CSTR is a multidisciplinary research group that undertakes
application-oriented speech research. Its main work is in the areas of
speech synthesis and speech recognition. Its goals are both to do
original research which can be published in the open literature, and
to see the results of its work embodied in technologies that are of
benefit to users. It is interested in collaborating with outside
academic or industrial partners in pursuit of either or both of these
goalsCSTR has developed several large scale software systems,
including the Festival speech synthesis system and the Edinburgh
speech tools. CSTR has excellent facilities for recording, storing and
processing large volumes of speech. It has access to a range of speech
analysis and display software, developed both in house and elsewhere.
It also has access to large speech databases, some of which were
recorded in house.