Xiaoshen Han
Papers
1
Total Citations
3
H-Index
1
About
Xiaoshen Han is a rising researcher at the forefront of embodied AI and robotic manipulation, with a focus on grounding high-level language understanding in low-level physical control. His most-cited work, "RoboGround: Robotic Manipulation with Grounded Vision-Language Priors," introduces a novel intermediate representation—grounding masks—that bridges the semantic gap between vision-language models and precise robotic actions. By demonstrating how these masks provide effective spatial guidance for policy learning, Han’s research directly tackles the challenge of generalization in manipulation tasks, enabling robots to better interpret and act upon open-ended instructions. Though early in his career, his contributions are already shaping how the field thinks about integrating pre-trained priors with real-world robotics. With 3 citations to date, his work signals a promising trajectory toward more robust, adaptable robotic systems. Han’s approach stands out for its elegant simplicity: rather than relying on complex end-to-end models, he leverages grounded, interpretable representations that improve both performance and transparency—a philosophy that may define the next generation of intelligent manipulation.
Research Focus
Key Achievements
Top Papers
- 1RoboGround: Robotic Manipulation with Grounded Vision-Language Priors3 citations · 2025