Xiaobin Hu
Papers
1
Total Citations
18
H-Index
1
About
Xiaobin Hu is a leading researcher at the intersection of embodied AI, robotics, and multi-modal learning, with a core focus on bridging the gap between large language models and physical world interaction. His most notable contribution is the development of **ManipVQA**, a pioneering framework that injects robotic affordance and physically grounded information into Multi-modal Large Language Models (MLLMs). This work directly addresses a critical limitation in robotics: while MLLMs excel at understanding language, they often lack the robotics-specific knowledge needed for real-world manipulation tasks. By enabling models to reason about object properties, spatial relationships, and possible actions, Hu’s research empowers robots to move beyond simple instruction following toward context-aware, physically feasible execution. His work has quickly gained traction, with ManipVQA accumulating 18 citations within its first year—a strong indicator of its timely impact. Hu’s research is essential reading for anyone interested in grounding AI in the physical world, and his contributions are helping to define the next generation of intelligent, interactive robotic systems.
Research Focus
Key Achievements
Top Papers
- 1