Papers
15
Total Citations
255
H-Index
9
About
Baichuan Huang is a roboticist whose research lies at the intersection of manipulation, planning, and learning, with a particular focus on enabling robots to operate intelligently in cluttered, unstructured environments. His most impactful work includes the Deep Interaction Prediction Network (DIPN), which allows robots to "imagine" the physical effects of pushing actions on unknown objects—a key capability for tasks like clutter removal. This work has garnered 59 citations. He has also pioneered the use of Visual Foresight Trees for non-prehensile object retrieval, and developed EARL, a reinforcement learning framework for dynamic grasping of moving objects. Across his publications, which total over 200 citations, Huang consistently advances the state of the art in tabletop rearrangement, Monte Carlo tree search for long-horizon planning, and human-robot interaction through natural language and mixed reality. His work on minimizing running buffers for rearrangement has practical implications for warehouse and household robotics. Huang’s contributions are notable for their blend of theoretical rigor—including complexity analysis and game-theoretic coordination—with real-world robotic systems, making his research both foundational and immediately applicable.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Fast High-Quality Tabletop Rearrangement in Bounded Workspace27 citations · 2022
- 5
- 6
- 7
- 8
- 9Stackelberg Strategic Guidance for Heterogeneous Robots Collaboration9 citations · 2022
- 10Learning Generalizable Pivoting Skills8 citations · 2023