Event Location: GHC 2109
Bio: Pratik Agarwal received his B.E. degree in Computer Science and
Engineering from Manipal Institute of Technology, India in 2010, and
M.S.E. in Computer Science and Engineering from University of Michigan,
Ann Arbor in 2012, where he also worked as a research assistant at the
APRIL robotics laboratory. Currently, he is a Doctoral Research
Assistant at the Autonomous Intelligent Systems Lab, University of
Freiburg, Germany. His research interests include robust robot mapping
and localization, sparse linear algebra and multi-object multi-sensor
tracking.

Abstract: Accurate global localization is one of the central challenges in mobile
robotics. Most of the existing methods use GPS or model it as a pose
estimation problem in a previously computed map. In this talk, I will
present a novel approach to metric localization that neither requires
the construction of a consistent map, nor a previous visit of the
environment. The central idea is to leverage geotagged imagery on Google
Street View as an accurate source of global positioning. Only a
monocular camera with odometric estimates are used to make this approach
as general as possible. I will present the accuracy of the method when
using a robotic platform in a parking lot with visual fiducials as
ground truth, and results of the approach applied to urban scenarios
with Google’s Tango mobile device.