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.