Karen Leung

kymleung [at] uw [dot] edu

Research Scientist @ NVIDIA

Assistant Professor @ UW A&A

Affiliate Faculty @ AFSL

I am joining the University of Washington Aeronautics and Astronautics department as an Assistant Professor in Fall 2022

I will be the Juris Vagners and Linda Christianson Endowed Faculty Fellow in Aeronautics & Astronautics. The fellowship provides support to faculty involved in controls and controls-related research and teaching with the goal of maintaining and expanding the excellence of the program.

I received my PhD in Aeronautics and Astronautics from Stanford University. I was part of the Autonomous Systems Lab directed by Marco Pavone. Currently, I am a Research Scientist at NVIDIA Research working in the Autonomous Vehicle Research Group.

The goal of my research is to harness advances in learning-empowered robot autonomy and unite them with assurances provided by formal methods to develop powerful yet safe and trustworthy autonomous systems. My research draws upon techniques from control theory, robot motion planning, optimization, formal methods and verification, and machine learning.

recent news

Jun, 2022

I attended my graduation ceremony and was thrilled to receive the William F. Ballhaus Prize for best PhD Thesis. Thank you everyone that has supported me along the way and made my PhD journey a wonderful experience.

Jun, 2022

I gave an invited talk at the Learning and control for safety-critical systems workshop. Thank you Yagiz Savas for inviting me to speak. I also presented my paper at American Control Conference 2022 at Atlanta, Georgia. It was my first in-person conference since the start of the pandemic, my first time in Atlanta, and my first time co-chairing a session. It was a great conference with great people!

Jun, 2022

I was invited to give a (virtual) talk at the workshop From Benchmarking Behavior Prediction to Socially Compatible Behavior Generation in Autonomous Driving at the Intelligent Vehicles Symposium 2022. Thank you Sascha Hornauer for inviting me and organizing the workshop! My collaborator Sever Topan at NVIDIA also presented our joint work titled Interaction-Dynamics-Aware Perception Zones for Obstacle Detection Safety Evaluation that was accepted to the conference. Arxiv link to be posted soon…

Mar, 2022

I gave a talk at the Princeton Robotics Seminar about Autonomous Vehicle Safety. Thank you Ani Majumdar for the invite—there were lots of great questions!! My next upcoming talk will at ACC where I will be presenting my paper and also an invited talk at the Learning and Control for Safety-Critical Systems workshop.

Jan, 2022

My paper Semi-Supervised Trajectory-Feedback Controller Synthesis for Signal Temporal Logic Specifications got accepted into American Control Conference 2022!

selected publications

  1. Multimodal Deep Generative Models for Trajectory Prediction: A Conditional Variational Autoencoder Approach
    IEEE Robotics and Automation Letters, 2021
  2. On Infusing Reachability-Based Safety Assurance within Planning Frameworks for Human-Robot Vehicle Interactions
    Leung, K., Schmerling, E., Zhang, M., Chen, M., Talbot, J., Gerdes, J. C., and Pavone, M.
    Int. Journal of Robotics Research, 2020
  3. Backpropagation through signal temporal logic specifications: Infusing logical structure into gradient-based methods
    Leung, K., Aréchiga, N., and Pavone, M.
    Int. Journal of Robotics Research, 2022 (In press)
  4. Towards the Unification and Data-Driven Synthesis of Autonomous Vehicle Safety Concepts