Guiding Attention in Real-World Scenes
University Of California At Davis, Davis CA
Investigators
Linked publications & trials
Abstract
Project Summary Real-world scenes contain far more information that we can perceive and comprehend at any given moment. A key mechanism for making real-world scene perception tractable is visual attention?the mechanism of preferentially processing only part of the scene at any given time. What we attend to in a scene determines what we see, understand, and remember. Attention is guided by both the visual properties of the scene itself and by our knowledge about similar scenes and the world in general. How knowledge is used to guide attention through a meaningful scene remains largely unknown. The central idea behind this proposal is to address this fundamental scientific question by focusing on two critical aspects of scene knowledge: knowledge about where a given object is likely to appear in a scene, and knowledge about which regions and objects in a scene are meaningful and informative. The studies aim to determine how spatial and meaning constraints are used to guide attention in scenes. This proposed research is innovative in combining high-resolution eyetracking with novel experimental paradigms for manipulating and measuring knowledge-based constraints. First, a new fusion of spatial learning methods with eyetracking is used to study the influence of spatial knowledge on attentional guidance. Second, new quantitative scene-rating and information-theoretic metrics are used to index meaning in scenes, providing a new theoretical approach to scene meaning and new empirical tools for investigating meaning. Third, real- time scene manipulation based on the viewer?s eye movements is combined with manipulations of spatial and meaning constraints to investigate how quickly knowledge about a scene becomes available to guide attention. The project is significant in challenging current models to explain the role of knowledge in guiding attention in scenes. The experiments are designed to advance the field regardless of the outcome, and will provide rich and theoretically constraining results that may have a transformative effect on current theory. In addition, the proposed research has important translational implications because deficits in attention and perception are suffered by many psychiatric and neurological populations. By understanding how knowledge influences the guidance of attention in real scenes, the proposed studies can ultimately lead to the development of targeted rehabilitation strategies for the real world that better capitalize on both disrupted and spared functions.
View original record on NIH RePORTER →