Skip to content

← ArchivePaper2015

Dynamic 3D Avatar Creation from Hand-Held Video Input

Alexandru-Eugen Ichim, Sofien Bouaziz, Mark Pauly

SIGGRAPHAcademic276 cites2 descendantsFacialRigging

System for creating personalized animatable 3D face avatars from casual handheld video using blendshape fitting and rig transfer.

Abstract

We present a complete pipeline for creating fully rigged, personalized 3D facial avatars from hand-held video. Our system faithfully recovers facial expression dynamics of the user by adapting a blendshape template to an image sequence of recorded expressions using an optimization that integrates feature tracking, optical flow, and shape from shading. Fine-scale details such as wrinkles are captured separately in normal maps and ambient occlusion maps. From this user- and expression-specific data, we learn a regressor for on-the-fly detail synthesis during animation to enhance the perceptual realism of the avatars. Our system demonstrates that the use of appropriate reconstruction priors yields compelling face rigs even with a minimalistic acquisition system and limited user assistance. This facilitates a range of new applications in computer animation and consumer-level online communication based on personalized avatars. We present realtime application demos to validate our method.

How to read this

Category
Method / system: personalized facial avatar creation from video
Contributions
  • A complete pipeline producing fully rigged, personalized 3D facial avatars from hand-held video
  • Blendshape-template adaptation via an optimization integrating feature tracking, optical flow, and shape from shading, with wrinkle-scale detail captured in normal and ambient-occlusion maps
  • A learned regressor for on-the-fly detail synthesis during animation, demonstrated in real-time application demos
Context
Sits in the consumer-level facial-capture and blendshape-rig lineage, emphasizing reconstruction priors that let a minimalist acquisition setup still yield compelling rigs.
Correctness
Quality hinges on appropriate reconstruction priors and on the recorded expression range; with a minimalistic rig and limited user assistance, results depend on capture coverage and may degrade for expressions or details not observed in the input.
Clarity
Readable end-to-end system description; one pass conveys the pipeline, a second covers the optimization and detail-regressor stages.
How to read it
Read for the staged optimization (tracking plus flow plus shape-from-shading) and the offline-detail vs runtime-synthesis split; second pass if you need to reproduce the regressor.

Builds on

Nothing in the archive, this is a starting point.

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 →