Machine Learning at the Network Edge: A Survey

Murshed, M. G. Sarwar; Murphy, Christopher; Hou, Daqing; Khan, Nazar; Ananthanarayanan, Ganesh; Hussain, Faraz

doi:10.1145/3469029

Cited by 205 publications

(111 citation statements)

References 81 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…GANs have shown impressive performance in various tasks [75,41,35]. It not only provides a kind of novel learning algorithm but also offers a vital alternative to generative models.…”

Section: Gan-based Methodsmentioning

confidence: 99%

3D Human Motion Prediction: A Survey

Lyu¹,

Chen²,

Liu³

et al. 2022

Preprint

View full text Add to dashboard Cite

3D human motion prediction, predicting future poses from a given sequence, is an issue of great significance and challenge in computer vision and machine intelligence, which can help machines in understanding human behaviors. Due to the increasing development and understanding of Deep Neural Networks (DNNs) and the availability of large-scale human motion datasets, the human motion prediction has been remarkably advanced with a surge of interest among academia and industrial community. In this context, a comprehensive survey on 3D human motion prediction is conducted for the purpose of retrospecting and analyzing relevant works from existing released literature. In addition, a pertinent taxonomy is constructed to categorize these existing approaches for 3D human motion prediction. In this survey, relevant methods are categorized into three categories: human pose representation, network structure design, and prediction target. We systematically review all relevant journal and conference papers in the field of human motion prediction since 2015, which are presented in detail based on proposed categorizations in this survey. Furthermore, the outline for the public benchmark datasets, evaluation criteria, and performance comparisons are respectively presented in this paper. The limitations of the state-of-the-art methods are discussed as well, hoping for paving the way for future explorations.

show abstract

“…GANs have shown impressive performance in various tasks [75,41,35]. It not only provides a kind of novel learning algorithm but also offers a vital alternative to generative models.…”

Section: Gan-based Methodsmentioning

confidence: 99%

3D Human Motion Prediction: A Survey

Lyu¹,

Chen²,

Liu³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Edge computing is a concept where processing is done close to the device that produced the data, which generally means on devices with much less memory than regular computing servers. There are many surveys about classification for edge computing [15], [16], [17], but most of the work focuses on deep learning, which is not applicable in our case because it requires a lot of data and time to train the model. They discuss inherent problems related to learning with edge devices, in particular about lighter architecture and distributed training.…”

Section: Related Workmentioning

confidence: 99%

“…Finally, these studies draw future work opportunities such as data augmentation, distributed training, and explainable AI. Aside from the deep learning approaches, the survey in [15] discusses two machine learning techniques with a small memory footprint: the Bonsai and ProtoNN methods. Bonsai [18] is a tree-based algorithm designed to fit in an edge device memory.…”

Section: Related Workmentioning

confidence: 99%

Mondrian Forest for Data Stream Classification Under Memory Constraints

Khannouz¹,

Glatard²

2022

Preprint

View full text Add to dashboard Cite

Supervised learning algorithms generally assume the availability of enough memory to store their data model during the training and test phases. However, in the Internet of Things, this assumption is unrealistic when data comes in the form of infinite data streams, or when learning algorithms are deployed on devices with reduced amounts of memory. In this paper, we adapt the online Mondrian forest classification algorithm to work with memory constraints on data streams. In particular, we design five out-of-memory strategies to update Mondrian trees with new data points when the memory limit is reached. Moreover, we design trimming mechanisms to make Mondrian trees more robust to concept drifts under memory constraints. We evaluate our algorithms on a variety of real and simulated datasets, and we conclude with recommendations on their use in different situations: the Extend Node strategy appears as the best out-of-memory strategy in all configurations, whereas different trimming mechanisms should be adopted depending on whether a concept drift is expected. All our methods are implemented in the OrpailleCC open-source library and are ready to be used on embedded systems and connected objects.

show abstract

“…Extreme-edge inference is achievable in practical cases since it can be performed with low precision integer operations that help increase the energy efficiency, reduce the memory footprint and the area overhead, with reduced accuracy loss [1], [2]. Eyeriss [3] is a hardware accelerator designed for Convolutional Neural Networks (CNNs) inference with INT16 arithmetic and implemented in 65 nm technology.…”

Section: Introduction and Related Workmentioning

confidence: 99%

RedMulE: A Compact FP16 Matrix-Multiplication Accelerator for Adaptive Deep Learning on RISC-V-Based Ultra-Low-Power SoCs

Tortorella

Bertaccini

Rossi

et al. 2022

2022 Design, Automation &Amp; Test in Europe Conference &Amp; Exhibition (DATE)

View full text Add to dashboard Cite

The fast proliferation of extreme-edge applications using Deep Learning (DL) based algorithms required dedicated hardware to satisfy extreme-edge applications' latency, throughput, and precision requirements. While inference is achievable in practical cases, online finetuning and adaptation of general DL models are still highly challenging. One of the key stumbling stones is the need for parallel floating-point operations, which are considered unaffordable on sub-100 mW extreme-edge SoCs. We tackle this problem with RedMulE (Reduced-precision matrix Multiplication Engine), a parametric low-power hardware accelerator for FP16 matrix multiplications -the main kernel of DL training and inference -conceived for tight integration within a cluster of tiny RISC-V cores based on the PULP (Parallel Ultra-Low-Power) architecture. In 22 nm technology, a 32-FMA RedMulE instance occupies just 0.07 mm 2 (14% of an 8-core RISC-V cluster) and achieves up to 666 MHz maximum operating frequency, for a throughput of 31.6 MAC/cycle (98.8% utilization). We reach a cluster-level power consumption of 43.5 mW and a full-cluster energy efficiency of 688 16-bit GFLOPS/W. Overall, RedMulE features up to 4.65× higher energy efficiency and 22× speedup over SW execution on 8 RISC-V cores.

show abstract

Machine Learning at the Network Edge: A Survey

Cited by 205 publications

References 81 publications

3D Human Motion Prediction: A Survey

3D Human Motion Prediction: A Survey

Mondrian Forest for Data Stream Classification Under Memory Constraints

RedMulE: A Compact FP16 Matrix-Multiplication Accelerator for Adaptive Deep Learning on RISC-V-Based Ultra-Low-Power SoCs

Contact Info

Product

Resources

About