Abstract
Individual differences in face recognition accuracy are likely to be linked to the way visual information is sampled and processed. Here we compared visual sampling of super-recognisers (SRs) – individuals that achieve the highest levels of accuracy in face recognition tasks – to typical viewers, using a novel gaze-contingent technique. Participants performed a face recognition task in which they learned and recognised novel faces while their gaze position was recorded. The face on the screen was modified in real-time to constrict the information around the gaze position at different aperture sizes. Super-recognisers displayed superior recognition accuracy for all but the smallest aperture viewing sizes. Underlying this superiority are qualitative differences in visual sampling: (i) SRs exhibited greater distribution of fixations across face images, suggesting enhanced visual exploration; (ii) SRs focused less on the eye region; (iii) SRs produced more fixations to the central region of faces. Importantly, these differences were most apparent in the learning phase of the experiment, suggesting that the superior accuracy of SRs was founded on enhanced encoding of faces into memory. Together, our results point to a process whereby SRs construct a more robust memory trace by accumulating samples of complex visual information across successive eye movements. Because super-recognisers display superior accuracy with restricted viewing – while also showing fixation patterns that are associated with holistic processing – SR’s superior performance appears to achieved by combining both local and global sources of information in memory representations.