CAREER: Disengaging attention from distractions

William Marsh Rice University, Houston TX

Investigators

Abstract

Human attention is limited and when distractions occur, they frequently result in accidents and other adverse events. Recent advances in Human-AI teaming aim to overcome limitations of human attention by combining the strengths of humans and AI to synergistically work together to accomplish shared objectives. For human-AI teaming to be truly successful in everyday life, more knowledge is needed about how humans shift attention between competing sources of information. For example, when driving, irrelevant distractions such as flashing billboards pull attention away from the road. To safely continue, the driver needs to disengage their attention from the distraction and direct it back to the road. The current work evaluates competing theories of how humans successfully disengage attention from visual distractions. One aim is to guide development of new technologies, such as computer vision and augmented reality, that aim to overcome limitations of human attention to improve performance in high-stakes situations (e.g., detecting potential threats within TSA scans or satellite images). Two competing explanations of how distractions compete for attention are tested using behavioral and neural (EEG) measures. The first explanation posits that visual distractions impair behavior by causing multiple shifts of attention, whereas the second explanation is that distractions impair behavior by overloading working memory. These ideas generate unique predictions for how distractions will impact behavioral and neural markers of attention and working memory. To test these ideas, the investigator combines a novel, large-scale EEG dataset (i.e., thousands of measurements per individual) with targeted manipulations of visual displays. The work informs understanding of the functioning of human attention, providing key information for building adaptive human-AI teaming applications that function safely and effectively. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.