16.640, Support: Comp Ling: PhD Student, University of Edinburgh

LINGUIST List linguist at linguistlist.org
Fri Mar 4 16:07:37 UTC 2005


LINGUIST List: Vol-16-640. Fri Mar 04 2005. ISSN: 1068 - 4875.

Subject: 16.640, Support: Comp Ling: PhD Student, University of Edinburgh

Moderators: Anthony Aristar, Wayne State U <aristar at linguistlist.org>
            Helen Aristar-Dry, Eastern Michigan U <hdry at linguistlist.org>

Reviews (reviews at linguistlist.org)
        Sheila Collberg, U of Arizona
        Terry Langendoen, U of Arizona

Homepage: http://linguistlist.org/

The LINGUIST List is funded by Eastern Michigan University, Wayne
State University, and donations from subscribers and publishers.

Editor for this issue: Jessica Boynton <jessica at linguistlist.org>
================================================================

To post to LINGUIST, use our convenient web form at
http://linguistlist.org/LL/posttolinguist.html.


===========================Directory==============================

1)
Date: 04-Mar-2005
From: Simon King < Simon.King at ed.ac.uk >
Subject: Comp Ling: PhD Student, University of Edinburgh, Edinburgh, UK

	
-------------------------Message 1 ----------------------------------
Date: Fri, 04 Mar 2005 11:04:56
From: Simon King < Simon.King at ed.ac.uk >
Subject: Comp Ling: PhD Student, University of Edinburgh, Edinburgh, UK


University or Organization: Centre for Speech Technology Research
Job Rank: PhD
Specialty Areas: Computational Linguistics
Dynamic Bayesian networks for speech recognition
------------------------------------------------

Hidden Markov models (HMMs) are the current model of choice for automatic
speech recognition (ASR). Although they are seemingly very simple models,
in fact they require a complex system of context-dependent models,
parameter sharing and adaptation algorithms to achieve the best performance.

HMMs are a member of a wider family of models - dynamic Bayesian networks
(DBNs). There are an infinite variety of other DBNs waiting to be tried for
ASR. DBNs can be formulated to reflect our understanding of the speech
signal; one example of this would be multi-streamed DBNs (such as the
factorial HMM) in which the factors have explicit linguistic
interpretations - the factors might represent aspects of the speech
production process.

Dependencies can be introduced between these hidden factors, creating ever
richer model structures (at the cost of increased computational
complexity). The goal is to find model structures that improve recognition
accuracy whilst remaining computationally feasible.

We have already started exploring various forms of DBN, but there is still
a lot of scope for exciting and original research in this area.  The
toolkits and computer power are now available to work with models that were
intractable until recently. It may be that some of the techniques developed
for HMMs can be transferred to DBNs, or we could build things like
parameter tying, pronunciation variation, language modeling and adaptation
into the model structure itself.

Full funding (fees plus living allowance) for eligible UK or EU students is
available.

Address for Applications:
	Simon King
Simon.King at ed.ac.uk

	----------------
	Website: http://www.cstr.ed.ac.uk/opportunities/phd.html





-----------------------------------------------------------
LINGUIST List: Vol-16-640	

	



More information about the LINGUIST mailing list