Recognizing human activities from smartphone sensors using hierarchical continuous hidden Markov models

Charissa Ann Ronao, Sung Bae Cho

Research output: Contribution to journalArticle

24 Citations (Scopus)

Abstract

Human activity recognition has been gaining more and more attention from researchers in recent years, particularly with the use of widespread and commercially available devices such as smartphones. However, most of the existing works focus on discriminative classifiers while neglecting the inherent time-series and continuous characteristics of sensor data. To address this, we propose a two-stage continuous hidden Markov model framework, which also takes advantage of the innate hierarchical structure of basic activities. This kind of system architecture not only enables the use of different feature subsets on different subclasses, which effectively reduces feature computation overhead, but also allows for varying number of states and iterations. Experiments show that the hierarchical structure dramatically increases classification performance. We analyze the behavior of the accelerometer and gyroscope signals for each activity through graphs, and with added fine tuning of states and training iterations, the proposed method is able to achieve an overall accuracy of up to 93.18%, which is the best performance among the state-of-the-art classifiers for the problem at hand.

Original languageEnglish
JournalInternational Journal of Distributed Sensor Networks
Volume13
Issue number1
DOIs
Publication statusPublished - 2017 Jan 1

Fingerprint

Smartphones
Hidden Markov models
Classifiers
Gyroscopes
Sensors
Accelerometers
Time series
Tuning
Experiments

All Science Journal Classification (ASJC) codes

  • Engineering(all)
  • Computer Networks and Communications

Cite this

@article{6be18d9eb2364a87b74e259a775e88a1,
title = "Recognizing human activities from smartphone sensors using hierarchical continuous hidden Markov models",
abstract = "Human activity recognition has been gaining more and more attention from researchers in recent years, particularly with the use of widespread and commercially available devices such as smartphones. However, most of the existing works focus on discriminative classifiers while neglecting the inherent time-series and continuous characteristics of sensor data. To address this, we propose a two-stage continuous hidden Markov model framework, which also takes advantage of the innate hierarchical structure of basic activities. This kind of system architecture not only enables the use of different feature subsets on different subclasses, which effectively reduces feature computation overhead, but also allows for varying number of states and iterations. Experiments show that the hierarchical structure dramatically increases classification performance. We analyze the behavior of the accelerometer and gyroscope signals for each activity through graphs, and with added fine tuning of states and training iterations, the proposed method is able to achieve an overall accuracy of up to 93.18{\%}, which is the best performance among the state-of-the-art classifiers for the problem at hand.",
author = "Ronao, {Charissa Ann} and Cho, {Sung Bae}",
year = "2017",
month = "1",
day = "1",
doi = "10.1177/1550147716683687",
language = "English",
volume = "13",
journal = "International Journal of Distributed Sensor Networks",
issn = "1550-1329",
publisher = "Hindawi Publishing Corporation",
number = "1",

}

TY - JOUR

T1 - Recognizing human activities from smartphone sensors using hierarchical continuous hidden Markov models

AU - Ronao, Charissa Ann

AU - Cho, Sung Bae

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Human activity recognition has been gaining more and more attention from researchers in recent years, particularly with the use of widespread and commercially available devices such as smartphones. However, most of the existing works focus on discriminative classifiers while neglecting the inherent time-series and continuous characteristics of sensor data. To address this, we propose a two-stage continuous hidden Markov model framework, which also takes advantage of the innate hierarchical structure of basic activities. This kind of system architecture not only enables the use of different feature subsets on different subclasses, which effectively reduces feature computation overhead, but also allows for varying number of states and iterations. Experiments show that the hierarchical structure dramatically increases classification performance. We analyze the behavior of the accelerometer and gyroscope signals for each activity through graphs, and with added fine tuning of states and training iterations, the proposed method is able to achieve an overall accuracy of up to 93.18%, which is the best performance among the state-of-the-art classifiers for the problem at hand.

AB - Human activity recognition has been gaining more and more attention from researchers in recent years, particularly with the use of widespread and commercially available devices such as smartphones. However, most of the existing works focus on discriminative classifiers while neglecting the inherent time-series and continuous characteristics of sensor data. To address this, we propose a two-stage continuous hidden Markov model framework, which also takes advantage of the innate hierarchical structure of basic activities. This kind of system architecture not only enables the use of different feature subsets on different subclasses, which effectively reduces feature computation overhead, but also allows for varying number of states and iterations. Experiments show that the hierarchical structure dramatically increases classification performance. We analyze the behavior of the accelerometer and gyroscope signals for each activity through graphs, and with added fine tuning of states and training iterations, the proposed method is able to achieve an overall accuracy of up to 93.18%, which is the best performance among the state-of-the-art classifiers for the problem at hand.

UR - http://www.scopus.com/inward/record.url?scp=85011854067&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85011854067&partnerID=8YFLogxK

U2 - 10.1177/1550147716683687

DO - 10.1177/1550147716683687

M3 - Article

AN - SCOPUS:85011854067

VL - 13

JO - International Journal of Distributed Sensor Networks

JF - International Journal of Distributed Sensor Networks

SN - 1550-1329

IS - 1

ER -