Building Efficient, Accurate Character Skins from Examples

Alex Mohr and Michael Gleicher

Abstract

Good character animation requires convincing skin deformations. These deformations are often subtle and detailed like muscle bulges. Such high-fidelity characters are typically created in commercial animation packages using very general tools and are often too slow to compute or take too much memory for interactive applications. Instead, characters created for interactive environments often trade off deformation fidelity for fast computation and small memory size. This paper presents an automated framework for building compact and efficient approximations to detailed skin deformations, suitable for interactive tasks. Our method starts with an arbitrarily rigged character in an animation system. A set of examples is exported, consisting of skeleton configurations paired with the deformed geometry as static meshes. Using these examples, we fit the parameters of a deformation model that best approximates the original data. The underlying deformation model we use is an extension of the traditional linear blend skinning model widely used in interactive applications. This extension allows us to capture subtle and detailed effects required for believable characters but remains nearly as fast and compact as linear blend skinning.

Here is the pdf of the paper.

Here is the video accompanying the paper.