News on Views From Human and Computational Face Recognition

Although faces form a very homogenous stimulus class, adults are real experts in recognizing them. In the present study, we investigated to what extend the processing of such highly overlearned stimuli is dependent on viewpoint. Four experiments were conducted on the basis of the inter-extra-ortho experiments by Bülthoff and Edelman (1991), who used novel objects as stimuli (wire- and amoeba-like novel 3-D objects). First, in all experiments, systematic effects of viewpoint on face recognition performance were found that were consistent with computational approaches using interpolation of 2-D views. Second, sensitivity was better for horizontal vs. vertical views. Third, this effect was reduced in inverted faces, which indicates an important role of expertise in addition to effects of symmetry. The results are discussed within the framework of a new computational model based on key-frames, which entails local view interpolation and has been shown to be well suited to model human face recognition performance