Abstract
Synchrony provides critical information for perceiving spatiotemporal structures from parallel inputs. Human ability of synchrony-based feature binding can be accessed by asking observers to judge which features occur at the same time across two repeatedly changing stimuli in separate locations. In case of motion direction, the upper limit (critical temporal frequency) for this judgment is fairly high (~12 Hz) regardless of the spatial separation of compared local motions at least up to 100 deg (Maruya et al., VSS2010). This finding suggests that the long-range relationship between local motions may be rapidly encoded by special hardware. However, one might suspect that this hardware is not a novel mechanism, but what is already known as optic flow detectors, since most of the local-motion relationships that show rapid binding performance (e.g., up on the right and down on the left) are included in global rotation, expansion or translation. The critical difference between the two hypotheses is the range of spatial pooling. While we propose a mechanism that monitors only task relevant locations, optic flow detectors will respond to any relevant local motions presented in their large receptive field. To estimate the pooling range of the underlying mechanism, we measured the critical temporal frequency for a pair of vertically moving elements (targets) under concurrent presentation of masker elements. These maskers were designed to disturb the global rotation detector that could assist the task. Therefore, if the global rotation detector is indeed responsible, the critical frequency should significantly fall off by these maskers even when the separation from the targets was increased. However, the results showed that the magnitude of the masking effect gradually decreased as a function of the target-masker distance. These results are consistent more with local motion comparison in a point-to-point manner, than with global rotation detection.
Meeting abstract presented at VSS 2012