Pocket Sphinx is a frontend to the sphinxbase large-vocabulary,
speaker-independent, continuous speech recognition engine.

This is an early release of a research system.  The APIs and function
names are likely to change, and several tools still need to be made
available to make this all complete.

Once the system is built, try running the Perl script sphinx2-demo.
The pocketsphinx-test script should run pocketsphinx-batch over an example
utterance of "go forward ten meters."

WWW: http://www.speech.cs.cmu.edu/sphinx/