Cheng Yu | Ph.D. Student at Peking University (PKU)

Ph.D. student at Peking University (PKU)

I am Cheng Yu, a Ph.D. student in the School of Intelligence Science and Technology at Peking University (PKU), China. I am advised by Prof. Mengyu Chu in Visual Computing and Learning Lab. I received my bachelor’s degree from the School of Electronic Engineering and Computer Science at Peking University in July 2024.

Research

My research focuses on computer graphics and physics-based animation, especially robust and efficient methods for collision detection, contact, friction, and thin-shell simulation. I develop algorithms that balance stability, realism, and computational efficiency for interactive graphics, virtual reality, and digital content creation. See my publications for recent work.

selected publications

Floating-Point Robustness in Parametric Surface Continuous Collision Detection: From Algorithm to Benchmarking

Xuwen Chen, Junyu Wang, Cheng Yu, Xingyu Ni, Meng Zhang, and 3 more authors

ACM Transactions on Graphics(SIGGRAPH 2026) , Jul 2026

DOI Bib PDF Website

@article{Chen2026CCDdataset,
  title = {Floating-Point Robustness in Parametric Surface Continuous Collision Detection: From Algorithm to Benchmarking},
  author = {Chen, Xuwen and Wang, Junyu and Yu, Cheng and Ni, Xingyu and Zhang, Meng and Wang, Bin and Chu, Mengyu and Chen, Baoquan},
  journal = {ACM Transactions on Graphics},
  volume = {45},
  number = {4},
  articleno = {141},
  numpages = {14},
  year = {2026},
  month = jul,
  publisher = {Association for Computing Machinery},
  doi = {10.1145/3811310},
  url = {https://doi.org/10.1145/3811310},
}

JODA: Composable Joint Dynamics for Articulated Objects

Tianhong Gao, Cheng Yu, Yinghao Xu, and Mengyu Chu

In Advances in Neural Information Processing Systems (NeurIPS), Dec 2026

arXiv Bib PDF Website

@inproceedings{Gao2026JODA,
  title = {{JODA}: Composable Joint Dynamics for Articulated Objects},
  author = {Gao, Tianhong and Yu, Cheng and Xu, Yinghao and Chu, Mengyu},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2026},
  month = dec,
  url = {https://openreview.net/forum?id=y49DW1RjHQ},
}

Simulating Thin Shells by Bicubic Hermite Elements

Xingyu Ni, Xuwen Chen, Cheng Yu, Bin Wang, and Baoquan Chen

Computer-Aided Design, Dec 2024

Abs DOI

In this study, we present the bicubic Hermite element method (BHEM), a new computational framework devised for the elastodynamic simulation of thin-shell structures. The BHEM is constructed based on quadrilateral Hermite patches, which serve as a unified representation for shell geometry, simulation, collision avoidance, as well as rendering. Compared with the commonly utilized linear FEM, the BHEM offers higher-order solution spaces, enabling the capture of more intricate and smoother geometries while employing significantly fewer finite elements. In comparison to other high-order methods, the BHEM achieves conforming C1 continuity for Kirchhoff–Love (KL) shells with minimal complexity. Furthermore, by leveraging the subdivision and convex hull properties of Hermite patches, we develop an efficient algorithm for ray-patch intersections, facilitating collision handling in simulations and ray tracing in rendering. This eliminates the need for laborious remodeling of the pre-existing surface as the conventional approaches do. We substantiate our claims with comprehensive experiments, which demonstrate the high accuracy and versatility of the proposed method.
A Time-Dependent Inclusion-Based Method for Continuous Collision Detection between Parametric Surfaces

Xuwen Chen, Cheng Yu, Xingyu Ni, Mengyu Chu, Bin Wang, and 1 more author

ACM Trans. Graph., Nov 2024

Abs DOI

Continuous collision detection (CCD) between parametric surfaces is typically formulated as a five-dimensional constrained optimization problem. In the field of CAD and computer graphics, common approaches to solving this problem rely on linearization or sampling strategies. Alternatively, inclusion-based techniques detect collisions by employing 5D inclusion functions, which are typically designed to represent the swept volumes of parametric surfaces over a given time span, and narrowing down the earliest collision moment through subdivision in both spatial and temporal dimensions. However, when high detection accuracy is required, all these approaches significantly increases computational consumption due to the high-dimensional searching space. In this work, we develop a new time-dependent inclusion-based CCD framework that eliminates the need for temporal subdivision and can speedup conventional methods by a factor ranging from 36 to 138. To achieve this, we propose a novel time-dependent inclusion function that provides a continuous representation of a moving surface, along with a corresponding intersection detection algorithm that quickly identifies the time intervals when collisions are likely to occur. We validate our method across various primitive types, demonstrate its efficacy within the simulation pipeline and show that it significantly improves CCD efficiency while maintaining accuracy.