M'rk, son of Mogh wrote:
Owain Taggart wrote:
And that's just it, though. In a perfect world, everyone would enunciate perfectly, but this is a system that's been touted as being able to understand natural speech. Anything else is just awkward and kind of defeats the purpose of its usefulness.
Getting people to stop slurring their words and to speak and enunciate properly is a bad thing now?
Nope, never said that. I'm all for enunciating properly, in fact I get very frustrated when actors end up mumbling their words throughout a movie, and when turning up the volume won't help due to bad audio.
Point is, most people don't enunciate properly in the first place. I think what a lot of people look for in a system like this is being able to just use a normal every day conversational tone. And this isn't a knock on MS, but I don't think we're quite there yet, in terms of voice recognition in general. We're not up to the level of having something accepting commands in a similar way to HAL. Not speaking about any response back, mind you, but just being able to accept commands using an everyday conversational tone. Having to speak up for a machine is just simply an unnatural concept for many people, and until that happens, I don't think a system such as MS has with the Kinect will have very widespread appeal or success. It's certainly possible though, that over the course of the life of the console, that they'll find ways to improve it as voice recognition technologies improve. There are a lot of variables to consider though in a system like that, such as room size, ambient noise, accents, and the volume at which someone is speaking at, and I think that's a large margin for error right there, and why a lot of people may have trouble with it.