Drive Muscle Face Model Using Single Video
Cai Shang-shu · Acta Simulata Systematica Sinica · 2003
The paper presents a method to drive muscle-based facial model using single video. Compared with other conventional methods which drive facial model by using multi-view video sequences, this method first gets moving parameters of facial muscle by tracking specific control nodes on human face, then displacements of all nodes are gotten on facial model. In this way we can drive muscle-based facial model only using single video. We have gotten realistic human expressions and mouth shape by this method.