Fig. 1: Estimating spatial relations via eye movements.
From: Inferring visual space from ultra-fine extra-retinal knowledge of gaze position
A–C Experimental design. A Subjects reported the spatial configuration of a Vernier (left or right) viewed through a retinally-stabilized aperture. B The aperture moved together with the eye, to allow stimulation of only a thin vertical strip on the retina. The width of the aperture was equal to that of each bar in the Vernier (28\({}^{{\prime} }\) long; 1.4\({}^{{\prime} }\), the angle covered by one pixel on the CRT). C In this way, each Vernier bar was visible only when it directly overlapped with the aperture, resulting in vertically-aligned bar exposures on the retina. D Motor knowledge of eye movements is required to accomplish this task. The same visual input signals can be obtained with different configurations of the stimulus, when the eye drifts in opposite directions. E Example trace of eye movements in a trial. The shaded green regions mark the periods of exposure of each Vernier bars. The pink region indicates the inter-stimulus interval (ISI), here 100 ms. F–I Ocular drift characteristics and performance in the task. Data from N = 6 human observers. F Mean eye speed and displacement are virtually identical to those measured in the same subjects while fixating on a marker. Shaded regions represent ± one SEM across subjects. G Average probability distribution of gaze displacement in between bar exposures. H, I Subjects correctly reported the configuration of the stimulus. Both proportion of correct responses and discriminability index were significantly above chance (H: **p = 3.16 × 10−4; ***p = 3.44 × 10−6; I: **p = 5.02 × 10−4, ***p = 9.16 × 10−6, two-tailed t-test) and improved as the Vernier gap increased (H: *p = 0.0024; I: *p = 0.0016, paired two-tailed t-test). Gray circles are the individual subjects data. Diamonds and associated error bars represent averages ± one SEM across subjects. Source data are provided as a Source data file.
