ASR - Articulatory Speech Recognition
dc.contributor.author
Frankel, Joe
en
dc.contributor.author
King, Simon
en
dc.date.accessioned
2006-05-19T15:45:37Z
dc.date.available
2006-05-19T15:45:37Z
dc.date.issued
2001
dc.description.abstract
We propose that using a continuous trajectory model to describe an articulatory-based feature set will address some of the shortcomings inherent in the hidden Markov model (HMM) as a model for speech recognition. The articulatory parameters allow us to explicitly model effects such as co-articulation and assimilation. A linear dynamic model (LDM) is used to capture the characteristics of each segment type. These models are well suited to describing smoothly varying, continuous, yet noisy trajectories, such as we find present in speech data. Experimentation has been based on data for a single speaker from the MOCHA corpus. This consists of parallel acoustic and recorded articulatory parameters for 460 TIMIT sentences. We report the results of classification and recognition tasks using both real and recovered articulatory parameters, on their own and in conjunction with acoustic features.
en
dc.format.extent
63047 bytes
en
dc.format.extent
52793 bytes
en
dc.format.mimetype
application/postscript
en
dc.format.mimetype
application/pdf
en
dc.identifier.citation
Frankel, Joe / King, Simon (2001): "ASR - articulatory speech recognition", In EUROSPEECH-2001, 599-602.
dc.identifier.uri
http://www.isca-speech.org/archive/eurospeech_2001/index.html
dc.identifier.uri
http://hdl.handle.net/1842/1140
dc.language.iso
en
dc.publisher
International Speech Communication Association
en
dc.subject
Articulatory Speech Recognition
en
dc.subject
hidden Markov model
en
dc.title
ASR - Articulatory Speech Recognition
en
dc.type
Conference Paper
en
This item appears in the following Collection(s)