Wearable eye gaze tracking

• Grant number: 14/12048-0
• Support Opportunities: Scholarships abroad - Research Internship - Doctorate (Direct)
• Effective date (Start): September 02, 2014
• Effective date (End): September 01, 2015
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: Carlos Hitoshi Morimoto
• Grantee: Andrew Toshiaki Nakayama Kurauchi
• Supervisor: Margrit Betke
• Host Institution: Instituto de Matemática e Estatística (IME). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Research place: Boston University (BU), United States
• Associated to the scholarship: 13/06791-0 - Mobile eye tracker, BP.DD

Abstract

An eye gaze tracker (EGT) is a device that estimates a person's gaze location. Currently, the most common setup uses cameras placed on a computer monitor. Computer vision techniques are used to extract and track eye features and to estimate the gaze location on the computer screen after a calibration procedure. Head mounted eye gaze trackers (HM-EGT) have recently become very small and lightweight. They are typically composed of at least two cameras, one eye-camera, to track the movements of one eye and one scene-camera, that captures the scene. The calibration function maps eye features extracted from the eye-camera to a coordinate in the scene-camera. Considering that the performance of HM-EGTs will continue to improve as they become more affordable, it is easy to envision a near future scenario with people using wearable EGTs. Gaze information could then be used in many ways, for example, for gaze interaction with ahead-up display and for context awareness of the personal wearable computing network as an indicator of the level of attention and degree of interest on each object in the scene. The purpose of this research project is twofold, to improve the performance of current mobile eye tracking technology to be robust in everyday activities and to investigate new forms to apply eye gaze information once it becomes ubiquitous. This project is part of the Honors Doctorate fellowship FAPESP project no. 2013/06791-0. It will take place at the Computer Science Department of the Boston University under the supervision of Professor Margrit Betke. (AU)