Models of eye movements in multiple object tracking with many objects

During everyday tasks, people often have to focus on several objects at once. As eye movements and attention are closely related, modeling eye movements can help us understand, where people attend to. In this study, we conducted an experiment in which participants had to track several moving objects at once while their eye gaze is being measured. We investigated, whether consistency of eye movements was affected by increasing number of objects on the scene during repeated presentations. We employed and extended several top-down models for predicting the eye gaze during such tasks. Results show that the presence of additional objects did not affect the coherence of scan patterns. The model accounting for the crowding effect yielded the best performance. Our results address the question of how the presence of visual clutter affects human eye behavior.