About me

I will be joining Cornell CS as an Assistant Professor in July 2022 — if you are a student interested in working with me, feel free to reach out!

I am interested in efficient inference for robot decision making — how should robots efficiently interact with both the environment and humans to infer the optimal action? I work on theory and algorithms at the intersection of machine learning and motion planning. Much of my research has been deployed on real-world robot systems — full-scale helicopters, self-driving cars, and mobile manipulators.

Currently, I'm a researcher at Aurora, where we are solving self-driving at scale. Previously, I was a Postdoctoral fellow at the University of Washington, CSE where I worked at the Personal Robotics Lab with Sidd Srinivasa.

I have a PhD from The Robotics Institute at Carnegie Mellon University, where I was advised by Sebastian Scherer. My thesis showed how robots can learn from prior experience to speed up online planning.

I am a Siebel’s Scholar, class of 2013.


Recent Research

My research aims to build a unified algorithmic framework where a robot efficiently infers the optimal value function from a bounded set of interactions with both humans and the environment. It ties together insights from motion planning and imitation learning and applies them to robots deployed in the wild. For more information, please have a look at my research and publications page.


2021_moment_matching

Of Moments and Matching: Trade-offs and Treatments in Imitation Learning
Gokul Swamy, Sanjiban Choudhury, Zhiwei Steven Wu, and J Andrew Bagnell
arXiv preprint arXiv:2103.03236, 2021
paper+code+video


2021_blended_mpc

Blending MPC & Value Function Approximation for Efficient Reinforcement Learning
Mohak Bhardwaj, Sanjiban Choudhury, and Byron Boots
International Conference on Learning Representations (ICLR), 2021
paper / arXiv


2021_feedback_il

Feedback in Imitation Learning: The Three Regimes of Covariate Shift
Jonathan Spencer, Sanjiban Choudhury, Arun Venkatraman, Brian Ziebart, and J Andrew Bagnell
arXiv preprint arXiv:2102.02872 , 2021
arXiv / talk


Selected News

June '21 Excited to start at Cornell CS in July 2022! If you are a student interested in working with me, feel free to reach out!
Feb '21 Our paper on Game-theoretic Imitation Learning was accepted to ICML'21 -- check out paper+code+video.
Jan '21 Our paper on blending MPC and Value Function accepted at ICLR 2021.
Jul '20 Unveiled ALICE, work done at Aurora, at 2020 ICML Workshop on AI for Autonomous Driving. Watch our talk to learn more!
May '20 Our paper on Expert Intervention Learning accepted at RSS 2020. Our paper on Imitation Learning as f-Divergence Minimization accepted at WAFR 2020.
Oct '19 RCTA collaboration with ARL finishes with a bang! Check out Roman clearing debris in the wild and articles by Economist and Signal.
Sep '19 I joined Aurora to work on fundamental research problems in learning and planning for self-driving. Come join (or intern with) our fantastic team!
Aug '19Presenting our year-long effort on MuSHR, an open-source robotic racecar! Checkout Allen School News and Geekwire articles, or read the full paper.
Aug '19I'm giving an invited talk at ONR Science of Autonomy 2019.
May '19Congratulations to Aditya for Best Student Paper Award for our paper on Generalized Lazy Search at ICAPS 2019!
Apr '19I'll be teaching a course on Mobile Robotics this quarter at University of Washington featuring brand new lectures and robot cars.
Feb '18AACUS wins the Howard Hughes Award. Also nominated for Collier Trophy.
Feb '18I defended my PhD on Adaptive Motion Planning! My thesis and talk are both available.
May '16I spent a lovely summer at Microsoft Research where I worked with Debadeepta Dey, Ashish Kapoor and Gireeja Ranade on imitation learning for POMDPs. Our IJRR 2018 paper was nominated for the best paper of the year.
May '14I received the Best Paper Award for my paper on Planner Ensemble in AHS 2014