Dynamic gesture recognition with laban movement analysis and hidden markov models

Arthur Truong, Titus Zaharia

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we propose a new approach for gesture recognition based upon the quantification of Laban Movement Analysis (LMA) concepts. The resulting body features are used to build a dictionary of key poses. Then, a soft abignment method is applied to the gesture sequences to obtain a gesture representation. The abignment results are used as input in a Hidden Markov Models (HMM) scheme for dynamic gesture recognition purposes. The proposed approach achieves high recognition rates (more than 92% for certain categories of gestures), when tested and evaluated on a corpus including 11 different actions. The high recognition rates obtained on two other datasets (Microsoft Gesture dataset [1] and UTKinect-Human Detection dataset [2]) show the relevance of our method.

Original languageEnglish
Title of host publicationProceedings of the 33rd Computer Graphics International Conference, CGI 2016
PublisherAssociation for Computing Machinery
Pages21-24
Number of pages4
ISBN (Electronic)9781450341233
DOIs
Publication statusPublished - 28 Jun 2016
Externally publishedYes
Event33rd Computer Graphics International Conference, CGI 2016 - Heraklion, Greece
Duration: 28 Jun 20161 Jul 2016

Publication series

NameACM International Conference Proceeding Series
Volume28-June-01-July-2016

Conference

Conference33rd Computer Graphics International Conference, CGI 2016
Country/TerritoryGreece
CityHeraklion
Period28/06/161/07/16

Keywords

  • Body motion descriptors
  • Gesture recognition
  • Hidden Markov Model
  • Laban movement analysis
  • Soft abignment

Fingerprint

Dive into the research topics of 'Dynamic gesture recognition with laban movement analysis and hidden markov models'. Together they form a unique fingerprint.

Cite this