SCS Faculty Candidate
- Remote Access Enabled
- Virtual Presentation
- IGOR GILITSCHENSKI
- Senior Postdoctoral Associate
- Computer Science and Artificial Intelligence Lab (CSAIL)
- Massachusetts Institute of Technology
Robust Perception for Interactive Autonomy
Today we have the promise of a future with machines helping people with physical and cognitive work. This requires advancements in robust perception, control, planning, and learning. However, robust machine operation remains a challenging goal. Robustness in the “wild” cannot be achieved by merely optimizing algorithms that have been shown to work well under controlled lab conditions. In this talk, I will focus on robust perception for autonomous systems that operate in complex environments. First, I will address perception challenges for navigation in changing and challenging environmental conditions. Second, I will discuss the role of perception in interactive autonomy. Particularly, I will focus on the use of simulators for edge-case generation, learning, and real-world transfer of deep driving policies. Finally, I will discuss principled approaches for geometric reasoning under high uncertainty.
Igor Gilitschenski is a Senior Postdoctoral Associate at Massachusetts Institute of Technology (MIT). He works with Daniela Rus and Sertac Karaman in the Computer Science and Artificial Intelligence Lab (CSAIL), the Distributed Robotics Lab (DRL), and the Toyota-CSAIL Joint Research Center. His current research focuses on perception and learning for autonomous vehicles. Igor is the technical lead of DRL’s autonomous driving research team and serves as an advisor to MIT’s driverless racing team. Before joining MIT, he worked with Roland Siegwart at the Autonomous Systems Lab of ETH Zurich. Igor obtained a PhD in computer science at Karlsruhe Institute of Technology supervised by Uwe Hanebeck and Simon Julier. His research has received best paper awards at the International Conference on Information Fusion and the American Control Conference.
Faculty Host: Chris Atkeson
Zoom Partcipation Enabled. See announcement for registration details.