Skip to content

← ArchivePaper2022

Local Anatomically-Constrained Facial Performance Retargeting

Prashanth Chandran, Loic Ciccone, Markus Gross, Derek Bradley

SIGGRAPHDisney Research25 cites2 descendantsFacialRetargeting

Retargets facial performances across diverse rigs using local anatomical constraints that preserve expression fidelity and prevent implausible deformations.

Abstract

Generating realistic facial animation for CG characters and digital doubles is one of the hardest tasks in animation. A typical production workflow involves capturing the performance of a real actor using mo-cap technology, and transferring the captured motion to the target digital character. This process, known as retargeting, has been used for over a decade, and typically relies on either large blendshape rigs that are expensive to create, or direct deformation transfer algorithms that operate on individual geometric elements and are prone to artifacts. We present a new method for high-fidelity offline facial performance retargeting that is neither expensive nor artifact-prone. Our two step method first transfers local expression details to the target, and is followed by a global face surface prediction that uses anatomical constraints in order to stay in the feasible shape space of the target character. Our method also offers artists with familiar blendshape controls to perform fine adjustments to the retargeted animation. As such, our method is ideally suited for the complex task of human-to-human 3D facial performance retargeting, where the quality bar is extremely high in order to avoid the uncanny valley, while also being applicable for more common human-to-creature settings.

How to read this

Category
Method: facial performance retargeting
Contributions
  • A two-step high-fidelity offline facial retargeting that is neither expensive nor artifact-prone
  • First transfers local expression details, then runs a global anatomically-constrained surface prediction to stay in the target's feasible shape space
  • Familiar blendshape controls for artists to make fine adjustments
Context
Builds on anatomically constrained face modeling (related to Wu et al.'s An Anatomically Constrained Local Deformation Model for Monocular Face Capture, 2016), positioned against costly large blendshape rigs and artifact-prone deformation-transfer methods.Builds on: An Anatomically Constrained Local Deformation Model for Monocular Face Capture
Correctness
Targets the demanding human-to-human 3D retargeting case where the quality bar is high; the key assumption is that anatomical constraints keep predictions plausible, so a reader should remember it is an offline method and that fidelity rests on the local-then-global decomposition behaving well across diverse rigs.
Clarity
Clearly structured around the two steps; a first pass conveys the local-then-global idea, a second pass for the constraint formulation.
How to read it
Read the two-step pipeline and the anatomical-constraint rationale first; second pass on the local detail transfer and global prediction math if you work on retargeting.

Builds on

Built upon by

Related work

Keywords

This page summarises the entry and links to its original source. The archive never hosts or redistributes the publication itself.Show it in the full archive list →