Downloads & Free Reading Options - Results

Consider the problem of visually finding an object in a mostly unknown space witha mobile robot. It is clear that all possible views and images cannot be examinedin a practical system. Visual attention is a complex phenomenon; we view it asa mechanism that optimizes the search processes inherent in vision. Here, we describea particular example of a practical robotic vision system that employs someof these attentive processes. We cast this as an optimization problem, i.e., optimizingthe probability of finding the target given a fixed cost limit in terms of totalnumber of robotic actions required to find the visual target. Due to the inherentintractability of this problem, we present an approximate solution and investigateits performance and properties. We conclude that our approach is sufficient to solve this problem and has additional desirable empirical characteristics. ©2009 Microsoft Corporation. All rights reserved.