Something else that has really changed lately, is the pushing of computer vision down into video games.So one of the first places to do that was the Nintendo Wii. The remote control, there's actually a camera system built right into here, that tracks the two dots that are from the sensor bar and reports that information.

But the real game-changer in terms of computer vision being involved in games was the Microsoft Kinect. The Microsoft Kinect is a depth sensor. And what I mean by that, is it can produce a scene like this, that is, it can, it can produce an image like this from a scene.

And this is a depth image, right? So darker is farther away. And brighter is closer. And gray here is sort of in between.

So from a technological perspective. From a raw technology perspective, most people think that what was important about the Kinect was that it created a depth image.

The important thing about a Kinect is that it can produce skeletal descriptions. Using a combination of machine learning techniques applied to computer vision, the folks at Microsoft and this came out of Microsoft Cambridge in the UK. Were able to recover the skeleton geometry of people from the depth image. And by the way, they do it instantaneously. Every frame, they do it differently. They don't even track it, they just do it one frame at a time. It's that robust.

Right? And because they can get the skeleton information, you can build really cool user interfaces. So you could you know, do driving games and go like this, to steer your car.

Much more interesting of course, is playing with robots. And here you see, this is Simon, in Andrea Tomas' lab, here at Georgia Tech. And the reason that Simon is waving to this student, is that behind Simon over here, there's a Kinect that's looking at the student waving. And is getting back the depth information. Also the skeleton information is then interpreting the skeleton information, in terms of what the human is doing.And that in turn, allows the robot a decision about what the robot wants to do.

So even though the Kinect was sort of promoted and invented as a way of impacting games. And early on, there was some uncertainty at Microsoft whether they wanted to open up that, system and let people use it. They quickly realized that this is something that everybody wants to use. And it has really revolutionized the way people think about depth imaging, and computer vision applied to depth imagery.