Passive Wireless Sensing for Unsupervised Human Activity Recognition in Healthcare

Wenda Li; Bo Tan; Yangdi Xu; Robert Piechocki

Passive Wireless Sensing for Unsupervised Human Activity Recognition in Healthcare

Wenda Li, Bo Tan, Yangdi Xu, Robert Piechocki

University of Bristol

Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding › peer-review

17 Citations (Scopus)

Abstract

Physical activity classification is an important tool for various applications such as activity of daily living (ADL) recognition and fall detection. Additionally, the non-contact nature of radar systems provides minimally invasive sensing platform. Doppler-based radar has been used for activity classification in the past. However, most of these studies considered supervised classification which requires labeled training data sets. In this paper, we propose a novel procedure of using micro Doppler radar for unsupervised classification with Hidden Markov Models (HMM). A low-complexity time alignment method for capturing activity is developed and an Elbow test has been adopted for model selection. Test results confirm the efficacy of the selected feature set and the proposed methodology. The results prove the proposed system can deliver a very good performance in ADL recognition tasks.

Original language	English
Title of host publication	2017 13th International Wireless Communications and Mobile Computing Conference (IWCMC)
Publisher	IEEE
Pages	1528-1533
ISBN (Electronic)	978-1-5090-4372-9
ISBN (Print)	978-1-5090-4373-6
Publication status	Published - 20 Jul 2017
Event	International Wireless Communications and Mobile Computing Conference - Holiday Inn, Valencia, Spain Duration: 26 Jun 2017 → 30 Jun 2017 Conference number: 13 http://iwcmc.org/2017/ (Link to the conference website)

Conference

Conference	International Wireless Communications and Mobile Computing Conference
Abbreviated title	IWCMC2017
Country/Territory	Spain
City	Valencia
Period	26/06/17 → 30/06/17
Internet address	http://iwcmc.org/2017/ (Link to the conference website)

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Cite this

@inproceedings{6bb2c2ac05c3423eaeb7f08124355750,

title = "Passive Wireless Sensing for Unsupervised Human Activity Recognition in Healthcare",

abstract = "Physical activity classification is an important tool for various applications such as activity of daily living (ADL) recognition and fall detection. Additionally, the non-contact nature of radar systems provides minimally invasive sensing platform. Doppler-based radar has been used for activity classification in the past. However, most of these studies considered supervised classification which requires labeled training data sets. In this paper, we propose a novel procedure of using micro Doppler radar for unsupervised classification with Hidden Markov Models (HMM). A low-complexity time alignment method for capturing activity is developed and an Elbow test has been adopted for model selection. Test results confirm the efficacy of the selected feature set and the proposed methodology. The results prove the proposed system can deliver a very good performance in ADL recognition tasks.",

author = "Wenda Li and Bo Tan and Yangdi Xu and Robert Piechocki",

year = "2017",

month = jul,

day = "20",

language = "English",

isbn = "978-1-5090-4373-6",

pages = "1528--1533",

booktitle = "2017 13th International Wireless Communications and Mobile Computing Conference (IWCMC)",

publisher = "IEEE",

address = "United States",

note = "International Wireless Communications and Mobile Computing Conference, IWCMC2017 ; Conference date: 26-06-2017 Through 30-06-2017",

url = "http://iwcmc.org/2017/",

}

TY - GEN

T1 - Passive Wireless Sensing for Unsupervised Human Activity Recognition in Healthcare

AU - Li, Wenda

AU - Tan, Bo

AU - Xu, Yangdi

AU - Piechocki, Robert

N1 - Conference code: 13

PY - 2017/7/20

Y1 - 2017/7/20

N2 - Physical activity classification is an important tool for various applications such as activity of daily living (ADL) recognition and fall detection. Additionally, the non-contact nature of radar systems provides minimally invasive sensing platform. Doppler-based radar has been used for activity classification in the past. However, most of these studies considered supervised classification which requires labeled training data sets. In this paper, we propose a novel procedure of using micro Doppler radar for unsupervised classification with Hidden Markov Models (HMM). A low-complexity time alignment method for capturing activity is developed and an Elbow test has been adopted for model selection. Test results confirm the efficacy of the selected feature set and the proposed methodology. The results prove the proposed system can deliver a very good performance in ADL recognition tasks.

AB - Physical activity classification is an important tool for various applications such as activity of daily living (ADL) recognition and fall detection. Additionally, the non-contact nature of radar systems provides minimally invasive sensing platform. Doppler-based radar has been used for activity classification in the past. However, most of these studies considered supervised classification which requires labeled training data sets. In this paper, we propose a novel procedure of using micro Doppler radar for unsupervised classification with Hidden Markov Models (HMM). A low-complexity time alignment method for capturing activity is developed and an Elbow test has been adopted for model selection. Test results confirm the efficacy of the selected feature set and the proposed methodology. The results prove the proposed system can deliver a very good performance in ADL recognition tasks.

UR - http://iwcmc.org/2017/

M3 - Conference proceeding

SN - 978-1-5090-4373-6

SP - 1528

EP - 1533

BT - 2017 13th International Wireless Communications and Mobile Computing Conference (IWCMC)

PB - IEEE

T2 - International Wireless Communications and Mobile Computing Conference

Y2 - 26 June 2017 through 30 June 2017

ER -

Passive Wireless Sensing for Unsupervised Human Activity Recognition in Healthcare

Abstract

Conference

UN SDGs

Other files and links

Fingerprint

SPHERE: A sensor platform for healthcare in a residential environment

Cite this