A Morphable Model for the Synthesis of 3D Faces

Seminar über aktuelle Themen der Computergraphik
Marco Nef; Prof. Dr. M. Gross, ETH Zürich; Martin Roth
SS 2000

Abstract

At SIGGRAPH '99 Volker Blanz and Thomas Vetter introduced a very new way to produce a morphable model from a single photograph. Their idea is to manually and roughly align the projection of an average 3D head to the shape of the face on the picture. An automatic algorithm then constructs a total correspondance between the picture and the projection. What you get by that is a very realistic 3D face model of the face on the picture. With this morphable model you can do almost everything, e.g. change illumination and shape, rotate the head, add additional 3D objects, change facial expressions,... The results are just amazing.

References

Volker Blanz, Thomas Vetter
A Morphable Model for the Synthesis of 3D Faces
SIGGRAPH '99

F. Pighin, J. Hecker, D. Lischinski, R. Szeliski, D. Salesin
Synthesizing realistic facial expressions from photographs
SIGGRAPH '98