A Facial Rigging Survey

How to read this

Category

State of the art report surveying facial rigging as a field

Contributions

• Gives facial rigging a single critical review rather than leaving it scattered across production folklore and individual papers
• Frames rigging plainly as building a set of controls to operate a model, the strings of a puppet, and works outward from that
• Walks the techniques available for preparing a face for animation, with their uses and their failure modes side by side
• Names the recurring problems in the process, which is the part a survey can do that an individual paper cannot

Context

This grew out of Veronica Orvalho's own line of work on reusable and transferable facial rigs, and it is co authored by Frederic Parke, whose 1974 parametric face model opens the whole digital face lineage. Reading it next to those two is the point: it is the moment the field stopped to take stock of what forty years of facial rigging had actually produced.Builds on: Transferring the Rig and Animations from a Character to Different Face Models · A Parametric Model for Human Faces

Correctness

A survey's claims rest on coverage and judgement rather than on an experiment, so the honest question is what it left out and when it was written. It dates from 2012, which means it predates the blendshape state of the art report, the performance capture wave that followed, and everything neural. Treat it as a map of the classical foundations, not as a current picture.

Clarity

Written to be read by people who rig rather than only by researchers, which is unusual for a state of the art report and makes it a good entry point for someone new to facial work.

How to read it

First pass, read the taxonomy and the problem list, since that vocabulary is what you will actually reuse when arguing about a facial pipeline. Second pass, follow its references outward: a survey is most valuable as a reading list, and this one indexes the pre neural facial rigging literature well. A third pass is unnecessary; instead, pair it with a modern survey to see which of its open problems got solved and which are still open thirteen years later.