[PAST] Autonomous Aerial Navigation using Monocular Visual-Inertial Fusion
[PAST] Autonomous Aerial Navigation using Monocular Visual-Inertial Fusion 
2016 / 10 / 10 AM 11:00 - 12:00
Location: 313 - 224
Speaker: Prof. Shaojie Shen
Autonomous Aerial Navigation using Monocular Visual-Inertial Fusion
I will present recent results in the HKUST UAV group towards robust autonomous navigation using only the sensory information from a monocular camera and an IMU. We argue that this is the minimum sensor suite that enables the full capability for autonomous navigation, including state estimation, dense mapping, trajectory planning, and feedback control. We will present a self-calibrating visual-inertial estimator that handles high-speed motion, and is stable from the ground level to high altitude. We pay special attention to the initialization problem and propose a numerically stable method for high-altitude estimator initialization and failure recovery. Building on top of the estimator, we propose an onboard method to generate large-scale dense maps that are sufficient for obstacle detection and avoidance. We close the perception-action loop by an optimization-based trajectory generation method that finds safe trajectories in an online fashion. Experimental results are demonstrated on our custom-built quadrotor testbed.
Prof. Shaojie Shen
Department of Electrical & Computer Engineering
Hong Kong University of Science & Technology
Kowloon, Hong Kong
Shaojie Shen received his B.Eng. degree in Electronic Engineering from the Hong Kong University of Science and Technology in 2009. He received his M.S. in Robotics and Ph.D. in Electrical and Systems Engineering in 2011 and 2014, respectively, all from the University of Pennsylvania. He joined the Department of Electronic and Computer Engineering at the Hong Kong University of Science and Technology in September 2014 as an Assistant Professor. His research interests are in the areas of robotics and unmanned aerial vehicles, with focus on state estimation, sensor fusion, computer vision, localization and mapping, and autonomous navigation in complex environments. He and his research team won the best paper finalist in ICRA2011, best paper award in SSRR2015, and first price in IARC2015.