Model-level data-driven sub-units for signs in videos of continuous sign language

Theodorakis, S; Pitsikalis, V; Maragos, P

dc.contributor.author	Theodorakis, S	en
dc.contributor.author	Pitsikalis, V	en
dc.contributor.author	Maragos, P	en
dc.date.accessioned	2014-03-01T02:46:53Z
dc.date.available	2014-03-01T02:46:53Z
dc.date.issued	2010	en
dc.identifier.issn	15206149	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/32910
dc.subject	HMM	en
dc.subject	Sign language	en
dc.subject	Subunit modeling	en
dc.subject.other	American sign language	en
dc.subject.other	Boston University	en
dc.subject.other	Data-driven	en
dc.subject.other	Hand positions	en
dc.subject.other	Hier-archical clustering	en
dc.subject.other	HMM	en
dc.subject.other	Phonetic information	en
dc.subject.other	Qualitative analysis	en
dc.subject.other	Region-based	en
dc.subject.other	Sign language	en
dc.subject.other	Sub-units	en
dc.subject.other	Subunit modeling	en
dc.subject.other	Time segmentation	en
dc.subject.other	Unit constructions	en
dc.subject.other	Visual feature	en
dc.subject.other	Visual-processing	en
dc.subject.other	Hidden Markov models	en
dc.subject.other	Linguistics	en
dc.subject.other	Quality control	en
dc.subject.other	Signal processing	en
dc.subject.other	Cluster analysis	en
dc.title	Model-level data-driven sub-units for signs in videos of continuous sign language	en
heal.type	conferenceItem	en
heal.identifier.primary	10.1109/ICASSP.2010.5495875	en
heal.identifier.secondary	http://dx.doi.org/10.1109/ICASSP.2010.5495875	en
heal.identifier.secondary	5495875	en
heal.publicationDate	2010	en
heal.abstract	We investigate the issue of sign language automatic phonetic sub-unit modeling, that is completely data driven and without any prior phonetic information. A first step of visual processing leads to simple and effective region-based visual features. Prior to the sub-unit modeling we propose to employ a pronunciation clustering step with respect to each sign. Afterwards, for each sign and pronunciation group we find the time segmentation at the hidden Markov model (HMM) level. The models employed refer to movements as a sequence of dominant hand positions. The constructed segments are exploited explicitly at the model level via hierarchical clustering of HMMs and lead to the data-driven movement sub-unit construction. The constructed movement sub-units are evaluated in qualitative analysis experiments on data from the Boston University (BU)-400 American Sign Language corpus showing promising results. ©2010 IEEE.	en
heal.journalName	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings	en
dc.identifier.doi	10.1109/ICASSP.2010.5495875	en
dc.identifier.spage	2262	en
dc.identifier.epage	2265	en