Shengchao Yan
Papers
2
Total Citations
7
H-Index
2
About
Shengchao Yan is a robotics researcher whose work lies at the intersection of geometric deep learning and long-term autonomous navigation. His research focuses on two critical challenges: enabling robots to learn more efficiently by exploiting inherent symmetries in their physical structure, and building robust systems that can understand and adapt to dynamic environments over extended periods. In his highly cited 2024 paper, Yan introduced a novel approach to learning continuous control policies by leveraging geometric regularity from robot intrinsic symmetry, demonstrating how data symmetries can be exploited to overcome the curse of dimensionality in high-dimensional control tasks—a fundamental problem in modern robotics. His 2025 work, "BYE: Build Your Encoder With One Sequence of Exploration Data for Long-Term Dynamic Scene Understanding," tackles the persistent challenge of dynamic scene understanding, proposing a method that moves beyond traditional short-term masking approaches to enable robots to adapt to long-term environmental changes. With his papers already garnering attention in the robotics community, Yan’s contributions are shaping how robots learn from their physical structure and perceive changing environments, paving the way for more efficient and adaptable autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2