AN ACTIVE VISION APPROACH TO UNDERSTANDING AND IMPROVING VISUAL TRAINING IN THE GEOSCIENCES (Invited Presentation)

We are using an Active Vision approach to learn how novices and expert geologists acquire visual information in the field. The Active Vision approach emphasizes that visual perception is an active process wherein new information is acquired about a particular environment through exploratory eye movements. Eye movements are not only influenced by physical stimuli, but are also strongly influenced by high-level perceptual and cognitive processes. Eye-tracking data have been collected on novices (undergraduate geology students) and experts during a 10-day field trip across California. The field trip is the centerpiece of a course at the University of Rochester that introduces students to geology through examples found in California. Topics that are emphasized on the trip include the change from the subduction to transform margin and neotectonics. In addition to eye-tracking data, high-resolution panoramic images were captured at each key locality for use in a semi-immersive laboratory environment. Examples of each data type will be presented. We will review efforts to quantify patterns of individual eye-tracking records (including gaze time, gaze pattern and geologic feature recognition), development of semi-immersive environments to display geologic scenes, and initial eye-tracking efforts in the semi-immersive environment. The research is a collaborative effort between Earth scientists, Cognitive scientists and Imaging scientists at the University of Rochester and the Rochester Institute of Technology, with funding from the National Science Foundation.