Charmed Quark Systems, Ltd. - Support Forums and Community

Full Version: What is the current state for voice input?
You're currently viewing a stripped down version of our content. View the full version with proper formatting.

What, if any, is the current best solutions for voice input to CQC. I want to add a few simple voice commands to a system I am designing. It would use a very limited pre-defined vocabulary, but must cope with multiple speakers without a software training phase.

I have found some old posts about a Microsoft macro plug in for XP and Dean has added a teaser about 4.3 development that could put the required tools in place. What solution could I use in the next 4-6 weeks?

The stuff for 4.3 is more oriented towards phones, and initially will likely be more oriented towards intercom and notification. It's just that a lot of the stuff done for that will ultimately be useful in voice activation as well, so once that stuff is done it won't be nearly as big a step to do it. But I don't foresee it happening in that time frame. There's a lot to do just on the first steps.
I haven't played with it myself but perhaps something could be done using VoxCommando. It can send keyboard events to navigate the IV, or you could use it to send commands to EventGhost which could trigger events in CQC.
If you wanted a more solid and secure setup, an external program to do the voice recognition which then used the XML Gateway to do things in response (perhaps invoke a global action) would be one way to do it.
I think one good possibility would be to dig into the Kinect SDK and try to write something to interface with CQC. The VoxCommando site says they haven't had a lot of luck with it, but I think they are using it as a standard Mic.

I know with my XBox I can yell Pause at it from 15 feet away with a movie playing and it responds pretty well, maybe 75% of the time.