Abstract
Adaptation has been found to affect most of the perceived characteristics of the face. Yet the functional form of the aftereffects remains poorly defined because they are typically assessed only at a single level such as a category boundary. Age is unique among facial attributes in that observers can reliably quantify levels throughout the natural range of variation (by assigning an age in years). We took advantage of this to ask how adaptation to a given age altered the appearance of faces throughout the adult lifespan, to assess the actual pattern of the response changes. Observers assigned ages to 80 photographs of faces with physical ages ranging from 18 to 89, before or during adaption to a random sequence of 10 younger (30-38) or older (63-67) faces which were not part of the test set. Average settings across the 20 observers were highly correlated with physical age though slightly reduced in range (with of slope of ~0.85 for perceived vs. physical age). Adapting to the young or old set caused all faces to appear older or younger, respectively, by roughly a constant amount (~+2.5 yrs). The aftereffects were thus well approximated by simple additive shifts with little change in slope. Importantly, this included biases in the adapting age range itself, so that observers were not simply becoming less sensitive to the adapting level. These results suggest that at least for the attribute of age, the effects of adaptation are consistent with a global and linear renormalization of the dimension, and thus with a recentering of face space.
Meeting abstract presented at VSS 2013