Ph.D. Research Studentship Opportunity

[Sue Denham <S.Denham@xxxxxxxxxxxxxx>]

Modelling human speech perception

Internal advisors: Dr Susan Denham, Dr Jeremy Goslin and Dr Caroline Floccia (1School of Psychology, University of Plymouth)
External advisor: Dr Steven Greenberg (Silicon Speech, USA)

Applications are invited for a University-funded studentship to start in April 2010

Although artificial speech recognition systems have improved considerably over the years, their performance still falls far short of human abilities, and their robustness in the face of changing conditions is limited. In contrast, humans and other animals are able to adapt, seemingly effortlessly, to different listening environments, and are able to communicate effectively with one another in many different circumstances. In this project we aim to investigate a novel theoretical model of human speech perception based on cortical oscillators. We take as our starting point the observation that natural communication sounds contain temporal patterns or regularities evident at many different times scales (Winkler, Denham et al. 2009). The proposal is that the speech message can be extracted through adaptation of a hierarchically organised system of neural oscillators to the characteristic multi-scale temporal patterns present in the speech of the target speaker, and that by doing so extraneous interfering sounds can be simultaneously rejected.

This proposal will be tested using electrophysiological measurements of listeners attending to speech in different background sounds, analyzing activity at various pre-lexical and lexical processing levels (e.g. (Goslin, Grainger et al. 2006)), for application in the development of a biologically inspired computational model of human speech perception.

We are looking for a highly qualified and motivated student with a strong interest in auditory perception, sounds and speech perception. You will join a well-established research environment, and work alongside the brain-technology team which is currently funded by a multi-centre European project SCANDLE (http://www.scandle.eu), and a new joint British ESRC/ French ANR project, RECONVO (investigating multi-lingual speech development).
Requirement: Knowledge of experimental methods and/or programming experience with a high level language; Desirable: Knowledge of signal processing techniques, models of auditory perception, electrophysiological techniques.
Candidates should have a first or upper second class honours degree in an area related to Cognitive Neuroscience (Computer Science, Maths, Physics, Electrical Engineering, Neuroscience, or Psychology). Applicants with a relevant MSc or MRes are particularly welcome. The studentship will provide a fully funded full-time PhD post for three years, with a stipend of approximately £13,290 per annum. The position is open to UK citizens and EU citizens with appropriate qualification who have been resident or studied in the UK for three years.

For informal queries please contact: Dr Susan Denham (sdenham@xxxxxxxxxxxxxx<mailto:sdenham@xxxxxxxxxxxxxx>).
For an application form and full details on how to apply, please visit www.plymouth.ac.uk/pghowtoapply<http://www.plymouth.ac.uk/pghowtoapply>
Applicants should send a completed application form along with the following documentation to The University of Plymouth, Postgraduate Admissions Office, Hepworth House, Drake Circus, Plymouth, PL4 8AA – United Kingdom

• Two references in envelopes signed across their seals
• Copies of transcripts and certificates
• If English is not your first language, evidence that you meet our English Language requirements (www.plymouth.ac.uk/elr<http://www.plymouth.ac.uk/elr> )
• CV
• Ethnic and Disability Monitoring Form

Closing Date: 5PM, Monday 15 February 2010.

Interviews will be held at the end of February 2010, with a proposed start date of 1 April 2010.

References
Goslin, J., J. Grainger, et al. (2006). "Syllable frequency effects in French visual word recognition: an ERP study." Brain Res 1115(1): 121-34.
Winkler, I., S. L. Denham, et al. (2009). "Modeling the auditory scene: predictive regularity representations and perceptual objects." Trends Cogn Sci 13(12): 532-40.