Lanke Frank Tarimo Fu

I'm a PhD student in the Dynamic Robot Systems Groups at the University of Oxford, supervised by Prof Maurice Fallon. My research is on multi-sensor fusion for robotics with an emphasis on camera-LiDAR calibration using gradient-based optimization.

Previously, I worked on multi-sensor tracking and prediction at Oxa

Email  /  Scholar  /  Twitter  /  Github

Research

I'm interested in deep learning, differentiable optimization, factor graphs, and projective geometry. Most of my research is about spatially aligning a camera image and a LiDAR pointcloud using either deep-learned, or hand-crafted features.

Batch differentiable pose alignment enables efficient self-supervised training of aligned LiDAR and Camera features. This approach is capable of zero-shot transferring to new sensors and environments, retrieving calibration parameters with centimeter accuracy.

Approximating the checkerboard pattern with a continuous function enables aligning the LiDAR point cloud to the camera image using gradient-based optimization.

This webpage is adapted from Jon Barron's site.