Open source speech recognition

There are at least three open-source speech recognition projects out there:
CMUSphinx: The Carnegie Mellon Sphinx Project, and HTK, and Julius (Japan).