Yet more news form the world of Kinect Hacking – this time a video demoing some awesome work on ‘teaching Kinect to recognize objects.’

In the video, embedded below, we see developer yankeyan using a combination of OpenKinect drivers alongside OpenCV (computer vision) for image processing and recognition plus FestVox for the (awesome 80’s sci-fi style) speech synthesis and, for speech recognition, the CMU Sphinx toolkit.

Yankeyan notes that he’s ‘wearing a headset because OpenKinect does not yet support audio input.’

Before you ask ‘Whats the link with Ubuntu?’ you’ll see in the video that this is yet another developer using Ubuntu as their development platform of choice.

Thanks, as ever, to yo2boy

Video kinect