David Ross (postdoc in Rosie Cowell’s Computational Memory and Perception Lab) will be presenting Norm-Based versus Exemplar-Based Models of Face Recognition in the Cognitive Brown Bag series in Tobin 521B at noon Wednesday, March 4. Everyone is welcome – the abstract is below.
Abstract: Face space models have successfully explained a range of findings in the face recognition literature. Debate has centered on whether the faces in face space are represented with respect to norms or exemplars. Findings from a number of face adaptation studies have been taken as conclusive support for a norm-based model, wherein faces are represented with respect to an abstracted norm/prototype, over an exemplar-based model, wherein faces are represented with respect to other exemplars. Here I will summarize work from my PhD that tested and refuted these claims using computational formalizations of norm and exemplar models. I will also present the results of some new simulations that indicate that it is actually norm-based models that are unable to account for the data.