Collisions and Attention
Citation:
Carol O'Sullivan, Collisions and Attention, ACM Transactions on Applied Perception, 2, (3), 2005, p309 - 321Download Item:
Abstract:
Attention is an important factor in the perception of static and dynamic scenes, which should, therefore, be taken into account when creating graphical images and animation. Recently, researchers have recognized this fact and have been investigating how the focus of attention can be measured, predicted, and exploited in graphical systems. In this article, we explore some preliminary strategies for developing an automatic means of predicting and exploiting attention in the processing of collisions and other dynamic events. Recent work on the perception of causality has shown that attention can change the way in which a dynamic scene consisting of collision events is perceived. We describe a series of experiments designed to determine the source of biases in the perception of anomalous collision dynamics and, in particular, whether attention plays a role. Using an eyetracker, eye-movements were recorded while participants viewed animations of simple causal launching events in 3D involving two colliding spheres. Results indicated that there was indeed a de?nite pattern to the allocation of attention based on the nature of the event, which is promising for the goal of developing a predictive metric. As a follow-up, a paper-based experiment was carried out in which participants were asked to sketch the predicted post-collision trajectories of the same two spheres printed on paper. These experiments demonstrated that attention alone was not suf?cient in determining performance, but rather the nature of the dynamic event itself also played a role.
Author's Homepage:
http://people.tcd.ie/osullicaDescription:
PUBLISHED
Author: O'SULLIVAN, ANN CAROL
Type of material:
Journal ArticleCollections
Series/Report no:
ACM Transactions on Applied Perception;2, 3Availability:
Full text availableKeywords:
Algorithms, Experimentation, Human Factors, Measurement, Animation, collisions, perception, eye movements, attentionMetadata
Show full item recordLicences: