Please use this identifier to cite or link to this item: http://scholarbank.nus.edu.sg/handle/10635/43226
Title: A new probabilistic model for recognizing signs with systematic modulations
Authors: Ong, S.C.W. 
Ranganath, S. 
Issue Date: 2007
Source: Ong, S.C.W.,Ranganath, S. (2007). A new probabilistic model for recognizing signs with systematic modulations. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 4778 LNCS : 16-30. ScholarBank@NUS Repository.
Abstract: This paper addresses an aspect of sign language (SL) recognition that has largely been overlooked in previous work and yet is integral to signed communication. It is the most comprehensive work to-date on recognizing complex variations in sign appearances due to grammatical processes (inflections) which systematically modulate the temporal and spatial dimensions of a root sign word to convey information in addition to lexical meaning. We propose a novel dynamic Bayesian network the Multichannel Hierarchical Hidden Markov Model (MH-HMM)- as a modelling and recognition framework for continuously signed sentences that include modulated signs. This models the hierarchical, sequential and parallel organization in signing while requiring synchronization between parallel data streams at sign boundaries. Experimental results using particle filtering for decoding demonstrate the feasibility of using the MH-HMM for recognizing inflected signs in continuous sentences. © Springer-Verlag Berlin Heidelberg 2007.
Source Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
URI: http://scholarbank.nus.edu.sg/handle/10635/43226
ISBN: 9783540756897
ISSN: 03029743
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

Page view(s)

49
checked on Dec 9, 2017

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.