Filters








2,772 Hits in 6.1 sec

Enabling Deep Learning on Edge Devices through Filter Pruning and Knowledge Transfer [article]

Kaiqi Zhao, Yitao Chen, Ming Zhao
2022 arXiv   pre-print
Second, it proposes a novel knowledge transfer method to enable the on-device model to update incrementally in real time or near real time using incremental learning on new data and enable the on-device  ...  Deep learning models have introduced various intelligent applications to edge devices, such as image classification, speech recognition, and augmented reality.  ...  The second approach to learning on devices is based on knowledge transfer which uses the knowledge distilled from a cloud-based deep model (termed teacher) to improve the accuracy of a on-device small  ... 
arXiv:2201.10947v1 fatcat:4frcqhypebboxcnsa5wck75qca

LOW LATENCY DEEP LEARNING INFERENCE MODEL FOR DISTRIBUTED INTELLIGENT IOT EDGE CLUSTERS

Soumyalatha Naveen, Manjunath R Kounte, Mohammed Riyaz Ahmed
2021 IEEE Access  
Therefore, intelligence needs to be applied through an efficient deep learning model to optimize resources like memory, power, and computational ability.  ...  Evaluation was done by deploying the proposed model on five IoT edge devices and a gateway device enabled with hardware accelerator.  ...  ACKNOWLEDGMENT The authors are greatly indebted to the anonymous reviewers whose thought-provoking and encouraging comments have motivated them to modify significantly and update the paper.  ... 
doi:10.1109/access.2021.3131396 fatcat:gqllzrrvcjhjfjjcgb45gux5bu

FedNILM: Applying Federated Learning to NILM Applications at the Edge [article]

Yu Zhang, Guoming Tang, Qianyi Huang, Yi Wang, Xudong Wang, Jiadong Lou
2021 arXiv   pre-print
cloud model compression via filter pruning and multi-task learning, and iii) personalized edge model building with unsupervised transfer learning.  ...  Specifically, FedNILM is designed to deliver privacy-preserving and personalized NILM services to large-scale edge clients, by leveraging i) secure data aggregation through federated learning, ii) efficient  ...  FedNILM realized data privacy-preserving through federated learning, efficient model compression via filter pruning and multi-task learning, and personalized model building by unsupervised transfer learning  ... 
arXiv:2106.07751v1 fatcat:zzoindueibbw7enbaggrwrndl4

Edge Intelligence: Architectures, Challenges, and Applications [article]

Dianlei Xu, Tong Li, Yong Li, Xiang Su, Sasu Tarkoma, Tao Jiang, Jon Crowcroft, Pan Hui
2020 arXiv   pre-print
Edge intelligence refers to a set of connected systems and devices for data collection, caching, processing, and analysis in locations close to where data is captured based on artificial intelligence.  ...  We first identify four fundamental components of edge intelligence, namely edge caching, edge training, edge inference, and edge offloading, based on theoretical and practical results pertaining to proposed  ...  Similar training architecture is also used in [123] to enable knowledge transferring amongst edge devices. B.  ... 
arXiv:2003.12172v2 fatcat:xbrylsvb7bey5idirunacux6pe

Prune2Edge: A Multi-Phase Pruning Pipelines to Deep Ensemble Learning in IIoT [article]

Besher Alhalabi, Mohamed Gaber, Shadi Basurra
2021 arXiv   pre-print
In this paper, we propose a novel edge-based multi-phase pruning pipelines to ensemble learning on IIoT devices.  ...  However, learning and using deep learning models are computationally expensive, so an IoT device with limited computational power could not run such models.  ...  ACKNOWLEDGEMENT The authors would like to thank Google for the generous credit grant that allows using cutting-edge deep learning frameworks on their cloud platform.  ... 
arXiv:2004.04710v2 fatcat:dlxzje7gqnagpg3cevw4t2t3mi

SPATL: Salient Parameter Aggregation and Transfer Learning for Heterogeneous Clients in Federated Learning [article]

Sixing Yu, Phuong Nguyen, Waqwoya Abebe, Wei Qian, Ali Anwar, Ali Jannesari
2022 arXiv   pre-print
Federated learning (FL) facilitates the training and deploying AI models on edge devices.  ...  a shared encoder and a local predictor, and transferring its knowledge to heterogeneous clients via the locally customized predictor.  ...  Heterogeneous Knowledge Transfer Learning Inspired by transfer learning [7] , SPATL aims to train an encoder in FL setting and address the heterogeneity issue through transferring the encoder's knowledge  ... 
arXiv:2111.14345v2 fatcat:wogvgqhkhjbj3k644sxv4zhife

Enabling Deep Learning for All-in EDGE paradigm [article]

Praveen Joshi, Haithem Afli, Mohammed Hasanuzzaman, Chandra Thapa, Ted Scully
2022 arXiv   pre-print
In this regard, this survey paper investigates Deep Learning at the edge, its architecture, enabling technologies, and model adaption techniques, where edge servers and edge devices participate in deep  ...  Deep Learning-based models have been widely investigated, and they have demonstrated significant performance on non-trivial tasks such as speech recognition, image processing, and natural language understanding  ...  We presented an overview of the architectures, enabling technologies and model adaption techniques that enable EDGE intelligence through Deep Learning.  ... 
arXiv:2204.03326v1 fatcat:khhmeo5tyra55kvfbyohswff7u

State-of-the-art Techniques in Deep Edge Intelligence [article]

Ahnaf Hannan Lodhi, Barış Akgün, Öznur Özkasap
2020 arXiv   pre-print
Edge Intelligence (EI) has quickly emerged as a powerful alternative to enable learning using the concepts of Edge Computing.  ...  Deep Learning-based Edge Intelligence or Deep Edge Intelligence (DEI) lies in this rapidly evolving domain. In this article, we provide an overview of the major constraints in operationalizing DEI.  ...  The end device houses multiple smaller submodels trained on knowledge partitions whereas baseline models are kept at the edge device.  ... 
arXiv:2008.00824v3 fatcat:aofzt6tfbzhvdhkuqlp6njvh6e

A Survey on Efficient Convolutional Neural Networks and Hardware Acceleration

Deepak Ghimire, Dayoung Kil, Seong-heum Kim
2022 Electronics  
In this review, to improve the efficiency of deep learning research, we focus on three aspects: quantized/binarized models, optimized architectures, and resource-constrained systems.  ...  Over the past decade, deep-learning-based representations have demonstrated remarkable performance in academia and industry.  ...  Acknowledgments: We appreciate our reviewers and editors for their precious time in providing valuable comments and improving our paper.  ... 
doi:10.3390/electronics11060945 fatcat:bxxgccwkujatzh4onkzh5lgspm

Adaptive Dynamic Pruning for Non-IID Federated Learning [article]

Sixing Yu, Phuong Nguyen, Ali Anwar, Ali Jannesari
2021 arXiv   pre-print
Learning models at edge devices such as cell phones is one of the most common use case of FL.  ...  However, the limited computing power and energy constraints of edge devices hinder the adoption of FL for both model training and deployment, especially for the resource-hungry Deep Neural Networks (DNNs  ...  Machine learning models' increasing memory and computing power requirements (such as deep neural networks) make their deployment on edge devices a grand challenge.  ... 
arXiv:2106.06921v1 fatcat:4igg3t2h4ff45fjs5m4hmc5jei

Bringing AI To Edge: From Deep Learning's Perspective [article]

Di Liu, Hao Kong, Xiangzhong Luo, Weichen Liu, Ravi Subramaniam
2020 arXiv   pre-print
As various and diverse methods which are applicable to edge systems are proposed intensively, a holistic review would enable edge computing engineers and community to know the state-of-the-art deep learning  ...  However, the development of edge intelligence systems encounters some challenges, and one of these challenges is the computational gap between computation-intensive deep learning algorithms and less-capable  ...  This research was conducted in collaboration with HP Inc. and supported by National Research Foundation (NRF) Singapore and the Singapore Government through the Industry Alignment Fund-Industry Collaboration  ... 
arXiv:2011.14808v1 fatcat:g6ib7v7cxbdglihkizw5ldsxcu

Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing

Zhi Zhou, Xu Chen, En Li, Liekang Zeng, Ke Luo, Junshan Zhang
2019 Proceedings of the IEEE  
We then provide an overview of the overarching architectures, frameworks, and emerging key technologies for deep learning model toward training/inference at the network edge.  ...  | With the breakthroughs in deep learning, the recent years have witnessed a booming of artificial intelligence (AI) applications and services, spanning from personal assistant to recommendation systems  ...  The overview of PipeDream's automated mechanism is shown in Fig. 10 . 5) Knowledge Transfer Learning: Knowledge transfer learning, or transfer learning for simplicity, is closely connected with DNN  ... 
doi:10.1109/jproc.2019.2918951 fatcat:d53vxmklgfazbmzjhsq3tuoama

TinyML for Ubiquitous Edge AI [article]

Stanislava Soro
2021 arXiv   pre-print
TinyML is a fast-growing multidisciplinary field at the intersection of machine learning, hardware, and software, that focuses on enabling deep learning algorithms on embedded (microcontroller powered)  ...  TinyML will open the door to the new types of edge services and applications that do not rely on cloud processing but thrive on distributed edge inference and autonomous reasoning.  ...  Index-4 In event-based applications, data analytics on the edge device can filter out invaluable data before sending out the processed information.  ... 
arXiv:2102.01255v1 fatcat:if5ny6kcirdkhnj56mswfaptlm

Efficient Synthesis of Compact Deep Neural Networks [article]

Wenhan Xia, Hongxu Yin, Niraj K. Jha
2020 arXiv   pre-print
For example, autonomous driving requires fast inference based on Internet-of-Things (IoT) edge devices operating under run-time energy and memory storage constraints.  ...  Deep neural networks (DNNs) have been deployed in myriad machine learning applications.  ...  One such method is DeepInversion [46] , an image synthesis methodology that enables data-free knowledge transfer. One of its many applications is data-free pruning.  ... 
arXiv:2004.08704v1 fatcat:g6gu7ng2zjda7minnahitn455a

Edge Machine Learning for AI-Enabled IoT Devices: A Review

Massimo Merenda, Carlo Porcaro, Demetrio Iero
2020 Sensors  
In this work, a detailed review on models, architecture, and requirements on solutions that implement edge machine learning on Internet of Things devices is presented, with the main goal to define the  ...  Deploying machine learning on such edge devices improves the network congestion by allowing computations to be performed close to the data sources.  ...  Acknowledgments: Programma di Azione Coesione PAC Calabria 2014-2020, Asse Prioritario 12, Azione 10.5.12, is gratefully acknowledged by one of the authors (D.I.).  ... 
doi:10.3390/s20092533 pmid:32365645 pmcid:PMC7273223 fatcat:aug36rsafndb5ghzinog3qhbnq
« Previous Showing results 1 — 15 out of 2,772 results