We present S 3, a novel approach to generating expressive, animator-centric 3D head and
eye animation of characters in conversation. Given speech audio, a Directorial script and a …