Abstract
The face viewpoint aftereffect (FVA) is an illusion that biases the perception of a frontward facing test face in the opposite direction to the adapting face (Fang & He, 2005). Changes in the outer contour of the head that highlight these changes in face viewpoint can be modelled using radial frequency (RF) patterns (Wilson et al., 2000). We used RF patterns together with a recently developed procedure to tease apart the FVA from observer response bias. The psychophysical procedure we used was necessary because traditional methods for measuring perceptual illusions confound perceptual changes, with changes in an observer’s criterion (Morgan, 2013; 2014). The contours describing the shape stimuli had a luminance profile that followed the fourth derivative of a Gaussian. Participants adapted to two vertically aligned, contrast reversing (1Hz) stimuli facing in opposite directions. Eight separate test conditions were randomly interleaved, each with two static test stimuli, vertically aligned to the same spatial position of the adaptors. Test stimuli could face in either the same or opposite directions and allowed us to make tangible predictions to distinguish adaptation from response bias. A two-alternate forced choice adaptive staircase procedure was used in which the participants indicated which of the two test stimuli appeared most asymmetric. Results showed a large FVA from RF patterns that could not be attributed to shifts in observer response bias. By varying the size of the test stimuli we were able to rule out a purely retinotopic account of adaptation. Halving the size of the test stimuli reduced the FVA by ~50%, suggesting an extra-striate locus of adaptation. These results are consistent with the proposal by Wilson et al., (2000) which suggests an extra-striate locus for the processing of face viewpoint from RF patterns.
Meeting abstract presented at VSS 2015