Date: Friday 20/1/2012
Time: 1.00 pm
Speaker: Prof. Paul Newman
Affiliation: Oxford University, UK
Large Scale Long Term Urban Navigation using Vision and Laser
By Prof. Paul Newman
Oxford University, UK
I will talk about large scale long term navigation and describe our work into vision and laser based localisation. I shall use intelligent transport as an application domain as it is here that modern robotics provides an opportunity to impact our every days lives. Surely we don't have to drive the car all the time? I will lay out and demonstrate a new formulation for "experienced based mapping" which allows us to incrementally learn a model whose complexity varies naturally in accordance with variation of scene appearance. It allows us to leverage the state of the art in pose estimation to build over many runs, a world model of sufficient richness to allow simple localisation despite a large variation in conditions. As our robot repeatedly traverses its workspace, it accumulates distinct visual experiences that in concert, implicitly represent the scene variation - each experience capturing a visual mode. This results in an ability to localise metrically with respect to previous trajectories. By adding topological connections derived from appearance based reasoning and matching online views to over head images we are able to demonstrate ever improving performance over the past 3 months of operation. This is a step towards life long learning.
Prof. Paul Newman obtained an M.Eng. in Engineering Science from Oxford University in 1995. He then undertook a Ph.D. in autonomous navigation at the Australian Center for Field Robotics, University of Sydney, Australia. In 1999 he returned to the United Kingdom to work in the commercial sub-sea navigation industry. In late 2000 he joined the Dept of Ocean Engineering at M.I.T. where as a post-doc and later a research scientist, he worked on algorithms and software for robust autonomous navigation for both land and sub-sea agents. In early 2003 he returned to Oxford as a Departmental Lecturer in Engineering Science before being appointed to a University Lectureship in Information Engineering and becoming a Fellow of New College in 2005.
Over the course of his career he has developed a particular expertise in the application of probabilistic methods to robotic navigation and mapping. His focus lies on pushing the boundaries of navigation techniques in terms of both endurance and scale. He is on the editorial board for the top two robotics science journals (IJRR, and JFR), was an IEEE Dis- tinguished Lecturer for Europe in 2009 and 2010 and is the IEEE media spokesman for robotics and automation. In 2012 he will be Program Chair for the premier robotics conference - Robotics Science and Systems.In 2010 he was awarded an EPSRC Leadership Fellowship.