[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
I have Sphinx working with a British English dictionary (BEEP) and with
models trained from British English speakers (WSJCAM0). If you have
access to WSJCAM0 I can send training scripts.
On 13/03/2012 12:40, Stefan Bleeck wrote:
A group of students of mine are investigating how state of the art
automatic speech recognition systems cope with noisy speech compared
to hearing impaired listeners. They are planning to use Sphinx,
because it's use is well documented and it's open source.
Unfortunately it is a bugger to install. Does anyone has (or know of
someone who has) a running Sphinx version with a BRITISH English
dictionary? We are interested in getting a running program (Windows)
or to collaborate by sending a few hundred wave files over (prize:
your name on a poster).
Background: objective measures of speech intelligibility fail to
analyse contextual information. Hearing impaired people often obtain
higher scores then they should in speech tests. ASR use contextual
Speech, Hearing and Phonetic Sciences, University College London