A technique for estimating three-dimensional volume-of-interest using eye gaze.

摘要

Assistive robotics promises to be of use to those who have limited mobility or dexterity. Moreover, those who have limited movement of limbs can benefit greatly from such assistive devices. However, to use such devices, one would need to give commands to an assistive agent, often in the form of speech, gesture, or text. The need for a more convenient method of Human-Robot Interaction (HRI) is prevalent, especially for impaired users because of severe mobility constraints.;For a socially responsive assistive device to be an effective aid, the device generally should understand the intention of the user. Also, to perform a task based on gesture, the assistive device requires the user's area of attention in three-dimensional (3D) space. Gaze tracking can be used as a method to determine a specific volume of interest (VOI). However, heretofore gaze tracking has been under-utilized as a means of interaction and control in 3D space.;The main objective of this research is to determine a practical VOI in which an individual's eyes are focused by combining existing methods. Achieving this objective sets a foundation for further use of vergence data as a useful discriminant to generate a proper directive technique for assistive robotics.;This research investigates the accuracy of the Vector Intersection (VI) model when applied to a usable workspace. A neural network is also applied to gaze data for use in tandem with the VI model to create a Combined Model. The output of the Combined Model is a VOI that can be used to aid in a number of applications including robot path planning, entertainment, ubiquitous computing, and others. An alternative Search Region method is investigated as well.