About
Hi, I’m Junying Wang (王君瑛). I am currently a Ph.D. candidate at
CGIT
and the
Physical Superintelligence (PSI) Lab
in the
Department of Computer Science
at the
University of Southern California
,
advised by
Prof. Ulrich Neumann
and
Prof. Yue Wang
.
Previously, I received my bachelor’s degree from
Ocean University of China
.
During my time at USC, I also worked with the
Vision and Graphics Lab
at USC ICT, advised by
Prof. Hao Li
.
If you’d like to know more, please see my
CV
.
I am also happy to discuss research ideas and potential collaborations — please feel free to reach out!
Research
Humans experience the world in three dimensions. We perceive our surroundings through vision, sound, language, and touch, and we act on the world with our bodies — intelligence is grounded in this embodied, multimodal experience, not only in what we read, hear, and write. Building machines with this kind of intelligence requires rich digital representations of how people look, move, and interact in 3D environments.
My research lies at the intersection of digital humans and humanoid intelligence, spanning computer graphics and computer vision. I build technologies that understand, reconstruct, and generate realistic humans in digital form — capturing how people look, move, and interact in dynamic environments — and turn them into the data, assets, and simulation environments needed to train embodied AI.
Key problems I work on:
- Creating faithful, controllable, and editable digital humans from images, videos, or text
- Building high-quality 3D assets for virtual applications and humanoid robotics
- Developing efficient, scalable real-to-sim environments and data generators for training embodied AI
Core technical areas:
- 3D/4D human reconstruction and generation
- Neural rendering and relighting
- Generative AI for image and video synthesis
- Vision-language-action (VLA) training and post-training
Through this work, I aim to develop digital humans and intelligent agents that can better understand, interact with, and assist people in both virtual and real-world settings. :)
Education
Aug. 2020 – Present, USC Viterbi | Department of Computer Science, University of Southern California
Ph.D. Candidate
Jan. 2018 – Aug. 2020, USC Viterbi | Department of Computer Science, University of Southern California
Master of Engineering (Multimedia and Creative Technologies)
Aug. 2013 – Aug. 2017, Department of Computer Science and Engineering, Ocean University of China
Bachelor of Engineering
Publications
Humanoid Everywhere: Large-Scale Real-to-Sim Environment and Data Generator for Humanoid Loco-Manipulation
Robotics: Science and Systems Workshops (RSSW) 2026, Data-Centric Robotics
[Paper],
[Project page]
DGH: Dynamic Gaussian Hair
Conference on Neural Information Processing Systems (NeurIPS) 2025
[Paper],
[Project page]
Comprehensive Relighting: Generalizable and Consistent Monocular Human Relighting and Harmonization
Junying Wang,
Jingyuan Liu,
Xin Sun,
Krishna Kumar Singh,
Zhixin Shu,
He Zhang,
Jimei Yang,
Nanxuan Zhao,
Tuanfeng Y. Wang,
Simon S. Chen,
Ulrich Neumann,
Jae Shin Yoon
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2025
[Paper],
[Project page]
Boosting Generalizability towards Zero-Shot Cross-Dataset Single-Image Indoor Depth by Meta-Initialization
Cho-Ying Wu,
Yiqi Zhong,
Junying Wang,
Ulrich Neumann
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2024
[Paper],
[Code]
Complete 3D Human Reconstruction from a Single Incomplete Image
Junying Wang,
Jae Shin Yoon,
Tuanfeng Y. Wang,
Krishna Kumar Singh,
Ulrich Neumann
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023
[Paper],
[Project page]
Meta-Optimization for Higher Model Generalizability in Single-Image Depth Prediction
Cho-Ying Wu,
Yiqi Zhong,
Junying Wang,
Ulrich Neumann
IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) 2023
[Paper]
Patents
-
Complete 3D Object Reconstruction from an Incomplete Image
U.S. Patent Application No. 18/242,380 | Filed September 5, 2023 | Published March 6, 2025 | Patent Link -
Systems and Methods for Generating a Relighted Image
U.S. Patent Application No. 18/775,063 | Filed July 17, 2024 | Published January 22, 2026 | Patent Link
Experience
Research Scientist Intern | Meta (May 2025 – Dec. 2025)
Mentor: Dr. Yuanlu Xu
Contributed to a NeurIPS 2026 submission (under review).
Research Scientist Intern | Meta (Jun. 2024 – Dec. 2024)
Mentor: Dr. Tony Tung
Work published at NeurIPS 2025 (DGH: Dynamic Gaussian Hair).
Research Scientist Intern | Adobe (Sep. 2023 – Jan. 2024)
Mentor: Dr. Jae Shin Yoon
Work published at CVPR 2025 (Comprehensive Relighting).
Research Scientist Intern | Adobe (May 2022 – Dec. 2022)
Mentor: Dr. Jae Shin Yoon
Work published at CVPR 2023 (Complete 3D Human Reconstruction).
Professional Services
Reviewer
- IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
- ACM International Conference on Multimedia (MM)
- AAAI Conference on Artificial Intelligence (AAAI)
- International Conference on 3D Vision (3DV)
- International Conference on Computer Vision (ICCV)
- Conference on Neural Information Processing Systems (NeurIPS)
- European Conference on Computer Vision (ECCV)
- IEEE International Conference on Robotics and Automation (ICRA)
- British Machine Vision Conference (BMVC)
Teaching
-
Teaching Assistant | University of Southern California (Jan. 2022 – Dec. 2025)
CSCI-576: Multimedia Systems Design
-
Teaching Assistant | University of Southern California (Jan. 2021 – May 2021)
CSCI-570: Analysis of Algorithms