Yinghao Shuai

Papers

Total Citations

H-Index

About

Yinghao Shuai is pushing the boundaries of robotic perception by bridging the gap between 2D vision and 3D physical understanding. His research centers on zero-shot learning for robotics, enabling machines to interpret and interact with the world without task-specific training. His major contributions include PUGS, a groundbreaking method that uses Gaussian splatting to predict physical properties like mass, friction, and hardness directly from 3D reconstructions—a critical step toward robots that can manipulate objects as intuitively as humans. He also developed RE0, a framework that achieves zero-shot 3D instance segmentation by leveraging vision foundation models, overcoming the scarcity of high-quality 3D training data. Though early in his career, his work has already garnered attention, with PUGS receiving 2 citations shortly after its 2025 release. Shuai’s research is notable for tackling the “physical understanding gap” in robotics, moving beyond simple object recognition to enable nuanced, real-world interaction. His innovative use of Gaussian splatting and foundation models positions him as a rising leader in embodied AI and 3D scene understanding.