Papers
19
Total Citations
800
H-Index
11
About
Annie Xie is a robotics and machine learning researcher whose work sits at the intersection of deep reinforcement learning, imitation learning, and vision-based robotic control. She is best known for her contributions to model-based deep RL, most notably through "Visual Foresight" (2018, 264 citations), which demonstrated that robots could learn complex manipulation skills directly from raw visual inputs without task-specific reward engineering. Her early work also pioneered one-shot imitation learning from human video demonstrations using domain-adaptive meta-learning (2018, 113 citations), pushing robots closer to human-like observational learning. Xie has been an active contributor to large-scale, community-driven efforts to advance generalizable robotics, including the landmark Open X-Embodiment project (2024, 119 citations) and the DROID manipulation dataset (2024, 108 citations), both of which aim to build foundation models for robotics through diverse, large-scale data. Her research has also explored tool use with novel objects, multi-agent interaction, and systematically diagnosing the generalization gap in imitation learning. With over 700 cumulative citations, her body of work has meaningfully shaped how the field approaches scalable, generalizable robot learning from sensory data.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3One-Shot Imitation from Observing Humans via Domain-Adaptive Meta-Learning113 citations · 2018
- 4DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset108 citations · 2024
- 5
- 6
- 7Few-Shot Goal Inference for Visuomotor Learning and Planning16 citations · 2018
- 8
- 9Learning Latent Representations to Influence Multi-Agent Interaction13 citations · 2020
- 10Robotics: Science and Systems XV12 citations · 2019