Wednesday, May 29, 2013
Robotics Institute Faculty Talk - Bernardo Pires
Visible-Light Gaze Tracking for Wearable Devices
Bernardo Pires, postdoctoral research fellow of the Robotics Institute at Carnegie Mellon University
May 29, 2013, 10:00 a.m.
Carnegie Mellon University, Newell Simon Hall, Room 3305
As technology advances, always-on assistive and enhancing wearable devices are close to becoming a reality. Gaze tracking can play an important part in such devices by giving insight into the users perception and intent. However, because they are intended for use during extended periods of time, these devices present a challenge for traditional infrared (IR) based gaze tracking. Non-IR gaze trackers are desirable for safety reasons, to minimize interference from IR light sources such as the sun, and to reduce device size and power consumption.
In this talk, I will discuss the challenges of building the First Person Vision wearable device at CMU and, in particular, of incorporating real-time non-IR gaze tracking into the device. I will present a novel method that locates the iris contour by fitting a circle on the estimated eye surface, and discuss some applications of the system, including gaze-assisted face detection and recognition, and real-time detection of the gaze location on a known object. If time permits, I will also discuss an application were the video recorded from the device is used for localization based on matching to an overhead satellite image.
Bernardo Pires is a Postdoctoral Fellow in the Robotics Institute at Carnegie Mellon University. His research interests lie in the areas of computer vision, image and signal processing, and wearable devices. Dr. Pires received his Licensiatura (B.Sc.) in Electrical and Computer Engineering from the Instituto Superior Tecnico (IST), Technical University of Lisbon in 2004 and Ph.D. from the Department of Electrical and Computer Engineering at Carnegie Mellon Univeristy in 2011. His research at IST won the award for the Best Graduation Thesis in the fields of electrical engineering and computer science.