Effects of smooth pursuit and second-order stimuli on visual motion prediction.

The purpose of this study was to determine whether smooth pursuit eye movements affect visual motion prediction using a time-to-contact task where observers anticipate the exact instant that a partially occluded target would coincide with a stationary object. Moreover, we attempted to clarify the influence of second-order motion on visual motion prediction during smooth pursuit. One target object moved to another stationary object (6 deg apart) at constant velocity of 3, 4, and 5 deg/s, and then the two objects disappeared 500 ms after the onset of target motion. The observers estimated the moment the moving object would overlap the stationary object and pressed a button. For the pursuit condition, both a Gaussian window and a random dots texture moved in the same direction at the same speed for the first-order motion, whereas a Gaussian window moved over a static background composed of random dots texture for the second-order motion. The results showed that the constant error of the time-to-contact shifted to a later response for the pursuit condition compared to the fixation condition, regardless of the object velocity. In addition, during smooth pursuit, the constant error for the second-order motion shifted to an earlier response compared to the first-order motion when the object velocity was 3 deg/s, whereas no significant difference was found at 4 and 5 deg/s. Therefore, our results suggest that visual motion prediction using a time-to-contact task is affected by both eye movements and motion configuration such as second-order motion.