作者: F. Lavagetto , R. Pockaj
DOI: 10.1109/76.752095
关键词:
摘要: We propose a method for implementing high-level interface the synthesis and animation of animated virtual faces that is in full compliance with MPEG-4 specifications. This allows us to implement simple facial object profile part calibration profile. In fact, starting from wireframe set configuration files, developed system capable automatically generating rules suited model driven by stream parameters. If parameters (feature points texture) are available, able exploit this information suitably modifying geometry performing its means calibrated computed ex novo on adapted somatics model. Evidence achievable performance reported at end paper figures showing capability reshape according decoded effectiveness expressions.