 | Veena D Singampalli
|
Research interests: Coarticulation modelling, Automatic Speech Recognition (ASR) based on articulatory models.
Hi, welcome to my webpage. I worked under the supervision of
Dr Philip Jackson during the course of my PhD (2005-2009). I was one of the team members of the DANSA project. My work focuses on using articulatory information to enhance the performance of models for speech recognition. My current work relvolves around exploring different feature spaces for representation of articulatory data and building a compact statistical model for coarticulation which would eventually be embedded in the architecture of a speech recogniser. I used articulatory data provided in the MOCHA-TIMIT corpus for my work.
Publications, posters and presentations
Academic journal
- PJB Jackson, V Singampalli (2009), "Statistical identification of critical articulators in the production of speech" In press, Speech Communication.
Conferences
- PJB Jackson, V Singampalli (2008). "Coarticulatory constraints determined by automatic identification from articulograph data". In Proc. ISSP 2008, Strasbourg.[pdf: paper]
- PJB Jackson, V Singampalli (2008). "Statistical identification of critical, dependent and redundant articulators". In Proc. Acoustics'08, Paris.
- V Singampalli, PJB Jackson (2007). "Statistical identification of critical, dependent and redundant articulators". In Proc. Interspeech 2007, 70-73, Antwerp, Belgium [pdf: presentation and paper].
Workshops
- V.Singampalli, PJB Jackson (2008)."Towards deriving compact and meaningful articulatory representations: an analysis of feature extraction techniques". In Proc. one-day meeting for Young Speech Researchers, University of Surrey, Guildford, 16th July 2008.
- V.Singampalli, PJB Jackson (2007). "A statistical technique for identifying articulatory roles in speech production". In Proc. one-day meeting for Young Speech Researchers, UCL, London, 12th April 2007 [pdf: abstract and poster].
- V Singampalli, PJB Jackson (2007). "Coarticulatory relations in a compact model of articulatory dynamics". In Proc. one-day meeting on Unified Models for Speech Recognition and Synthesis,* Birmingham, UK, p.3 (A)[pdf: abstract and presentation]
*
- V Singampalli, PJB Jackson (2005). "Statistical models to relate speech gestures to meaning". Presented at One-day meeting on Trajectory models for speech processing,* Edinburgh, UK.
*The current format of the audio file is '.dvf' and can be played in windows media player(Ver. 6.4 or higher) with this plugin.