Xinyuan Yu
Papers
1
Total Citations
8
H-Index
1
About
Xinyuan Yu is a leading researcher in differentiable simulation and robotic manipulation, with a primary focus on cloth dynamics and contact-rich interactions. Their most notable contribution is the development of **DiffClothAI**, a groundbreaking differentiable cloth simulator that ensures intersection-free frictional contact while enabling differentiable two-way coupling with articulated rigid bodies. This work, published in 2023, addresses a critical challenge in robotic cloth manipulation—maintaining physically plausible, intersection-free simulations during gradient-based optimization. By providing a robust framework for differentiable cloth dynamics, Yu’s research enables robots to learn complex manipulation tasks, such as folding or draping fabrics, with unprecedented accuracy and efficiency. With 8 citations in just its first year, DiffClothAI has already gained recognition for its potential to bridge simulation and real-world robotic control. Yu’s work is particularly impactful for students and researchers in robotics, computer graphics, and machine learning, offering a powerful tool for advancing autonomous cloth handling—a long-standing challenge in household and industrial robotics. Their contributions exemplify how differentiable physics can unlock new capabilities in robotic learning and control.
Research Focus
Key Achievements
Top Papers
- 1