Are Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?

The last decade has seen exponential growth in the field of deep learning with deep learning on microcontrollers a new frontier for this research area. This paper presents a case study about machine learning on microcontrollers, with a focus on human activity recognition using accelerometer data. We...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Atis Elsts, Ryan McConville
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	machine learning deep learning neural networks activity recognition accelerometers Electronics TK7800-8360
Acceso en línea:	https://doaj.org/article/e4176e2bd89d4504a12875f6ad238f29
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:e4176e2bd89d4504a12875f6ad238f29
record_format	dspace
spelling	oai:doaj.org-article:e4176e2bd89d4504a12875f6ad238f292021-11-11T15:39:01ZAre Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?10.3390/electronics102126402079-9292https://doaj.org/article/e4176e2bd89d4504a12875f6ad238f292021-10-01T00:00:00Zhttps://www.mdpi.com/2079-9292/10/21/2640https://doaj.org/toc/2079-9292The last decade has seen exponential growth in the field of deep learning with deep learning on microcontrollers a new frontier for this research area. This paper presents a case study about machine learning on microcontrollers, with a focus on human activity recognition using accelerometer data. We build machine learning classifiers suitable for execution on modern microcontrollers and evaluate their performance. Specifically, we compare Random Forests (RF), a classical machine learning technique, with Convolutional Neural Networks (CNN), in terms of classification accuracy and inference speed. The results show that RF classifiers achieve similar levels of classification accuracy while being several times faster than a small custom CNN model designed for the task. The RF and the custom CNN are also several orders of magnitude faster than state-of-the-art deep learning models. On the one hand, these findings confirm the feasibility of using deep learning on modern microcontrollers. On the other hand, they cast doubt on whether deep learning is the best approach for this application, especially if high inference speed and, thus, low energy consumption is the key objective.Atis ElstsRyan McConvilleMDPI AGarticlemachine learningdeep learningneural networksactivity recognitionaccelerometersElectronicsTK7800-8360ENElectronics, Vol 10, Iss 2640, p 2640 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	machine learning deep learning neural networks activity recognition accelerometers Electronics TK7800-8360
spellingShingle	machine learning deep learning neural networks activity recognition accelerometers Electronics TK7800-8360 Atis Elsts Ryan McConville Are Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?
description	The last decade has seen exponential growth in the field of deep learning with deep learning on microcontrollers a new frontier for this research area. This paper presents a case study about machine learning on microcontrollers, with a focus on human activity recognition using accelerometer data. We build machine learning classifiers suitable for execution on modern microcontrollers and evaluate their performance. Specifically, we compare Random Forests (RF), a classical machine learning technique, with Convolutional Neural Networks (CNN), in terms of classification accuracy and inference speed. The results show that RF classifiers achieve similar levels of classification accuracy while being several times faster than a small custom CNN model designed for the task. The RF and the custom CNN are also several orders of magnitude faster than state-of-the-art deep learning models. On the one hand, these findings confirm the feasibility of using deep learning on modern microcontrollers. On the other hand, they cast doubt on whether deep learning is the best approach for this application, especially if high inference speed and, thus, low energy consumption is the key objective.
format	article
author	Atis Elsts Ryan McConville
author_facet	Atis Elsts Ryan McConville
author_sort	Atis Elsts
title	Are Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?
title_short	Are Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?
title_full	Are Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?
title_fullStr	Are Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?
title_full_unstemmed	Are Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?
title_sort	are microcontrollers ready for deep learning-based human activity recognition?
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/e4176e2bd89d4504a12875f6ad238f29
work_keys_str_mv	AT atiselsts aremicrocontrollersreadyfordeeplearningbasedhumanactivityrecognition AT ryanmcconville aremicrocontrollersreadyfordeeplearningbasedhumanactivityrecognition
_version_	1718434703005974528

Are Microcontrollers Ready for Deep Learning-Based Human Activity Recognition?

Ejemplares similares