Joanne Truong is a 4th year PhD student at Georgia Tech, co-advised by Prof. Dhruv Batra and Prof. Sonia Chernova. Her research lies at the intersection of machine learning and robotics. Her work focuses on pre-training AI agents for complex tasks in realistic simulators before transferring the skills learned to real robotic platforms. She is a recipient of the NSF Graduate Research Fellowship, Apple Scholars in AI/ML PhD Fellowship, Adobe Research Fellowship, and Google Women Techmakers Scholarship. She has interned at Meta AI, NVIDIA Research, and Google Robotics. Webpage: www.joannetruong.com
The goal of AI is to “construct useful intelligent systems”, such as mobile robots to assist in our day-to-day lives (e.g., robots delivering packages from one building to another). However, training robots in the real-world can be slow, dangerous, expensive, and difficult to reproduce. Thus, one paradigm in robot learning is to leverage simulation for training robots (where gathering experience is scalable, safe, cheap, and reproducible) before being deployed in the real world. My research leverages imperfect simulators for training robots for the real-world and has led to the development of: (1) a formal approach for simulator design by identifying how the simulator is not accurately reflecting reality; (2) sample-efficient methods for reducing the sim2real gap that improve robot learning and generalization; (3) an end-to-end learned approach that enables navigation in out-of-distribution environments zero-shot.