Gestop , Next generation desktop powered by perceptual computing
I fortunate to be one of 150 odd developers who are sent with Creative camera Kit ( RGB camera with IR based Depth sensor and Array Microphones).
As usual, I was glued to system for first two weeks figuring out the possibilities and power of the SDK.
From the initial work, I must tell you that perceptual computing is going to change a lot of things in near future.
I have finalized my demonstration application or Apps with PCS. It is called Gesture Desktop or in simple words Gestop.
Here is the screenshot of the App.

Design:
It basically hooks the mouse driver and allows you to take control of mouse with your hand.
App tiles are big and bolted because your hand is not as precise as mouse. While keeping the hand in front of camera, your hand shakes which keeps moving the mouse pointer to places. Hence larger action area.
When you minimize Gestop, you are again back with your usual mouse ( And of course multi-touch, for an Ultrabook device).
Speed
Processing of a frame is less than a millisecond, giving you a negligible processing lag. However with faster processing, shake of mouse will be more. Hence a good synchronization must be designed to control this behavior.
Features
You can control a media player, power point, windows photo viewer and acrobat reader. The controls involves starting and stopping Apps, reverse, forward, volume control, content zooming, content navigation, text highlighting while giving presentation.
Note, alongside Gestop, you can also control the Apps through usual way.
It also comes with a Voice Synthesis module that speaks out the time, to show how voice perception works.
Applications
Touch is not the most suitable environment in every aspect. For example if you are in kitchen or in a factory or driving a car, its not always the best solution to control the devices through touch. So perception is the answer their.
Stability
The "beta" level tells you that the platform is not yet production ready. However some gestures and features are ultra responsive and quite stable. I am almost addicted to Gestop for my media player control and presentation.
Demo
have a look at my week 1 video. Week 2 video with complete Gestop in action will be released soon.
Technology Used:
C#.Net, Wpf, Managed C++ Coding, 3-D depth sensor, Perceptual Computing SDK, Hand Gesture Recognition, Voice Synthesis.
End Note:
Do tell me your suggestions and insights about other applications which could be controlled through this interface and how we can smooth mouse movement with hands.