The MASLE website is here, but I don't think it has been update for a while.
The voice recognition and transcription functionality of Google's Web Speech API is very well developed. I know of some projects using that, but I don't think they are publicly available ... yet.