Levels of Visual Information Processing: Perception of Dynamic Properties and Events

Abstract

Objective: The here presented studies explore automatic and controlled per- ceptual processes in two dynamic paradigms and support a rationale of a multi- level approach to dynamic visual perception. Method: I investigate different perceptual levels of dynamic scenes, including factors within the perceiver, within the objects, and within the environment. Automatic processes are explored with a simple 3-D tracking task and an event perception recognition task; controlled processes are observed in a modified tracking task with specific object properties and an identification task. Results: Through analysis of tracking and report errors measured in the two paradigms, I observed similarities in the automatic processing of artificial 3-D tracking environments (Study 1: the scene-based relations are more important than positions of individual objects) and real-life video clips (Study 2: core as- pects are preferred over fine details). Despite the assumption that tracking is a cognitive-impenetrable mechanism, results of the modified tracking task (Study 3) point towards the ability of participants to strategically weigh visual information based on task-demands. Conclusion: The results of this dissertation illustrate that the identification of influential internal and external factors is important to enhance our under- standing of the multidimensional nature of perception – an understanding that will eventually and hopefully bring research to move beyond questions of how resources are limited, and start to focus on fundamental issues like how we can use mental resources to our benefit.