idiggory, King of Bards wrote:
But I just can't see your average consumer wanting kinect. Maybe that's me misreading the situation, but it just sounds... off to me. I mean, one thing they keep talking about is how you can just say "xbox on" and it'll be good to go. That's nice... except that there's plenty of reasons you would use xbox on when you didn't want it to turn on. For instance, "Is the xbox on?", "Don't leave the xbox on", "Don't turn that xbox on until after you've done your homework", etc.
Ditto for xbox off. Some kid's mom walks in and says "Don't you think it's time your turned the xbox off?" Or your kid sister just LOVES that she can control it with her voice, so she does so. Whenever she can. Even if you're in the middle of a COD deathmatch.
I mean, I can't possibly be the only one with a toddler in the house who would absolutely do this every chance she got... And with the rate she picks up new technology, there'd be no chance in hell she wouldn't know how to do it after the device had been in the house for a few days.
And that doesn't account for other things, like "I WANT TO WATCH CAILLOU!" Is "watch ____" going to be a command? Because that would be REALLY annoying.
Maybe your experience with Windows Speech Recognition (WSR) has been different than my own, but that is usually not how voice commands work.
If you have a command such as "xbox on", then it looks for that phrase and that phrase only. When WSR starts to hear speech, it then constructs the phrase/sentence that it picks up. As long as you keep speaking, it continues to build up and does not perform any other action (there may be some limit, but it was easily picking up full length sentences when I played around with it). Once you finish speaking, that entire list of words becomes the command phrase, not just a small portion within. An action is performed if that entire phrase matches a command.
Toddlers could be a problem, but only because I expect they would pick up on you saying "xbox on" and "xbox off" then start using it because it sounds funny, unaware of what it actually does. That would be my guess, anyway. Usually the WSR works around your accent and speech patterns, so it might not pick up on the toddler because of that either, but I'd be less certain of that part.
Finally, system voice commands ("xbox on", "xbox off", "watch Dr. Who", etc.), might end up being disabled while playing a game or possibly settings to turn those particular ones off. Given the disaster of the recent announcement, I have my doubts, but it would be something I would think is a reasonable approach.
Edited, May 22nd 2013 8:58am by xypin