This entire project is defunct! It has been replaced by TinySR, which you can find here: http://github.com/petersn/tinysr
TSR is a light weight speech recognizer written in C. It has the following major goals, in decreasing importance:
- Be small: tiny source, tiny binary.
- Be fast: RTF (real time factor) below 0.5 on any modern system.
- Be portable: Don't use any fancy libraries.