The challenge to effectively using visual information does not end with perception or the inverse problem often associated with David Marr. In many places, taking advantage of visual inputs requires cognition in order to compare encounters that are separated in time and space. Whereas visual perception has benefited from considerable analysis at the algorithmic level, visual cognition is rarely subjected to the same kind of scrutiny. In particular the limitations of human performance are rarely attributed to the inputs and algorithms that support visual cognition —and are instead attributed to hypothesized capacity and resource limits of various forms. This talk will describe recent research in areas including object tracking, spatial working memory, and visual working memory more broadly, illustrating the considerable extent to which imperfect inputs and algorithms inescapably hamstring human performance.
To view a recording of this talk click here (You will need a Rutgers NetID and password)