Shuang Li

Papers

Total Citations

H-Index

About

Shuang Li is a researcher working at the intersection of 3D scene understanding, robot learning, and embodied AI. Their work focuses on equipping robots and autonomous systems with richer, more human-like representations of the physical world, enabling more capable and generalizable manipulation and navigation behaviors. Li's most recognized contribution, "3D Neural Scene Representations for Visuomotor Control" (2021, 26 citations), addresses a fundamental gap between human spatial intuition and robotic capability. By leveraging neural 3D scene representations, the work advances how robots perceive and interact with objects of varying physical properties — a meaningful step toward bridging human-level physical reasoning with machine intelligence. More recently, Li contributed to "ConceptFusion" (2023), an ambitious open-set multimodal 3D mapping framework that moves beyond the limitations of closed-set semantic systems, allowing robots to reason about an open-ended range of concepts when building environmental maps — a critical capability for real-world deployment. Together, these contributions reflect Li's broader mission to develop robots that understand and act within complex 3D environments with greater flexibility and semantic awareness. Their research is gaining traction in the robotics and computer vision communities, signaling a promising trajectory in embodied AI research.