New paper on solving high-dimensional reachability problems using deep learning. So far we have tested the method on computing up to 10D backward reachable sets/tubes and reach-avoid sets/tubes and have seen very promising results.
New paper on designing a robust control framework for human motion prediction (accepted to RA-L 2020).
I am honored to receive the Eli Jury Award from UC Berkeley, EECS for "outstanding achievement in the area of systems, communications, control, or signal processing."
Successfully defended my thesis at UC Berkeley! Thank you, Claire. Check out what I have been upto for the past 5 years [dissertation talk].
Check out my talk on scaling Hamilton-Jacobi reachability analysis for robotics at IPAM.
Excited to share our work on viusal navigation among humans. The work was done in collaboration with Google Brain research.
We are organizing the 2nd workshop on safe and robust autonomy at RSS 2020. Check out our call for papers (due on 5 June, 2020).
Check out our work on safe navigation among humans. We will present this work at ICRA 2020.
New paper on generating robust supervision for learning-based visual navigation using Hamilton-Jacobi reachability. We will present this work at L4DC 2020.
I am applying for faculty positions this Fall (research statement).
I will be at CDC 2019! Check out our work on:
- A task-specific model learning framework using Bayesian Optimization, Wednesday (Dec 11), 11:20-11:40am, Risso 8 (WeA23.5).
- A provably safe framework for perception-based navigation in unknown environments, Wednesday (Dec 11), 2-2:20pm, Risso 8 (WeB23.1).
- A warm-starting method for efficiently updating reachability-based safety guarantees, Thursday (Dec 12), 6:10-6:30pm, Méditerranée 5 (ThC03.6).
- Meet the Faculty Candidate poster session, Wednesday (Dec 11), 6:30-8:30pm, Rhodes Exhibition area.