Sphinx 2 is a real-time, speaker-independent speech recognition system.

this package contains an acoustic model for sphinx-ii trained for close-talking microphones. it is the default acoustic model used by the demos, and should provide adequate performance for most desktop applications.