FASE - Facial Analysis and Synthesis of Expressions
by Paul ten Hagen
The objective of the FASE project is to develop a system which can recognise the facial expressions and movements of a user in front of a camera and use this information to steer a remote animation model that will mimic, with life-like realism, the expressions and movements of the user.
Such a system will receive much attention from the entertainment industry as a new and efficient means of creating life-like animations. There is however, a much wider audience for this system (and the techniques in it) ranging from user interface designers, multimedia product developers, (mobile-) communications (coding techniques optimised to represent human faces like in video conferencing), criminal identification, distance learning, speech aids for the deaf, and surgical medicine.
The FASE project is a collaboration between CWI, Delft University of Technology (TUD), the Dutch Broadcasting Company, NOB, Philips Multimedia and Electronic (a leading 3D computer Graphics company in The Netherlands).
Two technologies play a pivoting role: Facial Analysis (FA and Facial Synthesis (FS). Traditionally both FA and FS are studied separately. The FASE project integrates both technologies in one system in order to improve the life-likeness of the estimated facial information and motion. The FA side will be covered by TUD. CWI will do the FS part and the integration.
There is close cooperation with users such as Philips and NOB, as well as with companies that can increase usability, marketing and distribution.
The Institute for the Deaf and the Institute of Phonetical Sciences will embed knowledge about how expressions are formed within the final system.
Paul ten Hagen - CWI
Tel: +31 20 592 4133