It has been a while, I have been refining my speech recognition application over the last four months. It is very robust now. It has many abilities. I wrote the whole thing for the Kinect Sensor. I still am using version 1.7 of the SDK since 1.8 cause to many errors. they changed alot of the classes used in the sdk and 1.7 is good enough.

i was going to ask a question to you guys. Does Amulet use Prompt builders for when Paul is speaking?

I am totally mystified by how you guys are able to cancel the Asynchronous synthesis of speech using the astrix in your command lines, if you aren't using prompt builder.

I find that writing speech programs for the Amulet Remote is much, much better then kinect. For reasons such as calling the recognizer to cancel and stop each time the synth speaks. Where as the remote couldn't hear the synth because it shuts off when you put it down. Brilliant design still to this very day.

I actually out fitted my entire home with microphone arrays, and with the MB3 server, I am able to send play commands to any client in my home, or stop play back of any user session, on any client.

With Speech recognition I can ask what my son is watching upstairs, and if he has broken the parental controls on my user library (again... little smart a$$) and is watching something he shouldn't be, I am able to shut it down with my voice...awesome!

I have created gesture interactions, and perfected them with kinect. open and gripped hands pause playback (I stop speech recognition during media playback, for obvious reason with a microphone array). then the synth asked if i want to stop or wait and resume my media.

Pretty neat stuff.