Ding ding ding...correct! I work in the Computer Vision field and have already seen presentations of plans to overlay ads inside video content. The example shown was your classic soap opera. The program detects suitable advertising surfaces for example a table top or a wall in a scene and then renders an advert in that location. This is then embedded in the video and tada you now no longer have to worry so much about piracy because you have adverts embedded in the pirated video in a way you can't easily cut out. The program automatically calculates the value of an ad location depending on time on screen, angle to camera etc. and gives a web based front end for buying ads from the TV company.

This could, given enough computation power in phones and cameras, allow companies to place ads in our videos and photographs based on location and orientation. Although it is hard to see why this would be better than placing huge billboars - but still I put nothing past an advertiser with enough intent.

I can guarantee you that there is not a neural-network behind this. I work in almost exactly the same field except my research is targeted at 'guessing' the pose of people in still images rather than in video (Natal not only has video, but also a 3D point cloud making this a fairly trivial problem). Given the amount of data Microsoft have gathered they will essentially be doing a nearest neighbour search using a very efficient classifier. Neural networks are very out of fashion in Computer Vision and other AI fields at the moment and have been surpassed by devices such as Hidden Markov Models and Support Vector Machine classifiers.

In any case, this is a good thing. Is it goodwill from Microsoft? I'm not sure. When they made the big internet patent grab several years ago, it seemed about as evil as you can get. But in having done that, one could argue that they kept other more evil companies from grabbing and exploiting them. If Microsoft simply gave them away now, that would be goodwill. If they charged $100B, that would be evil. If they charge essentially what they've spent to acquire and hold them, well, that still seems like a good thing. Twenty-two patents at $50K each would be $1.1M. So if the price was >$5M, I'd call it evil.

My old Atari 400, I still blame your membrane keyboard for my two finger typing.

I totally disagree with you here. Apple feels 100% different from Microsoft (hell, they even put the minimize, maximize and close icons on the window manager completely on the opposite site) and yet they succeed.

No; you need quality. As Linux user I will be taking a fresh perspective on this from a higher purpose here. First up: Does it even work? The awnser is "NO". Yeah it is comming with Gallium3D bla bla bla but it is comming and it is not here yet. So no-go from graphics (unless you still feel like using a 2D desktop ala 2009) and no-go for OpenCL.

Next stop: "Does it have advantages over the competition?" The awnser is "It's on par". Linux, in an idiot-friendly distro, is not faster and about kind-of as secure as Mac (maybe more, maybe less; only the future will tell us).

Final destination: "Does it out-app the competition?" and the awnser is probably still "No".

So ladies and gentleman: Linux is not ready yet, although I believe it will be kicking serious ass in a year or four.

as the undergrads get younger and younger. From the age of 16 to 21/22 most of the younger students are at 3 years younger tops in the UK...once you hit post-grad studies they get 5, 6, 7 etc. Children born in the 90's are now at university!

Not launched yet, the app "Nearest Tube" is awaiting Apple approval...

From what I've read Natal is able to aquire depth information - using this I figure they would be able to distinguish the quite unique shape of a face (ie: nose, chin, eye sockets) from the generally round back of your head. Although things might screw up if you put long hair over your face I guess.

I agree with you here, we've discussed how unimaginative the majority of the Wii games are going with just a 'waggle' to perform actions...I suspect with Natal game designers will end up having to put in a load of crazy gestures because there is no standard controller. Also think about things like golf/baseball/driving - I'm pretty sure all of these would feel much more immersive if you actually had an object in your hands as with the Wii.

3D motion capture is possible from one camera actually (I'm working on it for my PhD). There's only a small number of configurations a body can be put into to fill the same 2D silhouette as seen from a single camera.

Besides that the demo's shown in the GP video really only need 2D motion capture anyway, except maybe the accelerator for the driving game. But seriously, why would I want to stand/sit with my leg in an awkward position when I could just hold "A" instead? For the same reason my Wii is also collecting dust.

most people on here seem to knock twitter but as has been mentioned it is pretty much the only source for news right now. Follow updates on the #IranElection hashtag here: http://hashtags.org/tag/iranelection/messages For what it's worth I don't even use twitter, but it's times like this that I realise it kicks the ass of TV news for real-time coverage.

Obama: Hey Ballmer, you mind if we borrow 90% of the world's computers for a quick cyber war?

Ballmer: Finally, the moment I've been waiting for! *Throws ceremonial war chair at wall*

Use the scan-line technique to scan each line with a rapidly adjustable mix of red, green and blue lasers.

As long as the laser light spreads out proportional to the distance it has travelled the picture should look good wherever the laser will reach. My only worry is the battery life.

Ouch!

Oh no!...members of the other sex!