Title :
Real-time facial motion analysis for virtual teleconferencing
Author_Institution :
British Telecom Res. Labs., Ipswich, UK
Abstract :
Image analysis for a virtual teleconferencing system is described. Eyes are tracked in real-time using multiresolution grey-scale correlation. Locating the eyes allows the positions of other facial features to be approximated. The head orientation is calculated by comparing the eye locations in the video images to the positions of the top and sides of the head. The direction and gesture parameters are transmitted to a three dimensional animated `clone´ in a virtual office or virtual conferencing room
Keywords :
computer animation; face recognition; feature extraction; image recognition; image resolution; motion estimation; office automation; real-time systems; teleconferencing; direction parameters; eye location; eye tracking; facial feature approximation; gesture parameters; head orientation calculation; image analysis; multiresolution grey-scale correlation; real-time facial motion analysis; three dimensional animated clone; video images; virtual office; virtual teleconferencing; Animation; Cloning; Eyes; Facial features; Head; Image motion analysis; Motion analysis; Teleconferencing; Teleworking; Videoconference;
Conference_Titel :
Automatic Face and Gesture Recognition, 1996., Proceedings of the Second International Conference on
Conference_Location :
Killington, VT
Print_ISBN :
0-8186-7713-9
DOI :
10.1109/AFGR.1996.557288
