Abstract
Motion perception can be distorted (Filehne, 1922) or enhanced (Spering, Schütz, Braun, & Gegenfurtner, 2011) by smooth pursuit. Here we investigated the role of smooth pursuit in discriminating curvature of motion trajectories. Subjects viewed a white 0.5 deg diameter Gaussian blob on a black background in total darkness as it moved along an arc of constant curvature for 1s (standard). Then after a 1.5s delay, a second motion trajectory (comparison) was viewed for 1s after which subjects used a button press to report which path appeared "flatter". No feedback was given. Viewing condition was blocked. Subjects either fixated a point at the center of the motion trajectory or were instructed to smoothly follow the target as it moved across the screen. In order to prevent the use of the fixation point as a spatial reference, the comparison trajectory was randomly rotated away from the standard by +/- 0, 5,20, or 90 degrees. Psychophysical discrimination thresholds were lower during pursuit (M = 1.09,SD = 0.27o) compared to fixation (M = 1.33,SD = 0.44o), where M and SDs represent differences in radius of curvature between standard and comparison. We evaluated oculomotor curvature discrimination by calculating oculometric functions, with curvature judgments derived from the average signed distance between each de-saccaded unfiltered position sample in a particular interval and a line connecting the first and last points of the interval. Larger distances indicated more curvature. The results indicated that a window of 300 ms beginning at the start of steady state was required to successfully decode curvature from eye positions. Even then, pursuit thresholds never reached psychophysical thresholds and on average were larger by a factor of 3. These results indicate that smooth pursuit may be useful in discriminating curvature in motion, even though the pursuit itself indicates little curvature.
Meeting abstract presented at VSS 2016