Human activity recognition using symbolic sequences

María Mejía, Anh Tran, Paul R Cohen

Research output: Contribution to journalArticle

Abstract

Human activity recognition research is an active area in an early stage of development. We present two approaches to activity recognition based on symbolic representations of multivariate time series of joint locations in articulated skeletons.One approach uses pairwise alignment and nearest-neighbour classification, and the other uses spectrum kernels and SVMs as classifiers. We tested both approaches on three datasets derived from RGBD cameras (e.g., Microsoft Kinect) as well as ordinary video, and compared our results with those of other researchers.

Original languageEnglish (US)
Pages (from-to)12571-12581
Number of pages11
JournalARPN Journal of Engineering and Applied Sciences
Volume11
Issue number21
StatePublished - 2016

Fingerprint

Time series
Classifiers
Cameras

Keywords

  • Activity recognition
  • Artificial intelligence
  • Computer vision
  • Gesture
  • Machine learning
  • Pose

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Human activity recognition using symbolic sequences. / Mejía, María; Tran, Anh; Cohen, Paul R.

In: ARPN Journal of Engineering and Applied Sciences, Vol. 11, No. 21, 2016, p. 12571-12581.

Research output: Contribution to journalArticle

Mejía, María ; Tran, Anh ; Cohen, Paul R. / Human activity recognition using symbolic sequences. In: ARPN Journal of Engineering and Applied Sciences. 2016 ; Vol. 11, No. 21. pp. 12571-12581.
@article{2a100e1e1e714b1f8080903e34e10549,
title = "Human activity recognition using symbolic sequences",
abstract = "Human activity recognition research is an active area in an early stage of development. We present two approaches to activity recognition based on symbolic representations of multivariate time series of joint locations in articulated skeletons.One approach uses pairwise alignment and nearest-neighbour classification, and the other uses spectrum kernels and SVMs as classifiers. We tested both approaches on three datasets derived from RGBD cameras (e.g., Microsoft Kinect) as well as ordinary video, and compared our results with those of other researchers.",
keywords = "Activity recognition, Artificial intelligence, Computer vision, Gesture, Machine learning, Pose",
author = "Mar{\'i}a Mej{\'i}a and Anh Tran and Cohen, {Paul R}",
year = "2016",
language = "English (US)",
volume = "11",
pages = "12571--12581",
journal = "ARPN Journal of Engineering and Applied Sciences",
issn = "1819-6608",
publisher = "Asian Research Publishing Network (ARPN)",
number = "21",

}

TY - JOUR

T1 - Human activity recognition using symbolic sequences

AU - Mejía, María

AU - Tran, Anh

AU - Cohen, Paul R

PY - 2016

Y1 - 2016

N2 - Human activity recognition research is an active area in an early stage of development. We present two approaches to activity recognition based on symbolic representations of multivariate time series of joint locations in articulated skeletons.One approach uses pairwise alignment and nearest-neighbour classification, and the other uses spectrum kernels and SVMs as classifiers. We tested both approaches on three datasets derived from RGBD cameras (e.g., Microsoft Kinect) as well as ordinary video, and compared our results with those of other researchers.

AB - Human activity recognition research is an active area in an early stage of development. We present two approaches to activity recognition based on symbolic representations of multivariate time series of joint locations in articulated skeletons.One approach uses pairwise alignment and nearest-neighbour classification, and the other uses spectrum kernels and SVMs as classifiers. We tested both approaches on three datasets derived from RGBD cameras (e.g., Microsoft Kinect) as well as ordinary video, and compared our results with those of other researchers.

KW - Activity recognition

KW - Artificial intelligence

KW - Computer vision

KW - Gesture

KW - Machine learning

KW - Pose

UR - http://www.scopus.com/inward/record.url?scp=85006274788&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006274788&partnerID=8YFLogxK

M3 - Article

VL - 11

SP - 12571

EP - 12581

JO - ARPN Journal of Engineering and Applied Sciences

JF - ARPN Journal of Engineering and Applied Sciences

SN - 1819-6608

IS - 21

ER -