Our goal is to allow, for the first time, the joint creation and progressive refinement of a shape and of its animation. This will enable users to first design and visualize their digital story-board using rough shapes such as construction volumes. They will then iteratively refine both geometry and motion, in any preferred order, until both are fully designed. We would like to remain as close as possible to the methodology we adopted for shape design. In particular, the animation models used to generate motion should remain totally transparent to the user, which has never been attempted, to our knowledge, in a general setting. Let us precise the scientific problem and detail our methodology.
Scientific problem: Extend sketch-based modeling paradigms to the design of animations and enable further editing over time through over-sketching and appropriate time-warps. In particular, an appropriate animation model should be created to control each responsive shape over time.
Our first results :
Our contributions include methods for real-time physically-based simulation of passive material with the image-based intersection volume [Wang 2012], a very novel work on using a sketch-based approach, the “line of action” for controlling the pose of virtual characters [Guay 2013], several works on virtual crowd simulation [Lim 2013, Jordao 2013], and some work on storyboarding and autonomous cameras [Galvane 2013,Ronfard 2013], which are an excellent complement to interactive control of animations.
