Sign Language Recognition using Sub-Units

Cooper, HM, Ong, EJ, Pugeault, N and Bowden, R (2012) Sign Language Recognition using Sub-Units Journal of Machine Learning Research, 13. pp. 2205-2231.

Preview	Text Cooper_JMLR_2012.pdf - ["content_typename_Submitted version (pre-print)" not defined] Available under License : See the attached licence file. Download (2MB) | Preview
Preview	PDF (licence) SRI_deposit_agreement.pdf Available under License : See the attached licence file. Download (33kB) | Preview

Abstract

This paper discusses sign language recognition using linguistic sub-units. It presents three types of sub-units for consideration; those learnt from appearance data as well as those inferred from both 2D or 3D tracking data. These sub-units are then combined using a sign level classifier; here, two options are presented. The first uses Markov Models to encode the temporal changes between sub-units. The second makes use of Sequential Pattern Boosting to apply discriminative feature selection at the same time as encoding temporal information. This approach is more robust to noise and performs well in signer independent tests, improving results from the 54% achieved by the Markov Chains to 76%.

Item Type:	Article
Divisions :	Faculty of Engineering and Physical Sciences > Electronic Engineering > Centre for Vision Speech and Signal Processing
Authors :	Name Email ORCID Cooper, HM UNSPECIFIED UNSPECIFIED Ong, EJ UNSPECIFIED UNSPECIFIED Pugeault, N UNSPECIFIED UNSPECIFIED Bowden, R UNSPECIFIED UNSPECIFIED
Date :	25 July 2012
Contributors :	Contribution Name Email ORCID http://www.loc.gov/loc.terms/relators/EDT Guyon, I UNSPECIFIED UNSPECIFIED http://www.loc.gov/loc.terms/relators/EDT Athitsos, V UNSPECIFIED UNSPECIFIED
Uncontrolled Keywords :	Sign Language Recognition
Additional Information :	Copyright 2012 The Author(s) Published by Journal of Machine Learning Research
Depositing User :	Symplectic Elements
Date Deposited :	17 Nov 2015 18:26
Last Modified :	17 Nov 2015 18:26
URI:	http://epubs.surrey.ac.uk/id/eprint/808972

Actions (login required)

View Item

Downloads