Multi-user Activity Recognition in a Smart Home

Liang Wang, Tao Gu, Xianping Tao, Hanhua Chen, Jian Lu

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

Resumé

The advances of wearable sensors and wireless networks offer many opportunities to recognize
human activities from sensor readings in pervasive computing. Existing work so far focus
mainly on recognizing activities of a single user in a home environment. However, there
are typically multiple inhabitants in a real home and they often perform activities together.
In this paper, we investigate the problem of recognizing multi-user activities using wearable
sensors in a home setting. We develop a multi-modal, wearable sensor platform to collect
sensor data for multiple users, and study two temporal probabilistic models—Coupled Hidden
Markov Model (CHMM) and Factorial Conditional Random Field (FCRF)—to model
interacting processes in a sensor-based, multi-user scenario. We conduct a real-world trace
collection done by two subjects over two weeks, and evaluate these two models through
our experimental studies. Our experimental results show that we achieve an accuracy of
96.41% with CHMM and an accuracy of 87.93% with FCRF, respectively, for recognizing
multi-user activities.
OriginalsprogEngelsk
TidsskriftActivity Recognition in Pervasive Intelligent Environments
Vol/bind4
Sider (fra-til)59-81
DOI
StatusUdgivet - 2010

Fingeraftryk

Sensors
Ubiquitous computing
Sensor networks
Wireless networks
Wearable sensors

Citer dette

Wang, Liang ; Gu, Tao ; Tao, Xianping ; Chen, Hanhua ; Lu, Jian . / Multi-user Activity Recognition in a Smart Home. I: Activity Recognition in Pervasive Intelligent Environments. 2010 ; Bind 4. s. 59-81.
@article{b5d21ab5aebd4d0694d202e0581c3980,
title = "Multi-user Activity Recognition in a Smart Home",
abstract = "The advances of wearable sensors and wireless networks offer many opportunities to recognizehuman activities from sensor readings in pervasive computing. Existing work so far focusmainly on recognizing activities of a single user in a home environment. However, thereare typically multiple inhabitants in a real home and they often perform activities together.In this paper, we investigate the problem of recognizing multi-user activities using wearablesensors in a home setting. We develop a multi-modal, wearable sensor platform to collectsensor data for multiple users, and study two temporal probabilistic models—Coupled HiddenMarkov Model (CHMM) and Factorial Conditional Random Field (FCRF)—to modelinteracting processes in a sensor-based, multi-user scenario. We conduct a real-world tracecollection done by two subjects over two weeks, and evaluate these two models throughour experimental studies. Our experimental results show that we achieve an accuracy of96.41{\%} with CHMM and an accuracy of 87.93{\%} with FCRF, respectively, for recognizingmulti-user activities.",
author = "Liang Wang and Tao Gu and Xianping Tao and Hanhua Chen and Jian Lu",
year = "2010",
doi = "10.2991/978-94-91216-05-3_3",
language = "English",
volume = "4",
pages = "59--81",
journal = "Activity Recognition in Pervasive Intelligent Environments",

}

Multi-user Activity Recognition in a Smart Home. / Wang, Liang ; Gu, Tao; Tao, Xianping ; Chen, Hanhua ; Lu, Jian .

I: Activity Recognition in Pervasive Intelligent Environments, Bind 4, 2010, s. 59-81.

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

TY - JOUR

T1 - Multi-user Activity Recognition in a Smart Home

AU - Wang, Liang

AU - Gu, Tao

AU - Tao, Xianping

AU - Chen, Hanhua

AU - Lu, Jian

PY - 2010

Y1 - 2010

N2 - The advances of wearable sensors and wireless networks offer many opportunities to recognizehuman activities from sensor readings in pervasive computing. Existing work so far focusmainly on recognizing activities of a single user in a home environment. However, thereare typically multiple inhabitants in a real home and they often perform activities together.In this paper, we investigate the problem of recognizing multi-user activities using wearablesensors in a home setting. We develop a multi-modal, wearable sensor platform to collectsensor data for multiple users, and study two temporal probabilistic models—Coupled HiddenMarkov Model (CHMM) and Factorial Conditional Random Field (FCRF)—to modelinteracting processes in a sensor-based, multi-user scenario. We conduct a real-world tracecollection done by two subjects over two weeks, and evaluate these two models throughour experimental studies. Our experimental results show that we achieve an accuracy of96.41% with CHMM and an accuracy of 87.93% with FCRF, respectively, for recognizingmulti-user activities.

AB - The advances of wearable sensors and wireless networks offer many opportunities to recognizehuman activities from sensor readings in pervasive computing. Existing work so far focusmainly on recognizing activities of a single user in a home environment. However, thereare typically multiple inhabitants in a real home and they often perform activities together.In this paper, we investigate the problem of recognizing multi-user activities using wearablesensors in a home setting. We develop a multi-modal, wearable sensor platform to collectsensor data for multiple users, and study two temporal probabilistic models—Coupled HiddenMarkov Model (CHMM) and Factorial Conditional Random Field (FCRF)—to modelinteracting processes in a sensor-based, multi-user scenario. We conduct a real-world tracecollection done by two subjects over two weeks, and evaluate these two models throughour experimental studies. Our experimental results show that we achieve an accuracy of96.41% with CHMM and an accuracy of 87.93% with FCRF, respectively, for recognizingmulti-user activities.

U2 - 10.2991/978-94-91216-05-3_3

DO - 10.2991/978-94-91216-05-3_3

M3 - Journal article

VL - 4

SP - 59

EP - 81

JO - Activity Recognition in Pervasive Intelligent Environments

JF - Activity Recognition in Pervasive Intelligent Environments

ER -