The following new features and improvements were
added: confidence scoring, support for dynamic
grammars, out-of-grammar utterance rejection, and
improved JSGF support. This release also includes
some new demos, a narrow bandwidth acoustic model,
as well as a number of bugfixes.
