Been playing around with speech synthesis and voice recognition. This is a new area for me and still haven't caught up with all the reading that I have to do.
The toolkits and APIs out there look very impressive particularly the ones by IBM and Free TTS. Unfortunately Free TTS does not have voice recognition (duh isn't that why they call it TTS - Text To Speech) but it does seem a lovely piece of software. IBM web sphere voice SDK does have speech recognition in addition to TTS - unfortunately the manual is extremely wooden and leads to very boring reading. They haven't really included any usefull examples either.
Will keep you posted about my finding.