Skip to content

← ArchivePaper2020

Expression Packing: As-Few-As-Possible Training Expressions for Blendshape Transfer

Emma Carrigan, Eduard Zell, Cedric Guiard, Rachel McDonnell

CGFAcademic13 citesFacial

Integer optimization selects a minimal set of scanning poses with non-overlapping blendshapes for efficient blendshape transfer.

Abstract

To simplify and accelerate the creation of blendshape rigs, using a template rig is a common procedure, especially during the creation of digital doubles. Blendshape transfer methods facilitate copy and paste functionality of the blendshapes from the template model to the digital double. However, for adequate personalization, such methods require a set of scanned training expressions of the original actor. So far, the semantics of the facial expressions to scan have been defined manually. In contrast, we formulate the semantics of the facial expressions as an integer optimization of the blendshape weights. By combining different blendshapes of the template model, our method creates facial expressions that serve as semantic references during scanning. Our method guarantees to compute as‐few‐as‐possible training expressions with minimal overlap of activated blendshapes. If the number of training expressions is limited, blendshapes are selected based on their power to personalize the resulting blendshapes compared to generic blendshape transfer methods.

How to read this

Category
Method: optimization for blendshape-rig authoring
Contributions
  • Formulates the choice of training expressions to scan as an integer optimization over blendshape weights
  • Combines template blendshapes into semantic reference expressions with minimal overlap of activated shapes
  • Computes as-few-as-possible scanning poses, ranking shapes by personalization power when the budget is limited
Context
Sits in the blendshape-transfer and digital-double literature, replacing the manual definition of scan poses, and relates to direct blendshape control as in Lewis and Anjyo's Direct Manipulation Blendshapes.Builds on: Direct Manipulation Blendshapes
Correctness
The method targets efficient blendshape transfer from a template rig and assumes the template's shapes meaningfully span the actor's expressions; readers should note that minimizing scan count trades coverage for capture cost, so personalization quality depends on the chosen budget.
Clarity
Accessible at a first pass for the goal; the integer-programming formulation rewards a second pass.
How to read it
First pass for the problem framing (why fewer scans matters); do a second pass on the integer-optimization setup and the overlap/personalization objective if you plan to implement or compare selection strategies.

Builds on

Built upon by

Nothing yet.

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 →