Temporal dynamics of feature integration in peripheral vision and saccadic eye movement
23.4023, Saturday, 16-May, 8:30 am - 12:30 pm, Pavilion
Masahiko Terao1, Ikuya Murakami1; 1The University of Tokyo
In peripheral vision, a stimulus surrounded by multiple stimuli in close proximity can appear to look similar to the nearby stimuli (e.g., Greenwood et al, 2009). This perceptual assimilation also affects the landing point of a saccade (Terao & Murakami, 2014, APCV), indicating that both the visual and oculomotor systems integrate feature information over a certain spatial window. Here we investigated the temporal dynamics of these phenomena. A target cross and two flanker crosses vertically surrounding the target were presented at 12 degrees to the right of the fixation point. The crossing points of the flankers were deviated leftward or rightward. We varied the stimulus onset asynchrony between the target and the flankers and asked observers to judge the apparent position of the crossing point of the target. The results showed that the flankers presented prior to the target were not effective whereas the flankers presented at or after the target onset biased the horizontal position of the target’s crossing point toward the flankers’. The temporal window of this effect was approximately 200 ms, with the maximum effect obtained at 25 ms after the target onset. In another experiment, observers were asked to make a saccade to the non-deviated crossing point of the target as soon as its onset. The results showed that the saccadic landing position was horizontally biased toward the deviated crossing points of the flankers even when they were presented 75 ms after the target onset. The effect was seen even when the flanker onset was only 80–40 ms earlier than the saccade onset. Our findings indicate that the visual and oculomotor systems integrate feature information not only over a large space but also over a long time in peripheral vision and that it is within a very short latency that the integration affects eye-movement statistics.