3D FACE MODELING - 3D Face Processing: Modeling, Analysis and Synthesis - page 16

3. Future Research Direction of 3D Face Modeling
In the future, one promising research direction is to improve face modeling
tools which use one face image, along the line of the “linear class geometries”
work [Blanz and Vetter, 1999]. Improvements in the following aspects are
highly desirable.
3D face databases: The expressiveness of the “linear class geometries”
approach is decided by the 3D face model databases. In order to generate
convincing 3D face models for more people other than young Caucasian in
the Blanz and Vetter'99 database [Blanz and Vetter, 1999], more 3D face
scan data are needed to be collected for people of different races and different
ages.
Registration techniques for images and models: For the collected 3D face
geometries and textures, the corresponding facial points need to be aligned.
This registration process is required before linear subspace model can be
learned. This registration is also required for reconstructing a 3D face model
from an input face image. The original registration technique in [Blanz and
Vetter, 1999] is computationally expensive and need good initialization.
More recently, Romdhani and Vetter improved the efficiency, robustness
and accuracy of the registration process in [Romdhani and Vetter, 2003].
Recent automatic facial feature localization techniques can also help the
automatic generation of 3D face models [Hu et al., 2004].
Subspace modeling: When the 3D face database includes more geometry
variations, PCA may no longer be a good way to model the face geometry
subspace. Other subspace learning methods, such as Independent Com-
ponent Analysis (ICA) [Comon, 1994], local PCA [Kambhatla and Leen,
1997], need to explored the find better subspace representation.
Illumination effects of texture model: The textures of the 3D face models
also need to be collected to model the appearance variation, as in [Blanz and
Vetter, 1999]. Because illumination affects face appearance significantly,
the illumination effects need to be modeled. Besides the illumination models
in Blanz and Vetter's work [Blanz and Vetter, 1999], recent advances in
theoretical studies of illumination have enabled more efficient and effective
methods. In this topic, we present an efficient illumination modeling method
based on a single input face image. This method is discussed in details in
Chapter 6.