Abstract
This paper presents a facial expression recognition-synthesis system. In the front, it detects a facial area within the given image and then classifies its facial expression into 7 emotional weightings. Such weighting information, transmitted to the PDA via a mobile network, is used for non-photorealistic facial expression animation. The cartoon-like shading method, adopted as a non-photorealistic 3-D technique, is developed to render a 3-D avatar that conveys a familiar and yet unique facial character, even without employing extensive polygons. We found that facial expression animation using emotional curves is more effective in expressing the timing of an expression comparing to the linear interpolation method. The present emotional avatar embedded on a mobile platform can be used in the cyberspace.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have