Danfei Xu
I am an Assistant Professor at the School of Interactive Computing at Georgia Tech and a Research Lead at NVIDIA GEAR. I work at the intersection of robotics and machine learning.
I received my Ph.D. in Computer Science from Stanford University, advised by Fei-Fei Li and Silvio Savarese, and my B.S. from Columbia University.
Research opportunities: Please fill out this questionnaire and email me once submitted.
Research mission
I direct the Robot Learning and Reasoning Lab (RL2). We build general-purpose, adaptable robot intelligence for homes, factories, healthcare, and search and rescue.
- Learning from human data. Enabling robots to learn from everyday human experience, including data captured by wearable devices such as smart glasses. I discuss this direction in Robot Learning from Human Experience and my CoRL 2024 keynote.
- Long-horizon reasoning. Combining generative and neuro-symbolic models to compose skills, predict outcomes, and solve complex tasks in collaboration with people. See my talk on long-horizon reasoning.
- Robot learning systems. Building open-source hardware and software that bring advances in learning-based robotics into the physical world.
Find our publications and latest news on the RL2 lab website .
Teaching
-
Georgia Tech
CS 7643 / 4644 -
Georgia Tech
CS 8803 DLMDeep Learning for Robotics
-
Stanford
CS 231n
Spring 2021Spring 2020