Attentional synchrony in static and dynamic scenes

When multiple viewers attend to the same static naturalistic scene there is a high degree of agreement in terms of the regions viewers attend to but no synchronization of when they attend there. By comparison, real-world scenes are rich with temporally-defined visual events such as motion which have the potential to involuntarily capture attention. Initial studies of attention during dynamic scenes have reported a high degree of spatiotemporal agreement, henceforth referred to as Attentional Synchrony. However, these dynamic scenes are composed with viewer attention in mind e.g. film and TV footage depicting a single subject tracked by a moving camera. Therefore, it is not currently known if attentional synchrony is caused by the presence of motion, scene content, or compositional factors.

In this study the degree of attentional synchrony was measured while participants memorized static and dynamic versions of the same real-world scenes. All scenes were filmed by a static camera without any deliberate composition. As predicted, the degree of attentional synchrony was greater in dynamic compared with static versions of the same scene. On average, the gaze position of 6 participants was clustered within 15% of the screen area during dynamic scenes and 20% for static scenes. Both static and dynamic scenes began with a 1200ms period of high synchrony ([[lt]]8%) as all participants attended to the screen center. Following this central-tendency the greatest moments of attentional synchrony appear to be caused by the presence of a single moving person (one person synchrony = 10%; more than one person = 17%). People are prioritized by attention in static scenes but do not create synchrony. These results indicate that there is greater attentional synchrony in dynamic compared with static naturalistic scenes when content and composition are controlled.