HMM-based human action recognition using multiview image sequences

Mohiuddin Ahmad, Seong Whan Lee

Research output: Chapter in Book/Report/Conference proceedingConference contribution

76 Citations (Scopus)

Abstract

In this paper, we present a novel method for human action recognition from any arbitrary view image sequence that uses the Cartesian component of optical flow velocity and human body silhouette feature vector information. We use principal component analysis (PCA) to reduce the higher dimensional silhouette feature space into lower dimensional feature space. The action region in an image frame represents Q-dimensional optical flow feature vector and R-dimensional silhouette feature vector. We represent each action using a set of hidden Markov models and we model each action for any viewing direction by using the combined (Q + R)-dimensional features at any instant of time. We perform experiments of the proposed method by using KU gesture database and manually captured data. Experimental results of different actions from any viewing direction are correctly classified by our method, which indicate the robustness of our view-independent method.

Original languageEnglish
Title of host publicationProceedings - International Conference on Pattern Recognition
Pages263-266
Number of pages4
Volume1
DOIs
Publication statusPublished - 2006 Dec 1
Event18th International Conference on Pattern Recognition, ICPR 2006 - Hong Kong, China
Duration: 2006 Aug 202006 Aug 24

Other

Other18th International Conference on Pattern Recognition, ICPR 2006
CountryChina
CityHong Kong
Period06/8/2006/8/24

Fingerprint

Optical flows
Hidden Markov models
Flow velocity
Principal component analysis
Experiments

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Vision and Pattern Recognition
  • Hardware and Architecture

Cite this

Ahmad, M., & Lee, S. W. (2006). HMM-based human action recognition using multiview image sequences. In Proceedings - International Conference on Pattern Recognition (Vol. 1, pp. 263-266). [1698883] https://doi.org/10.1109/ICPR.2006.630

HMM-based human action recognition using multiview image sequences. / Ahmad, Mohiuddin; Lee, Seong Whan.

Proceedings - International Conference on Pattern Recognition. Vol. 1 2006. p. 263-266 1698883.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ahmad, M & Lee, SW 2006, HMM-based human action recognition using multiview image sequences. in Proceedings - International Conference on Pattern Recognition. vol. 1, 1698883, pp. 263-266, 18th International Conference on Pattern Recognition, ICPR 2006, Hong Kong, China, 06/8/20. https://doi.org/10.1109/ICPR.2006.630
Ahmad M, Lee SW. HMM-based human action recognition using multiview image sequences. In Proceedings - International Conference on Pattern Recognition. Vol. 1. 2006. p. 263-266. 1698883 https://doi.org/10.1109/ICPR.2006.630
Ahmad, Mohiuddin ; Lee, Seong Whan. / HMM-based human action recognition using multiview image sequences. Proceedings - International Conference on Pattern Recognition. Vol. 1 2006. pp. 263-266
@inproceedings{f85880d041ac47d28e8bf3d924b23797,
title = "HMM-based human action recognition using multiview image sequences",
abstract = "In this paper, we present a novel method for human action recognition from any arbitrary view image sequence that uses the Cartesian component of optical flow velocity and human body silhouette feature vector information. We use principal component analysis (PCA) to reduce the higher dimensional silhouette feature space into lower dimensional feature space. The action region in an image frame represents Q-dimensional optical flow feature vector and R-dimensional silhouette feature vector. We represent each action using a set of hidden Markov models and we model each action for any viewing direction by using the combined (Q + R)-dimensional features at any instant of time. We perform experiments of the proposed method by using KU gesture database and manually captured data. Experimental results of different actions from any viewing direction are correctly classified by our method, which indicate the robustness of our view-independent method.",
author = "Mohiuddin Ahmad and Lee, {Seong Whan}",
year = "2006",
month = "12",
day = "1",
doi = "10.1109/ICPR.2006.630",
language = "English",
isbn = "0769525210",
volume = "1",
pages = "263--266",
booktitle = "Proceedings - International Conference on Pattern Recognition",

}

TY - GEN

T1 - HMM-based human action recognition using multiview image sequences

AU - Ahmad, Mohiuddin

AU - Lee, Seong Whan

PY - 2006/12/1

Y1 - 2006/12/1

N2 - In this paper, we present a novel method for human action recognition from any arbitrary view image sequence that uses the Cartesian component of optical flow velocity and human body silhouette feature vector information. We use principal component analysis (PCA) to reduce the higher dimensional silhouette feature space into lower dimensional feature space. The action region in an image frame represents Q-dimensional optical flow feature vector and R-dimensional silhouette feature vector. We represent each action using a set of hidden Markov models and we model each action for any viewing direction by using the combined (Q + R)-dimensional features at any instant of time. We perform experiments of the proposed method by using KU gesture database and manually captured data. Experimental results of different actions from any viewing direction are correctly classified by our method, which indicate the robustness of our view-independent method.

AB - In this paper, we present a novel method for human action recognition from any arbitrary view image sequence that uses the Cartesian component of optical flow velocity and human body silhouette feature vector information. We use principal component analysis (PCA) to reduce the higher dimensional silhouette feature space into lower dimensional feature space. The action region in an image frame represents Q-dimensional optical flow feature vector and R-dimensional silhouette feature vector. We represent each action using a set of hidden Markov models and we model each action for any viewing direction by using the combined (Q + R)-dimensional features at any instant of time. We perform experiments of the proposed method by using KU gesture database and manually captured data. Experimental results of different actions from any viewing direction are correctly classified by our method, which indicate the robustness of our view-independent method.

UR - http://www.scopus.com/inward/record.url?scp=34047226617&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34047226617&partnerID=8YFLogxK

U2 - 10.1109/ICPR.2006.630

DO - 10.1109/ICPR.2006.630

M3 - Conference contribution

SN - 0769525210

SN - 9780769525211

VL - 1

SP - 263

EP - 266

BT - Proceedings - International Conference on Pattern Recognition

ER -