CS Seminar – Li Sun (University of Oxford)

Deep Learning for the Robot to Think Fast

Deep learning is empowering the world's robots to understand the complex scene. An advanced robotic vision system enables the robot arm to manipulate objects with the capability of detecting and tracking, and even forecasting the outcome and imagining to use tools. An intelligent mobile robot should be able to efficiently and precisely localize itself in a large-scale environment, and in the meanwhile, understand the semantics of surroundings to enhance the awareness in navigation. In this talk, we will introduce some pioneering research on deep learning for the robot to think fast, e.g. thinking-fast robot localization and mapping, end-to-end multi-object tracking, imagining tool-use for robot reaching task. In the end, we will discuss the bottlenecks and the potentials of the state-of-the-art deep learning for robotic applications.

August 15, 2019 (Thursday)
Room 205, Building E, Karlovo nám. 13


Li Sun (Kevin)

Kevin is a Postdoctoral Research Assistant at the Oxford Robotics Institute, University of Oxford, UK. His research interest focuses on robot learning for robot perception and localization, task-oriented manipulation, and long-term autonomy. He is working on solving the core challenges in the emerging autonomous robotics to enable the robot to manipulate with complex industrial objects or localize and navigate in the dynamic, real-life environment e.g. warehouse, agriculture, urban driving, etc. Kevin received his Ph.D. from the University of Glasgow, UK in 2016. Before joining Oxford, he was the postdoctoral research fellow at Lincoln Center of Autonomous Systems, University of Lincoln (2017-2018) and Extreme Robotics Lab, University of Birmingham (2016-2017). Since 2012, he has contributed to a few EU/EPSRC funded projects e.g. CloPeMa, RoMaNs, ILIAD, and MobileRobotics. Kevin will join the University of Sheffield as a lecturer (Assistant Professor) in September 2019.