Record Details

Landmark based recognition of stops: acoustic attributes versus smoothed spectra

DSpace at IIT Bombay

View Archive Info
 
 
Field Value
 
Title Landmark based recognition of stops: acoustic attributes versus smoothed spectra
 
Creator KARJIGI, V
RAO, P
 
Subject consonants
features
articulation
invariance
place
landmark based recognition
unvoiced stops
acoustic attributes
burst
vocalic transition
 
Description Landmark based recognition of unvoiced word-initial stops is investigated. The relative effectiveness of acoustic-phonetic attributes versus more global spectral shape features is experimentally evaluated for four-way place classification of unvoiced, unaspirated stops. Various feature sets derived from the burst and vocalic transition regions of word initial consonants are compared via GMM based classification under speaker, gender, and vowel-context variability. While a set of acoustic attributes derived from the burst shows the best invariance to vowel context, it is found that global spectral shape features provide the most robust representation of the vocalic transition region by overcoming the problem of errors in explicit formant tracking. A combination of features from the burst and vocalic regions was superior to burst-only cues, but still far from the near perfect identification achieved in human perception.
 
Publisher ISCA-INST SPEECH COMMUNICATION ASSOC
 
Date 2011-10-26T02:58:22Z
2011-12-15T09:11:29Z
2011-10-26T02:58:22Z
2011-12-15T09:11:29Z
2008
 
Type Proceedings Paper
 
Identifier INTERSPEECH 2008: 9TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2008, VOLS 1-5,1550-1553
978-1-61567-378-0
http://dspace.library.iitb.ac.in/xmlui/handle/10054/15889
http://hdl.handle.net/100/2099
 
Source 9th Annual Conference of the International-Speech-Communication-Association (INTERSPEECH 2008),Brisbane, AUSTRALIA,SEP 22-26, 2008
 
Language English