The patent for Microsoft's motion-sensing camera Kinect suggested that the device could understand American Sign Language. Well, it can't. At least, the version going on sale in November can't.
Microsoft has told Kotaku "We are excited about the potential of Kinect and its potential to impact gaming and entertainment. Microsoft files lots of patent applications to protect our intellectual property, not all of which are brought to market right away. Kinect that is shipping this holiday will not support sign language."
So why did the patent suggest it could? Well, sources close to the evolution of Kinect's development tell us it's because the version of the hardware that'll be available later this year isn't as capable as the a previous iteration of the device
The original Kinect had a much higher resolution (over twice that of the final model's 320x240), and as such, was able to not only recognise the limbs of a player as the current model version can, but their fingers as well (which the current version can't). And when the hardware could recognise fingers, it would have been able to read sign language.
But that capability came at a cost, and while Microsoft had always intended Kinect to sell for $US150, "dumbing down" the camera would have meant that Microsoft wouldn't be losing as much money on each unit sold, an important point should Kinect prove to be a failure. So dumb it down they did, reducing the camera's resolution (which in turn reduced the number of appendages it'd have to track) and placing the burden for some of the device's processing on the console and not Kinect's own hardware.
This probably isn't the first time you've heard such a rumour, but this latest time at least explains why Kinect can't read sign language!