Spankey said:
out of interest, have you ever tried MS's voice to text recognition software? or even text to speech? If they have trouble with that, how much trouble will this cause? Don't get me wrong, Natal is amazing, as is Milo, but the difficulties in getting it to work properly are huge. then there's slang and dialects to work into it. very frigging difficult. A sketch of a plane is peanuts for computers to recognise in comparrison But if MS says they can do it, I'm not going to argue. that's one talented bunch of people. |
I'm not for one moment saying that Milo works as well as has been presented to us.
I'm just saying that people are questioning how Microsoft can distinguish from the infinite possibilities of what people can say or how they say them yet Sonyseem to have somehow worked out how to recognise what has been drawn from the infinite possibilities of what can be drawn. If you see what I mean.
I just think that somewhere in there you have to tell it what you are drawing, from a select list of object. There are only a finite number of forms of transport for example.
I cannot lie and say the prospect of an eyepet flying round on a giant penis doesn't interest me







