Abstract
[Introduction] People commonly track objects moving in complex natural displays and their performance in the “multiple object tracking” (MOT) paradigm has been used to study this visual attention task for over three decades. Given the theoretical and practical importance of object tracking, it is critical to understand how people solve the correspondence problem to track objects; however, it remains unclear what information people use to achieve this feat. In particular, while people can track multiple moving objects based on their positions, there is ambiguity whether people can track objects via higher order kinematic information such as velocity. [Method] We explicitly test whether people can use instantaneous velocity direction to solve the correspondence problem during multiple object tracking when positional information is completely ambiguous. And we further test whether acceleration direction can also be used for tracking when object position and instantaneous velocity direction are both uninformative. [Results] We show that people use velocity direction when it is necessary to disambiguate targets from distractors. However, tracking via velocity is less accurate than tracking via position alone, indicating that velocity simply does not provide as an informative cue as position. Furthermore, observers fail to use acceleration to track objects, even though they can reliably detect the acceleration cue. This pattern of results indicates that the failure to use acceleration during object tracking reflects which kinematic information is used to resolve correspondence. Finally, we demonstrate that the extent to which velocity, but not acceleration, is used varies with “kinematic load” – the number of object pairs requires velocity information simultaneously. Together, these results indicate a hierarchy of kinematic information for tracking – position is most precise and most useful, but velocity can still be used to solve the correspondence problem; acceleration, in contrast, seems to be largely not used for multiple object tracking.