Fig. 4
Evidence of Context Specificity During Recognition of Identities and Expressions. Rendered “difference” models. The top half represents identity templates (i.e., target was identity) comparing shape information differences when the expression was neutral (i.e., 0) versus one of six expressions. The bottom half represents expression templates (i.e., target was expression) comparing shape information differences when the identity was the male average (i.e., 0) versus one of two identities. All differences are displayed in red. 3D face images were rendered using a plugin from FaReT16 via MakeHuman15 v1.2 (www.makehumancommunity.org).
