I keep seeing people ask the same question over and over again: “What can Kinect do that current motion controllers & camera-based devices can’t?” So here’s a list for folks to start from.
- Z-AXIS TRACKING
Kinect has depth perception – a result of is multiple camera lenses. Unlike the Sony Eyetoy which can assist in capturing 2D movement along the X & Y axis, Kinect can capture movement and imagery in 3D along X, Y, and Z axes. Additionally, it captures actual movement, acceleration, & position changes relative to the camera & other players in all 3 dimensions.
- IMAGE RECOGNITION
Kinect can recognize and understand the difference between physical real world objects like skateboards, tennis rackets, and other items that might be used in game play by matching patterns in real time using its on-board CPU. Taking image scans, loading them into the device’s memory and transferring these images for remapping to items in-game via bump mapping is trivial.
- FACE/PROFILE RECOGNITION
Kinect’s built-in software provides online/profile authentication using facial recognition and a simple hand gesture (a wave) that Kinect uniquely recognizes. Through its algorithms, it can also recognize people as important control objects distinguishing them from furniture and other items in the home through shape identification algorithms and heat mapping. Most importantly, through heat mapping and other visual characteristics, Kinect can differentiate between players despite overlapping each other (one standing behind the other) or changes to their position relative to Kinect; i.e. swapping sides does not impact Kinect.
- SIMULTANEOUS ARM/LEG/HEAD TRACKING
It’s one thing to recognize the difference between a sofa & a human. It’s another thing however to accurately identify movement of one’s arms & legs to ensure realistic game play. Kinect does this: Kids can’t line their bodies up and combine their visual profiles to fake Kinect out. It also tracks the movement of all arms & legs simultaneously to ensure the player is executing dance moves in total, forming the correct shape with their bodies in Yoga games, kicking one enemy while punching another, etc. Additionally, when your head turns to the left, Kinect is smart enough to know that you want to see more of what’s on the ‘left’ of the screen and can rotate it, as if you were scanning the cockpit of a car.
- ON-DEVICE COMPUTATION
This is a technical detail but an important one: With Kinect, all algorithmic/compute-intensive work is done on-device through its own CPU, memory, and software-in-ROM and does not requires the usage of the Xbox 360’s resources (such as the CPU or GPU), allowing the console to dedicate its resources exclusively to the game itself, rather than the player’s movements, their interpretation & recognition. Kinect does NOT burden the This is one of the reasons for Kinect’s current cost.
- VOICE RECOGNITION
Kinect’s on-board microphone provides a 4th dimension of control: Voice. Through vocal recognition software, Kinect is able to, without any training, comprehend commands and differentiate between player’s voice patterns. Call out commands, change strategies & tactics, order computer controller players… all without pushing a button or moving a limb.
- INFRARED OPTICS
Recognizing player identities & movement regardless of light differences is something unique to Kinect. It’s infrared camera provides accurate player tracking & movement even when the lights are turned out during a Netflix movie, a game of Alan Wake… or the upcoming Silent Hill.
More on Kinect from Microsoft’s press/media area:
http://www.microsoft.com/presspass/presskits/xbox/videogallery.aspx?contentID=360_kinect
kurtsh
Kurt's Thoughts 