Perceptive user interfaces workshop

Abstract

One of the shortcomings of traditional human-computer interaction is its failure to be tailored specifically for human cognition. Human cognition has particular virtues and limitations; designing interfaces to maximize the virtues and bolster the limitations could produce substantial gains in information management capability. These cognition-centric design principles strive to move beyond mere human-computer interaction and towards human-computer symbiosis – a catalytic marriage between the heuristic-driven, contextsensitive powers of human cognition and the detail-oriented, data crunching might of computer computation. This symbiosis becomes feasible due to progress made during the “Decade of the Brain” in expanding the understanding of brain mechanisms and introduction of novel non-invasive assessment tools (such as fMRI), the ongoing “Cognitive Revolution” in behavioral science producing advances in the science of problems solving, reasoning, and decision making, and the growth of digital technologies in pure computing power, miniaturization and ruggedization, data mining sophistication, and evolving advancements in robust input/output devices. These advances produce four significant content domains: multimodal interaction, shared context, interested management, and a new generation of human factors issues. Traditional computer systems rely almost solely on visual information (with a meager auditory component) – future systems will be inherently multimodal, relying on all sensory and motor processing channels for receiving and conveying information. Traditional computer systems also are restricted because humans and computers operate within different contexts – computers are wholly unaware of cues that humans give the highest priority or how to capitalize on those cues to help humans better process information. Similarly, computers lack the ability to truly ‘serve’ the user and determine what information in an environment should be omitted, what should be highlighted, and what should be portrayed with accuracy (and what determines sufficient accuracy). Finally, the advent of these new tools in the human and computer domain requires a new generation of human factors design issues be addressed. Our panel has LCDR Dylan Schmorrow acting as Chair. LCDR Schmorrow is the program officer for the Defense Advance Research Projects Agency Information Technology Office’s “Augmented Cognition” program. LT Jim Patrey will act as cochair; LT Patrey is the Director of International Programs for DARPA’s “Augmented Cognition” program. The proposed panel will each address one of the four content areas identified as significant within “Augmented Cognition.” Dr. Phil Cohen of the Center for Human Computer Interaction will address issues of multimodality. Dr. Denny Proffitt will discuss the value of context as an augmentation tool. Dr. Mike Zyda, of the