In this thesis, we implemented model creation system and face animation system for intelligent agent interface.
In model creation system, we build the 3-D model of a specific person, and reduce the user's input significantly by image processing technique. For given front and side images, the system modifies a generic model to a specific model. To modify more exactly, we must extract more feature points.
The animation system generates lip animation for given sentences and synthesizes voice synchronously. For the efficient transfer of emotion, it produces the facial expression. We implemented texture mapped and stereo graphic images for realistic effect.
We applied our system to voice-mail-reader, say-time and explanation system. Our system may be used for any systems that need to transfer messages. For more sophisticated applications, we must consider recognition of facial expression , gesture and voice of users.