PNNL

Abstract

In this paper we discuss RAPID (Revolutionary Accelerated Processing of Interest Detection), a research project that blends electroencephalogram and eye tracking technology in an attempt to improve the imagery analysis process. Imagery analysts, like many other specialized analysts in the intelligence community and knowledge workers everywhere, are inundated with information. As collection strategies and technologies improve, the backlog of imagery to be reviewed grows and restricts attempts to use imagery strategically. While many approaches to improving intelligence analysis have focused on replacing the trained eye of the analyst with complicated algorithms and learning techniques, our approach has been to utilize those strengths that make the human vision system so powerful. In effect, we use the analyst as a powerful image-processing engine, inferring from their eye movements and brain patterns salient features that can be investigated further by the analyst or passed along for further analysis using traditional techniques. We believe that our approach can be generalized and applied to other HCI situations such as adaptive user interfaces that can throttle information in tune with the cognitive load of the user. Here we discuss our approach and review the result of some recent experiments on identifying salient features.