Multi-Sensor SLAM

• Chris Choi (SRL Imperial College London)

Here, we fuse inertial measurements with visual measurements: due to the complementary characteristics of these sensing modalities, they have become a popular choice for accurate SLAM in mobile robotics. While historically the problem has been addressed with filtering, advancements in visual estimation suggest that non-linear optimisation offers superior accuracy, while still tractable in complexity thanks to the sparsity of the underlying problem. Taking inspiration from these findings, we formulate a probabilistic cost function that combines reprojection error of landmarks and inertial terms. We ensure real-time operation by limiting the optimisation to a bounded window of keyframes by applying various marginalisation strategies. Keyframes may be spaced in time by arbitrary intervals, while old measurements are still kept as linearised error terms.

Former collaborators:

• Simon Lynen (Previously ETH Zurich, now Google)
• Dr Mike Bosse (Previously ETH Zurich, Zoox)
• Dr Vincent Rabaud (Previously Willow Garage, now OpenCV Foundation)
• Dr Kurt Konolige (Previously Willow Garage, Google)
• Andreas Forster (Previously ETH Zurich, now Facebook)
• Dr Margarita Chli (ETH Zurich)
• Prof. Roland Siegwart (ETH Zurich)
• Dr Paul Furgale (Previously ETH Zurich, now Facebook)

Event cameras are novel camera systems that sense intensity change independently per pixel and report these events of change — brighter or darker by a specific amount — with a very accurate timestamp. As such, they are inspired from biology (retina) and offer the potential to overcome difficulties with motion blur or dynamic range that standard frame-based cameras face.

We have been looking at two different challenges: first, we tried to simply reconstruct both video and optical flow from the events: the approach should be able to deal with any scene content. Second, we tackled reconstruction of semi-dense depth and intensity keyframes along with general camera motion, where the scene is assumed to be static — effectively SLAM with an event camera.

Former collaborators:

• Patrick Bardow (Previously Dyson Robotics Lab at Imperial College London, now Google)
• Prof. Andrew Davison (Imperial College London)
• Hanme Kim (previously Robot Vision Group at Imperial College London, now Toyota Research Institute)