Advisor: Szymon Rusinkiewicz
TL;DR
Jarvis from Iron Man! Use your hands to pick up and move virtual objects in front of your screen!
Overview
3D graphics are an exciting and versatile method of visualization. However, with the standard interfaces of mouse, keyboard, and 2D screen, interacting with 3D graphics is highly cumbersome and unintuitive. The ideal system would combine a 3D display with a spatially accurate gesture system, so that the user could interact with the virtual object as if it were a real one.
I’ve implemented several prototype versions of this system that successfully aligns the user’s hand and the 3D object in space. For display, I have used 3D monitors powered by Nvidia 3D Vision. For input devices, the first iteration used the Intel Perceptual Computing SDK while later versions used the Leap Motion.
The main focus of the prototypes have been for user studies to compare whether the natural interaction and display is advantageous to users. Previous in-air gestural systems have shown low accuracy compared to the mouse; however our results have suggested that, when coupled with the 3D display space, gestural systems are much faster and more intuitive.
Brief demo showing a simple 3D modelling system built on top of Holograsp
System Implementation
Coming soon…
User Study Results
Coming soon…