3D/4D Scenes
Recovering real environments from sparse observations, modeling how they evolve over time, and grounding interactive world models in geometry and appearance.
Reconstruction · Rendering · SLAM · World Models
University of Science and Technology of China
Juyong Zhang（张举勇）
Professor · Ph.D. Advisor · NSFC Excellent Young Scholar
Vice Dean for Research, Institute of Advanced Technology, USTC
Executive Deputy Director, Anhui Provincial Key Laboratory of Graphics Computing and Perceptual Interaction
中国科大先进技术研究院副院长（主管科研）
安徽省图形计算与感知交互重点实验室常务副主任
Graphics & Geometric Computing Laboratory
School of Mathematical Sciences, USTC · Room 626
图形与几何计算实验室
中国科学技术大学数学科学学院 · 626 室
Research overview
From sensing the physical world to building intelligent, interactive digital worlds.
At a high level, our work asks one question: how can machines faithfully understand, reconstruct and generate people and environments across space and time? We connect computer vision, computer graphics, generative AI and numerical optimization to transform images, videos, audio and geometric observations into structured, controllable 3D/4D representations.
Our vision
We envision world models that are geometrically grounded, physically plausible and human-aware—systems that can perceive the real world, imagine how it may evolve, and create digital experiences in which people can naturally interact, communicate and explore.
研究简介
从感知真实世界，到构建智能、可交互的数字世界
从更高层面看，我们关注一个核心问题：如何让机器跨越空间与时间，真实地理解、重建并生成“人”与“环境”？我们融合计算机视觉、计算机图形学、生成式人工智能与数值优化，将图像、视频、语音和几何观测转化为结构化、可控制的三维/四维表达。
我们的愿景
构建具有几何基础、符合物理规律并能理解人的世界模型：它既能感知现实、推演世界如何变化，也能创造让人自然交流、互动和探索的数字体验。
Research Areas
Explore our research →Multimodal Affective Digital Humans
Combining appearance, motion, speech, expression and affect to build realistic digital humans that can communicate, express emotion and interact naturally.
Avatar · Expression · Speech · InteractionGeometry & Optimization
Developing efficient, robust and scalable methods for geometry processing, registration and sparse optimization to support reconstruction, generation, simulation and design.
Geometry · Registration · Numerical MethodsBiography
- Professor, USTCSchool of Mathematical Sciences
- Associate Professor, USTCSchool of Mathematical Sciences
- Postdoctoral ResearcherEPFL · Nanyang Technological University
- Ph.D., Nanyang Technological UniversityComputer Science and Engineering
- B.Eng., USTCComputer Science and Technology
- 中国科学技术大学教授数学科学学院
- 中国科学技术大学副教授数学科学学院
- 博士后研究瑞士 EPFL · 南洋理工大学
- 南洋理工大学博士计算机科学与工程
- 中国科学技术大学学士计算机科学与技术
Academic Service
IEEE Transactions on Mobile Computing (2025–present)
IEEE Computer Graphics and Applications (2023–present)
IEEE Transactions on Multimedia (2021–2025)
The Visual Computer (2018–2022)
CAD/Graphics 2021
National Conference on Geometric Design and Computing 2021
ChinaGraph 2024/2026
SGP 2015 · VCIP 2016 · CAD/Graphics 2017/2019 · GMP 2018–2021 · SPM 2019–2021 · CVM 2020–2022 · CVPR 2026/2027
IEEE Transactions on Mobile Computing（2025–至今）
IEEE Computer Graphics and Applications（2023–至今）
IEEE Transactions on Multimedia（2021–2025）
The Visual Computer（2018–2022）
CAD/Graphics 2021
全国几何设计与计算学术会议 2021
ChinaGraph 2024/2026
SGP 2015 · VCIP 2016 · CAD/Graphics 2017/2019 · GMP 2018–2021 · SPM 2019–2021 · CVM 2020–2022 · CVPR 2026/2027
Join Us
We welcome postdocs, Ph.D., master's and undergraduate students who are excited by 3D vision, graphics, generative AI and computational methods—and who value curiosity, rigor and collaboration.
加入我们
课题组长期欢迎对三维视觉、图形学、生成式 AI 与计算方法感兴趣，并重视好奇心、严谨性与合作精神的博士后、博士生、硕士生和本科生。