Abstract
There is a considerable amount of evidence that different kinds of objects are processed differently by the visual system and brain, but there is a need for more mechanistic explanations for those effects. One example is the face inversion effect. Many researchers have found that there is an advantage for upright faces compared to upside down ones that is not observed with other types of objects (Rossion, 2008; Yin, 1969). A common explanation for these effects is that upright faces are processed by a holistic system while everything else – including inverted faces -- goes through a featural system (Maurer, Grand, & Mondloch, 2002; McKone, 2009). An alternative account is that there exists a single system and that variation in perception of different kinds of objects is due to differential tuning of the parameters of that system (Sekuler et al., 2004). Here, we examined these ideas with a formal Drift Diffusion model (Ratcliff, 1978; Ratcliff & Rouder, 1998). If upright and inverted faces are processed by different systems, then one would expect qualitatively different patterns of parameter values, with parameters for inverted faces matching objects (both upright and inverted) and parameters for upright faces being unique. However, based on the single-system account, we hypothesized that lower performance with inverted faces would reflect either a lower rate of evidence accumulation (drift rate) or a higher decisional threshold (criterion). The modeling results showed that the large inversion effects found with faces were mediated solely by the rate of perceptual evidence accumulation – none of the other parameters changed across orientation. This suggests that inverted faces are processed by the same system as upright faces and that the inefficiency associated with the face inversion effect is due to a slower accumulation of perceptual evidence.
Meeting abstract presented at VSS 2013