Internet of Things (IoT) approaches.
Serving at the Edge: An Edge Computing Service Architecture Based on ICN
An edge computing service architecture based on Information-Centric Networking is proposed, which contains the edge Computing service session model, service request forwarding strategies, and service dynamic deployment mechanism, which can avoid unnecessary network transmissions and reduce the service completion time.
Efficient Federated Meta-Learning Over Multi-Access Wireless Networks
This paper rigorously analyzes the contribution of each device to the global loss reduction in each round and develops an FML algorithm with a non-uniform device selection scheme to accelerate the convergence and formulate a resource allocation problem integrating NUFM in multi-access wireless systems to jointly improve the convergence rate.


TensorFlow: A system for large-scale machine learning
The TensorFlow dataflow model is described and the compelling performance that Tensor Flow achieves for several real-world applications is demonstrated.
MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems
The API design and the system implementation of MXNet are described, and it is explained how embedding of both symbolic expression and tensor operation is handled in a unified fashion.
The Amazon Echo
Smart speakers such as the Amazon Echo Dot and the Google Home Mini are the most important new product category of recent years and are having a major impact on American culture and society. This is
AutoVAPS: an IoT-enabled public safety service on vehicles
An IoT-Enabled public safety service called AutoVAPS which integrates body-worn cameras and other sensors on the vehicle for public safety and a reference architecture that consists of the data layer for data management, the model layer for edge intelligence, and the access layer for privacy-preserving data sharing and access is proposed.
Big Data for Health
  • V. Persico
  • Medicine
    Encyclopedia of Big Data Technologies
  • 2019
Bayesian Dark Knowledge
This work describes a method for “distilling” a Monte Carlo approximation to the posterior predictive density into a more compact form, namely a single deep neural network.
MUVR: Supporting Multi-User Mobile Virtual Reality with Resource Constrained Edge Cloud
  • Yong Li, Wei Gao
  • Computer Science
    2018 IEEE/ACM Symposium on Edge Computing (SEC)
  • 2018
This paper envisions the key reason of such challenges as the ignorance of redundancy across VR frames being rendered, and aims to fundamentally remove this performance constraint on highly dynamic VR applications by adaptively reusing the redundant VR framesbeing rendered for different VR users.
Enabling Edge Intelligence for Activity Recognition in Smart Homes
This work proposes an Edge-based solution for addressing the activity recognition problem in smart homes from multiple perspectives, including architecture, algorithm design and system implementation, and proposes a convolutional neural network model to perform activity recognition tasks on Edge devices.
FastGRNN: A Fast, Accurate, Stable and Tiny Kilobyte Sized Gated Recurrent Neural Network
The FastRNN and FastGRNN algorithms are developed to address the twin RNN limitations of inaccurate training and inefficient prediction and to be deployed on severely resource-constrained IoT microcontrollers too tiny to store other RNN models.
Multiple Instance Learning for Efficient Sequential Data Classification on Resource-constrained Devices
This paper proposes a method, EMI-RNN, that improves standard LSTM model’s accuracy by up to 1% while requiring 72x less computation, which enables us to deploy such models for continuous real-time prediction on a small device such as Raspberry Pi0 and Arduino variants, a task that the baseline L STM could not achieve.