Seeking Optimum System Settings for Physical Activity Recognition on Smartwatches

被引：1

作者：

Ahmad, Muhammad ^{[1
,2
]}

Khan, Adil ^{[1
]}

Mazzara, Manuel ^{[1
]}

Distefano, Salvatore ^{[2
]}

机构：

[1] Innopolis Univ, Kazan, Russia

[2] Univ Messina, Messina, Italy

来源：

ADVANCES IN COMPUTER VISION, VOL 2 | 2020年 / 944卷

关键词：

Smartwatch; Accelerometer; Magnetometer; Gyroscope; Machine learning; Physical activity recognition; Health care services;

D O I：

10.1007/978-3-030-17798-0_19

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Physical activity recognition using wearable devices can provide valued information regarding an individual's degree of functional ability and lifestyle. Smartphone-based physical activity recognition is a well-studied area. However, research on smartwatch-based physical activity recognition, on the other hand, is still in its infancy. Through a large-scale exploratory study, this work aims to investigate the smartwatch-based physical activity recognition domain. A detailed analysis of various feature banks and classification methods are carried out to find the optimum system settings for the best performance of any smartwatch-based physical activity recognition system for both personal and impersonal models in real life scenarios. To further validate our hypothesis for both personal and impersonal models, we tested single subject out cross validation process for smartwatch-based physical activity recognition.

引用

页码：220 / 233

页数：14

共 50 条

[41] Improving physical activity recognition using a new deep learning architecture and post-processing techniques
Gil-Martin, Manuel
San-Segundo, Ruben
Fernandez-Martinez, Fernando
Ferreiros-Lopez, Javier
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2020, 92 (92)
[42] Predictors of Physical Activity for Preschool Children With and Without Disabilities From Socioeconomically Disadvantaged Settings
Brian, Ali
Taunton, Sally
Shortt, Chelsee
Pennell, Adam
Sacko, Ryan
ADAPTED PHYSICAL ACTIVITY QUARTERLY, 2019, 36 (01) : 77 - 90
[43] Physical Activity Recognition using Multiple Sensors Embedded in a Wearable Device
Nam, Yunyoung
Rho, Seungmin
Lee, Chulung
ACM TRANSACTIONS ON EMBEDDED COMPUTING SYSTEMS, 2013, 12 (02)
[44] Multi-sensor physical activity recognition in free-living
Ellis, Katherine
Kerr, Jacqueline
Godbole, Suneeta
Lanckriet, Gert
PROCEEDINGS OF THE 2014 ACM INTERNATIONAL JOINT CONFERENCE ON PERVASIVE AND UBIQUITOUS COMPUTING (UBICOMP'14 ADJUNCT), 2014, : 431 - 439
[45] Towards an Automatic Gait Recognition System using Activity Recognition (Wearable Based)
Bajrami, Gazmend
Derawi, Mohammad Omar
Bours, Patrick
PROCEEDINGS OF THE 2011 3RD INTERNATIONAL WORKSHOP ON SECURITY AND COMMUNICATION NETWORKS (IWSCN 2011), 2011, : 23 - 30
[46] Physical Activity Recognition Using a Single Tri-Axis Accelerometer
Lee, Mi-Hee
Kim, Jungchae
Kim, Kwangsoo
Lee, Inho
Jee, Sun Ha
Yoo, Sun Kook
WCECS 2009: WORLD CONGRESS ON ENGINEERING AND COMPUTER SCIENCE, VOLS I AND II, 2009, : 14 - +
[47] The Effect of Sensor Placement and Number on Physical Activity Recognition and Energy Expenditure Estimation in Older Adults: Validation Study
Davoudi, Anis
Mardini, Mamoun T.
Nelson, David
Albinali, Fahd
Ranka, Sanjay
Rashidi, Parisa
Manini, Todd M.
JMIR MHEALTH AND UHEALTH, 2021, 9 (05):
[48] Accelerometry-Based Motion Pattern Analysis for Physical Activity Recognition and Activity Level Assessment
Liu, Kai-Chun
Liu, Chung-Tse
Chen, Chao-Wei
Lin, Chih-Ching
Chan, Chia Tai
APPLIED SCIENCE AND PRECISION ENGINEERING INNOVATION, PTS 1 AND 2, 2014, 479-480 : 818 - 822
[49] Evaluation of a Smartphone-based Human Activity Recognition System in a Daily Living Environment
Lemaire, Edward D.
Tundo, Marco D.
Baddour, Natalie
JOVE-JOURNAL OF VISUALIZED EXPERIMENTS, 2015, (106):
[50] Towards Clustering of Mobile and Smartwatch Accelerometer Data for Physical Activity Recognition
Dobbins, Chelsea
Rawassizadeh, Reza
INFORMATICS-BASEL, 2018, 5 (02):

← 1 2 3 4 5 →