Chenrui Tie (铁宸睿)

I am a first-year Ph.D. student in the School of Computing, National University of Singapore, supervised by Prof. Lin Shao. Before this, I received my B.S. degree from School of EECS, Peking University in 2024. In my undergraduate study, I am privileged to work with Prof. Hao Dong and Prof. He Wang.

It's worth mentioning that before I study computer science, I majored in physics. My research focuses on constructing models by incorporating varying degrees of physics intuitions based on the specific characteristics of different robot manipulation tasks, so that the robot can more efficiently complete a variety of manipulation tasks in the physical world.

Email: chenrui.tie [at] u.nus.edu

Email  /  Github  /  Google Scholar

We propose a novel framework that enables VLM to understand human-designed manuals and acquire robotic skills for furniture assembly tasks.

We propose a new diffusion policy method to tackle tasks with certain symmetry, achieving better data efficiency and spatial generalization.

We propose EqvAfford framework, with novel designs to guarantee the SE(3) equivariance in point-level affordance learning for downstream robotic manipulation.

We Introduce a framework taking contact synthesis as a unified task representation that can generalizes over objects, robots, and manipulation tasks.

We developed Jade, a differentiable physics engine for articulated rigid bodies, using the continuous collision detection to prevent intersection between bodies with complex geometry shapes.

We tackle multi-part geometrically assembly task, leveraging SE(3) equivariance and invariance, which fits the natural characteristic of the task and narrows the solution space.

• Advisor: Prof. Lin Shao

• Advisor: Prof. Hao Dong, Prof. He Wang
• Lab: Hyperplane Lab and Epic Lab, Center on Frontiers of Computing Studies

• International Conference on Robotics and Automation (ICRA)
• Robotics and Automation Letters (RA-L)
• International Conference on Learning Representations (ICLR)