Ying He @NTU
Ying He

Associate Professor
College of Computing and Data Science
Nanyang Technological University

Office: S3-B1c-98
Tel: +65 6514 1008
Email: yhe `at' ntu.edu.sg


Research Overview

I am an Associate Professor at the College of Computing and Data Science (CCDS), Nanyang Technological University (NTU), Singapore. My research began with manifold splines during my PhD at Stony Brook University (supervised by Prof. Hong Qin), which laid the foundation for my work in geometry processing and geometric computing. It has since evolved along a continuous geometry-driven trajectory, from geometric modeling to digital geometry processing, and more recently to 3D vision and 3D deep learning. Across these directions, a unifying theme is the development of geometry-aware, representation-first methods for the reconstruction, representation, and understanding of 3D data, designed to operate robustly under real-world constraints such as noise, incompleteness, scale, and complex topology.

My research spans geometric modeling, digital geometry processing, 3D vision, and 3D deep learning, and is characterized by reformulating difficult geometric inference problems as reusable and scalable structures. Two long-running themes define this work most clearly. In discrete geodesics, my work on Saddle Vertex Graphs (SVG, SIGGRAPH Asia 2013) and later Discrete Geodesic Graphs (DGG, TOG 2020) helped change the way people think about shortest-path computation on surfaces. These works show that although geodesic computation is inherently a global problem, it can be handled efficiently by exploiting local geometric structures. This leads to representations that support fast, scalable, and repeated shortest-path queries on surfaces. In 3D reconstruction from raw or unoriented point clouds, my work on iterative Poisson surface reconstruction (iPSR, SIGGRAPH 2022) and subsequent extensions showed that orientation and reconstruction should be treated as coupled aspects of a single geometric inference problem, rather than as a brittle two-stage pipeline, enabling more robust and scalable reconstruction from real-world data.

In recent years, my work has focused on the following research directions:

Each webpage provides representative publications, with PDF or open-access links where available. Further details on research impact, teaching, and service can be found on the following pages:

Google Scholar | Scopus | Representative Works | Recognition: SMA Fellow (2024) | Teaching | Academic Roles and Service


Selected Publications (Since 2006)

Highlights: ACM TOG/SIGGRAPH/SIGGRAPH Asia (22), IEEE TVCG (35), IEEE TPAMI/IJCV (8), IEEE TCSVT (8), IEEE TIP (3), CAD (30), CAGD (5), CGF (5), CVMJ (5), TVCJ (4), AAAI (10), CVPR/ICCV/ECCV (18), NeurIPS/ICLR/ICML (8), ACM MM (4), ACM CHI (3), and ACM I3D (6).

Monograph

Journal papers

Referred conference papers


Awards


Professional Services