We synthesize whole-physique human images beginning from a supplied human pose with two focused steps. one) With some texts describing the shapes of clothes, the presented human pose is very first translated to your human parsing map. 2) The ultimate human impression is then generated by providing the procedure with https://www.humanizingaitext.com/