Li Ling

I am a PhD student at the Division of Robotics, Perception and Learning supervised by John Folkesson, co-supervised by Nils Bore.

The ocean covers more than 70% of the Earth's surface, yet we know less about the deep ocean floors than the surface of the moon. Understanding the detailed topology of the ocean floor, also called bathymetry, will provide us with more insights on the geological and oceanographic processes of the Earth. Such knowledge in turn informs us of the ocean's biological and climate importance, allowing us to make more informed decisions.

Given the vast surface area of the ocean floor and its relative depth, autonomous underwater vehicles (AUVs) with sonars is the most realistic sensing platform. As part of the Swedish Maritime Robotics Centre (SMaRC), my research focuses on utilizing deep learning techniques for the perception and navigation of these AUVs.  More specifically, I am interested in improving AUV navigation and seafloor mapping by adapting computer vision techniques for camera images to various sonar sensors.  Some of my previous research topics include point cloud registration for multibeam sonar and automatic multibeam sonar noise removal. More recently, I am interested in reconstructing higher resolution bathymetry by applying neural rendering techniques to fuse various sonar measurements. Besides my research, I also contribute to building the SMaRC project, an open source software stack for AUVs.

I have a software engineering background. Prior to my PhD studies, I obtained a computer science and engineering degree from KTH, with a master’s degree in machine learning.