Task 010/011: Fast Visual Servoing with an Event Camera

Presentation Abstract: The field of computer vision explores the capabilities of traditional cameras, but also reveals some of their downfalls. Capturing high speed motion or operating robustly in environments with low light remains a challenge for a traditional camera. Event cameras are biologically inspired vision sensors that asynchronously collect data providing advantages in these challenging situations. We believe these neuromorphic devices can contribute strongly to the problem of event based robotic control. In this work, we explore high speed line following through visual servoing with an event camera. Our method extracts a steering angle as a function of the event stream.  We test our method using a non-holonomic robotic car platform outfitted with an Inivation Davis346 event camera.

Presenter Bio: Kendall Queen is an Electrical & Systems Engineering PhD student at the University of Pennsylvania under the advisement of Kostas Daniilidis, PhD. He received his B.S. in Computer Engineering from the University of Maryland, Baltimore County (UMBC) as a Meyerhoff Scholar and has earned his M.S. with a focus in Robotics from Penn. Kendall’s interests lie at the intersection of robotics and computer vision. Currently Kendall is investigating robotic system applications of event cameras, with plans of publishing his work at international conferences for computer vision and robotic systems.