Profile Photo

Postdoc @ Stanford University

Email: szwu AT

Elliott / Shangzhe Wu

I am a postdoc researcher at Stanford University, working with Jiajun Wu. My current research focuses on unsupervised 3D perception and inverse rendering. I'm particularly fascinated by symmetries and structures in the world, as well as how they connect to the way we make sense of our perceptual experiences. The fundamental motivation of my work is to uncover the underlying statistical and geometric structures of the physical world (e.g., distributions of 3D shape, motion, appearance) from raw, casual observations (e.g., Internet photos, videos, audio) in an unsupervised manner akin to human perception.

I obtained my PhD from Oxford VGG, advised by Andrea Vedaldi and Christian Rupprecht, and BSc in Computer Science from HKUST, working with Chi-Keung Tang and Yu-Wing Tai. I also spent a lovely summer interning at Google Research with Noah Snavely's team.

Students: if you're passionate about 3D animal modeling and have extensive research experience in 3D computer vision, feel free to reach out for research internships and collaboration!


Sep 2024: We are organizing a Workshop on Visual Concepts at ECCV 2024.

Jun 2024: We are organizing the 4th CV4Animals Workshop at CVPR 2024.

Nov 2023: Received the BMVA Sullivan Doctoral Thesis Prize!

Dec 2022: Defended my PhD thesis on Unsupervised Learning of 3D Objects in the Wild!

Oct 2022: Collected another two Outstanding Reviewer badges from ECCV and NeurIPS!

Oct 2022: We are organizing an ECCV Workshop on Neural Geometry and Rendering with the CO3D Challenge.

Oct 2021: We are organizing an ICCV Workshop on Unsupervised 3D Learning in the Wild.

Jun 2021: I have set up some public office hours (following Jia-Bin Huang).

May 2021: Happy to be recognized as an Outstanding Reviewer at CVPR 2021!

Jun 2020: Our work on Unsup3D received the Best Paper Award at CVPR 2020!

Recent Talks:

[2024/05] Physical Natural World from Images

[2023/12] Learning 3D Fauna and Flora in the Wild


( * and indicate equal contribution or alphabetical order )



( This may take a while to load. Try refreshing the page if it's not loading. )