Xiangshan Tan
Papers
1
Total Citations
3
H-Index
1
About
Xiangshan Tan is a rising researcher at the forefront of embodied AI and human-robot interaction, with a primary focus on bridging natural language understanding and 3D scene perception. His work tackles the critical challenge of enabling robots to interpret and act upon free-form human references in complex three-dimensional environments. In his seminal paper, "Transcrib3D: 3D Referring Expression Resolution through Large Language Models" (2024), Tan pioneered a novel framework that leverages large language models to resolve ambiguous spatial descriptions—such as "the mug to the left of the laptop"—by grounding them in precise 3D coordinates. This contribution addresses a fundamental bottleneck in human-robot collaboration, where machines must parse both linguistic nuance and geometric structure simultaneously. Though early in his career, Tan's work has already garnered attention, with his most-cited paper accumulating 3 citations within its first year, signaling growing impact in the field. His research promises to make robots more intuitive partners in homes, warehouses, and factories, and positions him as a promising voice in the next generation of AI-driven robotics researchers.
Research Focus
Key Achievements
Top Papers
- 1