Abstract
Visual backward masking can be explained by two distinct processes, namely detection of low-level spatiotemporal correlations within a monocular pathway and object-level interference at a cyclopean stage. Here we investigated how these two potential mechanisms might impact on the phenomenon called “object substitution masking” that occurs due to the delayed delivery of four surrounding dots neither spatially adjacent nor similar to a target. We used a head-mounted display to compare dichoptic and monocular viewing conditions. In the dichoptic condition, four Landolt-C-like rings at the same eccentricity were presented to one eye, whereas a stimulus (referred to as a “masker”) comprised of four dots surrounding one of the rings was presented to the other eye. In the monocular condition, all these stimuli were presented to the same eye. Observers were asked to indicate which part of the ring (referred to as a “target”) surrounded by the dots was missing. For baseline conditions, only a single dot at the smallest eccentricity of the four dots was presented. We found a significant decline of correct responses from the baseline performances in some masker-target onset asynchronies as predicted from object substitution masking. This masking effect was quantitatively equivalent between the monocular and dichoptic conditions. Our results demonstrate that object substitution masking stems from interference at a cyclopean level, possibly in reentrant pathways from higher visual areas representing visual objects to lower areas representing images with high spatial resolution. In addition, we found that the single dot used in the baseline conditions deteriorated the performance to the same degree as the four dots did when it was presented at a nearer location to the missing part of the target, implying that multiple underlying mechanisms contribute to the visual masking in the current situation.
Acknowledgement: Supported by KAKENHI 18H05523