Mostly likely somewhere at some point over the holidays you've had a go on Microsoft's motion controller Kinect. Whether its to jig along to a song with Dance Central, or just flit about on the Dashboard menu with a wave of your hand, it's impressive stuff (once you've got the distance and lighting right first).
So how does the technology work? Well, two designers who worked on the Kinect, Ron Forbes and Arjun Dayal have gone into some detail to expose what makes the motion controller tick via a Microsoft engineering blog.
It's an interesting read, showcasing the challenges the team came against in using a system that'd read body gestures - noting the multiple nuances on even a simple 'push' gesture.
"Let's also consider the complexity of recognizing gestures with a seemingly simple movement like a push. When you initially think of pushing your hand out to select, you likely think of the direction of the push being completely perpendicular to the plane of your body. In reality, you don't push away from your body in a straight line because of how your shoulder and arm joints are designed. The result of this nuance is that every human performs a push slightly differently, but everyone expects that their action will result in the same outcome - a push. Successful gesture recognition is about understanding the nuances in the way humans perform interactions and teaching technology to be mindful of these differences.
Part of the excitement of working on revolutionary tech like Kinect is overcoming these challenges in gesture creation while creating a product that is approachable and easy to use. Every decision we made was breaking ground in Human Computer Interaction, and all of our work had the potential to define the way users interact with entertainment technology in the future."
Read the whole thing here.