VIDAS: Video assisted with audio coding and representation

• Research unit: EU RFP
• Project number: 95.0689-2

Key words (English)

Communication; facial features; facial animation; MPEG-4

Alternative project number (English)

EU project number: AC057

Research programs (English)

EU-programme: 4. Frame Research Programme - 1.2 Communications technologies

Short description (English)

See abstract

Partners and International Organizations (English)

DIST-University of Genova (I), MATRA (F), ELMER (I), Linkoping University (Sweden), Universitat Politecnica Catalunya (Spain), INRIA (F), EPFL (Switzerland), University of Geneva (CH), PHILIPS (UK), National Association for the Deaf (UK), Aristotle University of Thessanloniki (GR)

Abstract (English)

The main objective of VIDAS is to improve the subjective quality of audio-visual communication at very low bit rate. During this period of the project, the University of Geneva has contributed to the activities in Work Package (WP) 3 and 5.
In activity 3.1 we provided a whole chapter to the deliverable d.35, describing the methods and algorithms we use and have developed to reconstruct and animate a synthetic face in relation with the MPEG-4 SNHC definition of FDPs and FAPs.
We have worked on development of new mesh deformation for face that will enable tighter compliance with the standard and is expected to give optimized speed for real-time facial animation. This has resulted in low-level deformation software, which can be customized to various objects defined by feature points. This software, customized for MPEG-4 face object has been tested well for real time facial animation.
In activities 5.2 we have concentrated on the participation in the activities of MPEG-4 and the development of compatible facial and body animation and deformation software. We have participated in all meetings of MPEG, submitted several documents, taken part in Core Experiments and kept an activity on the SNHC-FBA reflector. In 5.2 we also have integrated several tools developed in the previous activities in order to have a synthetic head capable of interpreting FAPs files, phonemes and expressions in a real-time facial deformation. They have been integrated in an interactive user-friendly interface. The different types of data can be activated simultaneously at any time and blended to a continuous facial animation. Some efforts have been put into the development of a good blending method and of a realistic real-time phoneme-based speech implementation.
We have published our work in various international conferences and journals (Computer Animation 99, MultiMedia Modeling '98, Computer Graphics and Applications '98, Proc. of the IEEE. Vol.86, WSCG'99).

References in databases (English)

Swiss Database: Euro-DB of the
State Secretariat for Education and Research
Hallwylstrasse 4
CH-3003 Berne, Switzerland
Tel. [phone]
Swiss Project-Number: 95.0689-2