File(s) under permanent embargo
Classifying human actions using an incomplete real-time pose skeleton
chapter
posted on 2004-01-01, 00:00 authored by P Peursum, H Bui, Svetha VenkateshSvetha Venkatesh, G WestCurrently, most human action recognition systems are trained with feature sets that have no missing data. Unfortunately, the use of human pose estimation models to provide more descriptive features also entails an increased sensitivity to occlusions, meaning that incomplete feature information will be unavoidable for realistic scenarios. To address this, our approach is to shift the responsibility for dealing with occluded pose data away from the pose estimator and onto the action classifier. This allows the use of a simple, real-time pose estimation (stick-figure) that does not estimate the positions of limbs it cannot find quickly. The system tracks people via background subtraction and extracts the (possibly incomplete) pose skeleton from their silhouette. Hidden Markov Models modified to handle missing data are then used to successfully classify several human actions using the incomplete pose features.
History
Title of book
PRICAI 2004 : trends in artificial intelligence : 8th Pacific Rim International Conference on Artificial Intelligence, Auckland, New Zealand, August 9-13, 2004 : proceedingsSeries
Lecture notes in artificial intelligence ; 3157Chapter number
119Pagination
971 - 972Publisher
Springer-VerlagPlace of publication
Berlin, GermanyPublisher DOI
ISSN
0302-9743ISBN-13
9783540228172ISBN-10
3540228179Language
engPublication classification
B1.1 Book chapterCopyright notice
2004, Springer-Verlag Berlin HeidelbergExtent
142Editor/Contributor(s)
C Zhang, H Guesgen, W YeapUsage metrics
Categories
No categories selectedKeywords
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC