I am a Stephen Fleming Early Career Professor in Computer Science at Georgia Tech. I am in the School of Interactive Computing affiliated with Robotics and Machine Learning programs. I also hold courtesy appointments at University of Toronto and Vector Institute. I have previously held research leadership positions at Nvidia and Apptronik.
My research vision is to build the Algorithmic Foundations for Generalizable Autonomy, that enables robots to acquire skills, at both cognitive & dexterous levels, and to seamlessly interact & collaborate with humans in novel environments. My group focuses on understanding structured inductive biases and causality for decision making. In particular we are looking at multi-modal object-centric and spatiotemporal event representations, self-supervised pre-training for reinforcement learning & control, principle of efficient dexterous skill learning.
Research Interests: Foundation Models, Reinforcement Learning, Robotics, 3D Vision.
Current Applications: Self-Driving Labs, Surgical Robotics, Personal Robotics.
Check out PAIR Group for info on reseearch projects and how to join.
News
Recent Talks
Priors as Abstractions for Autonomy
June 2024, CVPR CORR Workshop
Towards Generalizable Autonomy
Mar 2023, NURO Seminar Series
Building Blocks of Embodied AI
Oct 2022 Stanford Robotics Seminar
Experience
Georgia Institute of Technology
Assistant Professor
Apptronik
Chief Scientific Officer
Nvidia Research
Senior Staff Research Scientist
University of Toronto / Vector Institute
Assistant Professor
Stanford University
Postdoc in Computer Science
Education
UC Berkeley
Ph.D. in Operations Research
MS in Computer Science
Georgia Tech
MS in Industrial Engineering
University of Delhi
B.E. in Manufacturing Processes & Automation
Teaching
Deep Reinforcement Learning (Grad)
F24, F25
Algorithmic Intelligence in Robotics (UG)
F20, Sp22
Introduction to Reinforcement Learning (UG)
Sp21, F21
Reinforcement Learning (Grad)
Sp20