Abstract
Perceptual asymmetries in vision arise from the brain's preferential response to particular stimulus types at different retinal locations. Where the lower visual field (LVF) outperforms the upper visual field (UVF) in many aspects of low-level vision, increasing evidence suggests the reverse is true for high level face-processing (e.g., sex categorization, Quek & Finkbeiner, 2014; 2016), perhaps due to facilitated projection of UVF inputs to the ventral visual stream. Here we asked whether this UVF advantage extends to face categorization, i.e., the ability to rapidly discriminate variable faces from many other object categories. We recorded 128 channel EEG while 20 participants performed a demanding conjunction target-detection task at central fixation. Simultaneously, we presented two synchronised 6Hz streams of natural object images (e.g., buildings, animals, common objects) 4.3° visual angle above and below fixation. Unbeknownst to participants, we embedded face images in the upper and lower object streams at different periodic intervals (1/8 images vs. 1/10 images) within the one minute sequence. This allowed us to capture differential processing of faces vs. objects in the EEG spectrum at each hemifield's face-presentation frequency (i.e., 0.75Hz and 0.60Hz), providing a separate yet simultaneous index of UVF and LVF face categorization. Consistent with previous studies, the common visual response at 6Hz was characterised by a central occipital topography, whereas face-selective activity was strongest over right occipito-temporal channels. Importantly, and as predicted, this differential response to faces amongst objects was much stronger in the UVF than the LVF. A control experiment (N=12) in which we stimulated the UVF and LVF on separate trials indicated that this increased face-selective activity for UVF faces could not be attributed to a general increase in visual processing in this region. Taken together, our results suggest the UVF advantage demonstrated for face-sex discrimination also extends to rapid face categorization.
Meeting abstract presented at VSS 2018