Some speech recognition applications require
simultaneous active grammars. For instance, an
application may have a global grammar that supports a
set of voice commands that are always active along with
a context dependent grammar with a set of voice
commands appropriate to the specific context of the
application. Sphinx-4 should support multiple active
grammars.
Logged In: NO
The JSGF allows you to add a grammar dynamically at the
expense of losing the previous grammar. However, a
dynamically generated and loaded grammar could reference a
baseline grammar, achieving a simple model of what you suggest.
Ticket moved from /p/cmusphinx/feature-requests/8/