Bardew,_Patrick2aI am a PhD-Student in the Dyson Robotics Lab at the Imperial College London. My research is centered around Simultaneous Tracking and Mapping (SLAM) and Scene Understanding for autonomous robotic navigation. I am particular interested in ‘silicon retinas’ or event cameras, which are bio-inspired vision sensor. These sensors mimic retinas to measure per-pixel intensity changes rather outputting an actual intensity image. This paradigm shift away from traditional frame cameras offers significant potential advantages: namely avoiding high data rates, dynamic range limitations and motion blur.

I received a Diploma in Computer Science (M.Sc.-equivalent in Computer Science) with honors from the University of Koblenz in 2011 and worked as a Research-Developer at metaio GmbH in Munich, Germany (now Apple Inc.)  to develop Augmented Reality applications for mobile devices, before starting my PhD in 2014.

Current Main Research Interests

As detailed in my projects page, my main areas of interest are the following

  • Precise optical flow estimation from an event camera
  • Efficient processing of sparse information
  • 3-D reconstruction from camera data and scene understanding
Delicious Twitter Digg this StumbleUpon Facebook

Projects

  • Precise motion estimation and reconstruction using an event camera

    We are investigating methods to estimate motion with an Event Camera. The Event Camera is a recently developed camera sensor, which introduces a new camera paradigm where each pixel is an independent and highly sensitive brightness change detector with great possible advantages in low power, high dynamic range and tracking of rapid motion. Due to its sparse nature, most vision algorithms fail when they are applied to the Event Camera. This research focuses on novel algorithms, without assumptions on the camera motion or the environment, to enable generic motion estimation.

    [+] more