Doodler

A graceful and intuitive interaction technique for controlling gesturing avatars.


[Francesca Barrientos home]

 

Doodler image and movie.

doodler.exe video

Idea

Gesturing with an avatar should be as simple as doodling on a piece of paper. To explore this concept, I created this prototype for controlling the free flowing gestural arm movements of an avatar by doodling on a graphics tablet. (It also works with a regular mouse.)  Communicative gestures are often creations of the moment, generated spontaneously to accompany a singular utterance within a particular context. In this interaction technique the user's spontaneous and continuous motion is reflected meaningfully in the avatar's motion.

Expressive, Constrained movement

This technique maps the two-dimensional movement of the pen's motion on the tablet to the the motion of an avatar's arm in virtual three-dimensional space. The avatar's motion is necessarily simple. Yet, this smooth motion which responds to the slightest movement of the pen's tip, can also be quite expressive and graceful. As any puppeteer or dancer can explain, even very simple motion, when combined with appropriate timing and intent, is rendered meaningful and evocative.

Insights from gesture research

Human gestural motion often follows a particular pattern:  just before someone gestures, they will raise their arm from a resting position into a into a ready position; then the gestural motion proceeds; finally, the arm returns to the rest position. The doodler technique mimics this motion. When the pen touches the tablet, the avatar arm moves to a ready position. After the pen is raised, the avatar lowers its arm to a resting position.

Demo

You can try using doodler. Doodler runs as a Java applet that interacts with a VRML browser. You will need a Java enabled web browser and a VRML browser to try it out. It has been successfully tested with IE6.0 with the Cortona VRML browser plug-in.

You can also download the source code here (31 KB).  It requires the VRML EAI package.

Related publication

The ideas behind this work and others is discussed in a workshop paper.

Francesca Barrientos, "Continuous Control of Avatar Gesture," presented at workshop Bridging the Gap: Bringing Together New Media Artists and Multimedia Technologists in Proceedings ACM Multimedia 2000 Workshops, 4 November 2000, Marina delRey, California, pp. 5-8. (152 KB)


This page last modified 18 December, 2002