Hardware Solutions for Low-Power Smart Edge Computing

Wisniewski, Lucas Martin; Bec, Jean-Michel; Boguszewski, Guillaume; Gamatié, Abdoulaye

doi:10.3390/jlpea12040061

Cited by 11 publications

(4 citation statements)

References 72 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The RP2040 is suitable for edge machine learning applications due to its low cost, low power consumption, and high performance. It has 264KB of RAM, and a dual Arm Cortex-M0+ core that can run at up to 133 MHz, with enough processing power to handle ML models [44]. The features of our edge device are:…”

Section: Edge Device Architecturementioning

confidence: 99%

Vibe - With people with hearing impairment

Velmurugan S,

Prabhakar A

2023

ITU J-FET

View full text Add to dashboard Cite

Hearing loss significantly impacts daily life, leading to communication difficulties, social isolation, and increased risk of accidents. This can cause frustration, anxiety, and depression. The use of sound indication devices improve the quality of life for individuals with hearing impairment by providing them with a means to detect sounds in their environment. We have developed Vibe, a device that alerts a user to specific sounds (a baby's cry, an alarm, a door knock, a vehicle horn, or a spoken name) in their environment. Our device uses EdgeML, a form of machine learning, to perform tasks locally on the device rather than relying on cloud-based services. This approach has the advantage of reduced latency and improved privacy. The EdgeML model was trained on a locally collected dataset, for better applicability to the Indian environment. Devices such as Vibe will help individuals with hearing impairment navigate their surroundings safely and confidently, leading to increased independence and improved mental health.

show abstract

Section: Edge Device Architecturementioning

confidence: 99%

Vibe - With people with hearing impairment

Velmurugan S,

Prabhakar A

2023

ITU J-FET

View full text Add to dashboard Cite

show abstract

“…In essence, fog is a vital link between the network's edge and the cloud. The fog devices in proximity enable the facilitation of the implementation of emerging IoT applications with low latency and stringent requirements [28,29]. As introduced in [30], edge computing represents a novel paradigm that aims to provide storage and computational capabilities in proximity to end-users and the IoT domain.…”

Section: Fog and Edge Computingmentioning

confidence: 99%

Exploring the Potential of Distributed Computing Continuum Systems

Donta,

Murturi,

Casamayor Pujol

et al. 2023

Computers

View full text Add to dashboard Cite

Computing paradigms have evolved significantly in recent decades, moving from large room-sized resources (processors and memory) to incredibly small computing nodes. Recently, the power of computing has attracted almost all current application fields. Currently, distributed computing continuum systems (DCCSs) are unleashing the era of a computing paradigm that unifies various computing resources, including cloud, fog/edge computing, the Internet of Things (IoT), and mobile devices into a seamless and integrated continuum. Its seamless infrastructure efficiently manages diverse processing loads and ensures a consistent user experience. Furthermore, it provides a holistic solution to meet modern computing needs. In this context, this paper presents a deeper understanding of DCCSs’ potential in today’s computing environment. First, we discuss the evolution of computing paradigms up to DCCS. The general architectures, components, and various computing devices are discussed, and the benefits and limitations of each computing paradigm are analyzed. After that, our discussion continues into various computing devices that constitute part of DCCS to achieve computational goals in current and futuristic applications. In addition, we delve into the key features and benefits of DCCS from the perspective of current computing needs. Furthermore, we provide a comprehensive overview of emerging applications (with a case study analysis) that desperately need DCCS architectures to perform their tasks. Finally, we describe the open challenges and possible developments that need to be made to DCCS to unleash its widespread potential for the majority of applications.

show abstract

“…Previous work in the area of hardware acceleration leveraged central processing unit (CPU), graphic processing unit (GPU), application-specific integrated circuit (ASIC), and field programmable array (FPGA) to design and implement embedded ML. In [7], low-power, ultra low-power, and powerful embedded devices for ML at the edge are listed and grouped by their GPU, CPU, acceleration capacity, ML usage, and application examples. A comprehensive review of the state-of-the-art tools and techniques for efficient edge inference is provided by Shuvo et al in [8].…”

Section: Introductionmentioning

confidence: 99%

A Study on the Application of TensorFlow Compression Techniques to Human Activity Recognition

Contoli

Lattanzi

2023

IEEE Access

View full text Add to dashboard Cite

In the human activity recognition (HAR) application domain, the use of deep learning (DL) algorithms for feature extractions and training purposes delivers significant performance improvements with respect to the use of traditional machine learning (ML) algorithms. However, this comes at the expense of more complex and demanding models, making harder their deployment on constrained devices traditionally involved in the HAR process. The efficiency of DL deployment is thus yet to be explored. We thoroughly investigated the application of TensorFlow Lite simple conversion, dynamic, and full integer quantization compression techniques. We applied those techniques not only to convolutional neural networks (CNNs), but also to long short-term memory (LSTM) networks, and a combined version of CNN and LSTM. We also considered two use case scenarios, namely cascading compression and stand-alone compression mode. This paper reports the feasibility of deploying deep networks onto an ESP32 device, and how TensorFlow compression techniques impact classification accuracy, energy consumption, and inference latency. Results show that in the cascading case, it is not possible to carry out the performance characterization. Whereas in the stand-alone case, dynamic quantization is recommended because yields a negligible loss of accuracy. In terms of power efficiency, both dynamic and full integer quantization provide high energy saving with respect to the uncompressed models: between 31% and 37% for CNN networks, and up to 45% for LSTM networks. In terms of inference latency, dynamic and full integer quantization provide comparable performance.

show abstract

Hardware Solutions for Low-Power Smart Edge Computing

Cited by 11 publications

References 72 publications

Vibe - With people with hearing impairment

Vibe - With people with hearing impairment

Exploring the Potential of Distributed Computing Continuum Systems

A Study on the Application of TensorFlow Compression Techniques to Human Activity Recognition

Contact Info

Product

Resources

About