Synthesis and Analysis Techniques for the Human Body: R&D Projects - 3D Modeling and Animation: Synthesis and Analysis Techniques for the Human Body - page 357

Emotional facial animation: The first activity carried out was related to the
enhancement of the facial expressions synthesis. With the help of an artist,
a database of high-level expressions has been developed by mixing low-
level expressions. A second activity was related to the development of a
facial animation engine based on Facial Animation Tables (FAT), in order
to accurately reproduce facial animations. For each Facial Animation
Parameter (FAP), we need the FAT to control the displacements of the
vertices corresponding to the Facial Description Parameter (FDP). The
FAT can be defined with an animation engine or by a designer. The
advantages of the FAT-based animation are that the expected deforma-
tions due to the animation are exactly mastered, the FAT can be defined by
a designer and, thus, the animation can be balanced between realistic and
cartoon-like behavior. The drawback is that each FAT is unique to a given
face and it is required to build one FAT for each face (even if the topology
of the mesh is the same), whereas an MPEG-4 animation engine can work
with any face mesh and FDP data.

Emotional body animation: The server calculates the emotion from all man-to-
machine action (text, speech and facial expression) and associates to this
emotion a gesture that is translated into the corresponding Body Animation
Parameters (BAP) file that is finally sent through Internet.

STAR

The (IST STAR) project Service and Training through Augmented Reality
(STAR) is a collaboration between research institutes and companies from
Europe and the USA. STAR's focus is the development of Mixed Reality
techniques for training, documentation, and planning purposes.
To achieve these, the following components were developed (Figure 7):

• Automated 3D reconstruction of industrial installations,

• Automated view selection and camera handover,

• Manipulation of mixed objects by virtual humans,

• Approaches with different tracking and recognition of objects from video,

• User interface to create own Augmented Reality applications.

Using these components, it is possible to realize the following scenario: A worker
is on-site preparing to perform a maintenance task. He is equipped with a laptop