Abstract
Face perception depends on two sources of information 3D shape and surface-based cues (texture, color). According to behavioral studies, both of them equally contribute to discrimination of individual faces (O'Toole, Vetter & Blanz, 1999). However, these behavioral measures can be contaminated by many factors, not allowing to distinguish and to quantify the respective contribution of each source of information. Using an established method of manipulating facial images (Tiddeman, Burt, & Perrett, 2001), four individual faces (2 males) were morphed with another 10 same-sex faces varying shape only, texture only or both. Electroencephalogram (EEG) was recorded from 10 participants during a rapid periodic oddball visual stimulation, providing an objective, implicit and robust quantifiable measure of visual discrimination (Liu-Shuang, Norcia & Rossion, in press). Stimuli were presented in 80 sec long trials, in which the same face (one of the 4 faces) was shown four times consecutively and the fifth face (the oddball) was one of the corresponding morphed faces, thus resulting in a sequence AAAABAAAACAAAAD. The base frequency F was 5.88 Hz and the oddball frequency 1.18Hz (5.88Hz/5) and its harmonics (nF/5) were used to measure individual face discrimination. This individual face discrimination response was observed at occipito-temporal sites, particularly over the right hemisphere. While shape was also discriminated at right occipito-temporal electrode sites, surface information was coded bilaterally (Jiang et al., 2009). However, and most importantly, shape and texture changes alone were associated with much weaker responses than when both sources of information were combined, revealing a supra-additive effect of the contribution of the two facial aspects to the discrimination of individual faces. Thus, these results suggest that the two kinds of information combined are necessary to provide a full face identity, i.e. face identity being more than the sum of the contribution of shape and surface cues.
Meeting abstract presented at VSS 2014