Abstract
Stereo depth cues can potentially provide a salient source of information to aid visual decision making. For example, in a visual search paradigm, stereo depth cues can literally cause part of an image to "pop out" at the observer (crossed binocular disparity cues). While some research has examined the mechanisms supporting search across various depth planes (Finlayson, et al., 2013), minimal research has examined the mechanisms engaged when stereo depth itself is the search target. The goal of this project was to examine the human information processing characteristics of stereo depth as a source of target information. We utilized the powerful measures and experimental methodology of Systems Factorial Technology (SFT; Townsend & Nozawa, 1995) to assess the workload capacity characteristics of stereo depth cues in a visual search paradigm. Standard group-level mean response time analyses indicated a nearly flat slope of mean target present response time as a function of distractor set size (although significantly nonzero) and a much larger slope of mean target absent response times. This pattern is characteristic of perceptual "pop-out" in visual search (Wolfe, 1998), and confirms that stimulus stereo disparity results in visual search "pop-out" effects. Furthermore, larger stereo disparity led to faster responses. Interestingly, workload capacity analyses at the individual participant level indicated worse than standard parallel performance with a filled array of distractors for nearly all participants. Participants fell into two groups with a sparser placement of distractors: most participants were still worse than standard parallel (limited capacity), but some were much better (super capacity). These results suggest that stereo pop-out visual search may utilize processing mechanisms that differ from a standard parallel architecture, and this architecture is influenced by the organization of the search display and varies across individuals. Distribution A: Approved for public release; distribution unlimited. 88ABW Cleared 12/02/2013; 88ABW-2013-5017.
Meeting abstract presented at VSS 2014