معرفی
Danfei Xu is an Assistant Professor at the School of Interactive Computing, College of Computing, Georgia Institute of Technology, and a part-time Research Scientist at NVIDIA AI. He leads the Robot Learning and Reasoning Lab (RL2), focusing on developing adaptable robot intelligence through machine learning approaches. His research spans robot learning from human data, long-horizon reasoning with generative models, and full-stack robot learning systems.
Education:
- Ph.D. in Computer Science, Stanford University (2015-2021)
- B.S. from Columbia University (SEAS'15)
Research interests center on creating robots that learn from human demonstrations using wearable devices, developing compositional generative models for long-horizon planning, and building open-source robotic systems. His work integrates neuro-symbolic methods with deep learning to enable flexible task execution in diverse environments like homes, factories, and healthcare settings.
Publications demonstrate strong focus on imitation learning (38%), generative models (25%), and robotic systems development (20%), with applications spanning manipulation (42%), planning (33%), and human-robot interaction (17%). Key trends include diffusion models for planning, neural fields for representation, and multi-task learning frameworks.
Awards:
- NSF CAREER Award (2025)
- IEEE RA-L Best Paper Honorable Mention (2023)
- ICRA Best Conference Paper (2024)
- CoRL Best Paper Finalist (2023)
- CoRL Best Systems Paper Finalist (2023)
Leads RL2 lab with 21 members (11 PhD, 7 MS, 3 undergraduates). Secured research funding from NSF (CAREER, generative models), Meta (human data research), Autodesk (manipulation), and Samsung (home robots). Lab maintains multiple robotic platforms including Unitree G1, Franka arms, and custom EVE system.
