time-series-5

Edge-computing

3260 papers • 126 benchmarks • 313 datasets

Deep Learning on EDGE devices

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in edge-computing-10

Trend

Dataset

Best Model

Actions

No benchmarks available.

Libraries

i

Use these libraries to find edge-computing-10 models and implementations

yoshitomo-matsubara/torchdistill

3 papers 1,256

Datasets

No datasets available.

Subtasks

Tiling & Deployment Device-Cloud Collaboration

Most implemented papers

TransVisDrone: Spatio-Temporal Transformer for Vision-based Drone-to-Drone Detection in Aerial Videos

M. Shah, Waqas Sultani, Tushar Sangam, I. Dave•Sat Oct 15 2022

This work utilizes CSPDarkNet-53 network to learn object-related spatial features and VideoSwin model to improve drone detection in challenging scenarios by learning spatio-temporal dependencies of drone motion and proposes a simple yet effective framework, TransVisDrone, that provides an end-to-end solution with higher computational efficiency.

36

Content

0

Paper Graph

Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach

Xiaodong Wang, Zhao Chen•Sat Dec 15 2018

Numerical results demonstrate that the proposed DDPG-based strategy can help each user learn an efficient dynamic offloading policy and verify the superiority of its continuous power allocation capability to policies learned by conventional discrete action space-based reinforcement learning approaches like deep Q-network (DQN).

173 0

Paper Graph

On the Convergence of FedAvg on Non-IID Data

Xiang Li, Kaixuan Huang, Wenhao Yang, Shusen Wang, Zhihua Zhang•Wed Jul 03 2019

This paper analyzes the convergence of Federated Averaging on non-iid data and establishes a convergence rate of $\mathcal{O}(\frac{1}{T})$ for strongly convex and smooth problems, where $T$ is the number of SGDs.

2794 0

Paper Graph

Distilled Split Deep Neural Networks for Edge-Assisted Real-Time Systems

M. Levorato, Sameer Singh, Yoshitomo Matsubara, S. Baidya, Davide Callegaro•Sun Oct 06 2019

Different from prior literature presenting DNN splitting frameworks, the architecture of the head DNN is distill to reduce its computational complexity and introduce a bottleneck, thus minimizing processing load at the mobile device as well as the amount of wirelessly transferred data.

109 0

Paper Graph

Scientific Image Restoration Anywhere

V. Abeykoon, Zhengchun Liu, R. Kettimuthu, G. Fox, Ian T Foster•Thu Oct 31 2019

The use of deep learning models within scientific experimental facilities frequently requires low-latency inference, so that, for example, quality control operations can be performed while data are being collected. Edge computing devices can be useful in this context, as their low cost and compact form factor permit them to be co-located with the experimental apparatus. Can such devices, with their limited resources, can perform neural network feed-forward computations efficiently and effectively? We explore this question by evaluating the performance and accuracy of a scientific image restoration model, for which both model input and output are images, on edge computing devices. Specifically, we evaluate deployments of TomoGAN, an image-denoising model based on generative adversarial networks developed for low-dose x-ray imaging, on the Google Edge TPU and NVIDIA Jetson. We adapt TomoGAN for edge execution, evaluate model inference performance, and propose methods to address the accuracy drop caused by model quantization. We show that these edge computing devices can deliver accuracy comparable to that of a full-fledged CPU or GPU model, at speeds that are more than adequate for use in the intended deployments, denoising a 1024x1024 image in less than a second. Our experiments also show that the Edge TPU models can provide 3x faster inference response than a CPU-based model and 1.5x faster than an edge GPU-based model. This combination of high speed and low cost permits image restoration anywhere.

19 0

Paper Graph

Graph Markov Network for Traffic Forecasting with Missing Data

Zhiyong Cui, Longfei Lin, Ziyuan Pu, Yinhai Wang•Mon Dec 09 2019

This study considers the traffic network as a graph and defines the transition between network-wide traffic states at consecutive time steps as aGraph Markov process, which proposes a new neural network architecture for spatial-temporal data forecasting, i.e. the graph Markov network (GMN) and a spectral graph MarkOV network (SGMN).

106 0

Paper Graph

Split Computing for Complex Object Detectors: Challenges and Preliminary Results

M. Levorato, Yoshitomo Matsubara•Sun Jul 26 2020

This is the first study to inject small bottlenecks to such object detectors and unveil the potential of a split computing approach, and it is shown that naive split computing methods would not reduce inference time.

31 0

Paper Graph

Head Network Distillation: Splitting Distilled Deep Neural Networks for Resource-Constrained Edge Computing Systems

M. Levorato, Sameer Singh, Yoshitomo Matsubara, S. Baidya, Davide Callegaro•Thu Nov 19 2020

This paper proposes to modify the structure and training process of DNN models for complex image classification tasks to achieve in-network compression in the early network layers to obtain aggressive compression while preserving accuracy.

78 0

Paper Graph

Systolic-CNN: An OpenCL-defined Scalable Run-time-flexible FPGA Accelerator Architecture for Accelerating Convolutional Neural Network Inference in Cloud/Edge Computing

A. Dua, Yixing Li, Fengbo Ren•Thu Apr 30 2020

Systolic-CNN is an OpenCLdefined scalable, run-time-flexible FPGA accelerator architecture, optimized for performing the low-latency, energy-efficient inference of various convolutional neural networks (CNNs) in the context of multi-tenancy cloud/edge computing.

14 0

Paper Graph

Neural Compression and Filtering for Edge-assisted Real-time Object Detection in Challenged Networks

M. Levorato, Yoshitomo Matsubara•Thu Jul 30 2020

This work focuses on DNNs for three different object detection tasks, and modify the architecture of the network to achieve in-network compression by introducing a bottleneck layer in the early layers on the head model, and prefilter pictures that do not contain objects of interest using a convolutional neural network.

65 0

Paper Graph

Adding a benchmark result helps the community track progress.