Real-Time Vision-Based Robot Localisation, Mapping, Scene Understanding and Interaction
The recently launched Dyson Robotics Lab at Imperial College is committed to pursuing cutting-edge research in the field of mobile robotics. Main related challenges comprise real-time computer vision for localisation, mapping and scene understanding – primary components of future robots for the home and beyond. The lab, directed by Prof. Andrew Davison and Dr Stefan Leutenegger, employs a team of post-doctoral researchers and PhD students, which we seek to reinforce with two additional PhD students. The students will work closely with the directors and other colleagues to advance the state-of-the-art in the fascinating topic of vision-based mobile robotics.
The main academic foci will be robot state estimation and mapping that aims at high robustness, accuracy, as well as scalability in operation space and time. Further challenges include the fusion with complementary sensors (e.g. odometry or Inertial Measurement Units), as well as developing an understanding of the scene and its contained hierarchies of objects. Moreover, we are interested in deriving ways of interaction with the environment through robot motion and control of manipulators.
For more information see
http://www3.imperial.ac.uk/computing/phd/scholarships
