Shaping the imagination : from 3D digital creation to animated virtual worlds

Abstract

This lesson was designed to give, in an hour's time, an overview of the importance of 3D graphics creation in many fields (arts and humanities, science, technology), its potential to revolutionize the tasks of imagination, design and expression, and a clear idea of the nature of theresearch being carried out in computer graphics to advance in this direction. Several examples of challenges and associated solutions were described, from the creation of isolated 3D objects (volumetric sculptures, clothing), to that of virtual worlds with myriads of interacting and often animated elements (vegetation, hair, rugged terrain with waterfalls). I've shown how certain solutions manage to harmoniously blend mathematical modeling, real-time algorithms and "trompe l'oeil" graphic representations, allowing the user to control his or her creation while abstracting from the underlying complexity. Below, I detail some elements of this methodology.

Recent research in computer graphics has focused on developing 3D creation methods that combine simplicity and speed of use with guaranteed visual quality and control of results. The aim is for users to be able to literally "shape" the forms and movements they imagine, while relying on the digital tool to automatically fill in details and maintain the constraints they desire in terms of realism. I have identified three methodological principles to advance towards these goals:

Firstly, to propose creation through gestures. These can be drawing gestures to sketch out a new form, sculpting or modeling gestures to improve it, add detail, or mime gestures to indicate movement. To these interaction metaphors inspired by the real world are added certain simple actions that have already revolutionized digital environments, such as copying and pasting to reproduce and transfer certain elements.
Secondly, to revisit the graphical models so that they react as a human user would expect, under these interaction gestures. This involves integrating appropriate knowledge into the models, enabling them to respond to the semantics that a human user associates, almost involuntarily, with his or her actions.
Finally, to develop different scaling methodologies, enabling the user to orchestrate the creation of a complex virtual world, made up of hierarchies of possibly animated elements, without having to manipulate them one by one.