Recognition and discrimination: Is there a role for context in face learning?

Abstract

Recognizing unfamiliar identities in naturalistic images is challenging (e.g., two images of the same person are misperceived as belonging to different people). Face learning involves increased tolerance of variability in appearance and improved discrimination. I examined how a perceiver determines the range of inputs attributable to a newly learned identity, such that novel images of that identity are recognized, yet similar identities are excluded. I propose that learning a new face in the context of a similar identity facilitates learning via more precise representation. In two experiments, participants learned three identities (two similar, NNs; one dissimilar, FN) and were asked to recognize of two of those identities (one NN and FN). Performance did not vary for the NNs and FNs. Thus, identity learning involves both increased tolerance of variability and improved discrimination. We find no evidence that face learning is best accounted for by the multi-dimensional face space model.