3D facial geometry reconstruction using patch database

Tsukasa Nozawa, Takuya Kato, Pavel A. Savkin, Naoki Nozawa, Shigeo Morishima

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

3D facial shape reconstruction in the wild environments is an important research task in the field of CG and CV. This is because it can be applied to a lot of products, such as 3DCG video games and face recognition. One of the most popular 3D facial shape reconstruction techniques is 3D Model-based approach. This approach approximates a facial shape by using 3D face model, which is calculated by principal component analysis. [Blanz and Vetter 1999] performed a 3D facial reconstruction by fitting points from facial feature points of an input of single facial image to vertex of template 3D facial model named 3D Morphable Model. This method can reconstruct a facial shape from a variety of images which include different lighting and face orientation, as long as facial feature points can be detected. However, representation quality of the result depends on the number of 3D model resolution.

Original languageEnglish
Title of host publicationSIGGRAPH 2016 - ACM SIGGRAPH 2016 Posters
PublisherAssociation for Computing Machinery, Inc
ISBN (Electronic)9781450343718
DOIs
Publication statusPublished - 2016 Jul 24
EventACM International Conference on Computer Graphics and Interactive Techniques, SIGGRAPH 2016 - Anaheim, United States
Duration: 2016 Jul 242016 Jul 28

Publication series

NameSIGGRAPH 2016 - ACM SIGGRAPH 2016 Posters

Other

OtherACM International Conference on Computer Graphics and Interactive Techniques, SIGGRAPH 2016
CountryUnited States
CityAnaheim
Period16/7/2416/7/28

Keywords

  • 3D reconstruction
  • Shape from X
  • Texture synthesis

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Human-Computer Interaction
  • Software

Fingerprint Dive into the research topics of '3D facial geometry reconstruction using patch database'. Together they form a unique fingerprint.

Cite this