SPHEAR TMR NETWORK - SPeech, HEaring and Recognition

This is the home page of the SPHEAR TMR network.

SPHEAR is funded by the EC Transfer and Mobility of Researchers programme. The contract began on 1st March 1998 and its duration has now been extended to 60 months, till 28th February 2003.
A network developed from SPHEAR has been funded by the framework 5 'Inproving Human Potential' Programme. See HOARSE

SPHEAR Objectives

The twin goals of the network are to achieve better understanding of auditory information processing and to deploy this understanding in automatic speech recognition for adverse conditions.

SPHEAR  researches new methodologies for Automatic Speech Recognition based on missing-data theory and multiple classification streams. It will provide a link between ASR and models of hearing, particularly work in Computational Auditory Scene Analysis, which it will also support. Recognition in cellular telephone and in-car applications will be used as a testbed.

SPHEAR Partners

SPHEAR has 7 participants

SPHEAR Programme

The programme has four themes.
  1. to model the temporal organisation of auditory events
  2. to model the interaction of primitive cues with schema-driven mechanisms in sound-source segregation .
  3. to develop speech recognition architectures which are capable of functioning within auditory scenes.
  4. to investigate the techniques developed in other themes within the application domains of voice control of cellular telephones and in-car recognition.

For details of the SPHEAR programme, click here.

Research Updates

The following pages provide details of what's going on in SPHEAR

Bochum (task 1.2)

IDIAP (task 3.1)


Sheffield (task 3.2)


SPHEAR Recruiting

If you are interested, send your CV and the names and addresses of two referees to the address below. You can do this be e-mail if you wish. Indicate what SPHEAR tasks (see the detail document) and what SPHEAR labs interest you, and when you expect to be available.