Search this site
Embedded Files
Zezhou Cheng

CV / Research Statement / Google Scholar / Twitter  

I am a tenure-track assistant professor in Computer Science at the University of Virginia, leading the Computer Vision Lab. Before joining UVA, I was a Postdoctoral Researcher at Caltech, advised by Georgia Gkioxari. I obtained my Ph.D. in Computer Science at UMass Amherst in 2023, co-advised by Subhransu Maji and Daniel Sheldon. 


My research interests include:  (1) Self-supervised Representation Learning; (2) 3D Vision, Video Understanding and Generation; (3)  Applications of Computer Vision in Other Domains (e.g., Ecology, Chemistry, etc.)


Prospective Students: I am actively looking for research interns, if you’re interestd,  please contact me (zc3bp@virginia.edu) with your CV or complete this form.  Please visit Computer Vision Lab @ UVa  for more information. 

News 

  • 02/26     One paper accepted to CVPR 2026: WildRayZer. Congrats Xuweiyi and Wentao! 

  • 02/26      Gave a talk at the Symposium on Computer and Autonomous Vision Systems  on Rethinking Self-Supervised Learning in Computer Vision: Challenges, Tasks, and Methods. Thanks to Lei, Cheng, and Sheng for the kind invitation! 

  • 01/26     One paper accepted to ICLR 2026: Point-MoE. Congrats Xuweiyi and Wentao! 

  • 01/26      Teaching Computer Vision this Spring.

  • 01/26      I will be co-organizing GPU-Accelerated Computing Research Group. Welcome to join us to explore the intersection of GPUs with AI, HPC, systems/cybersecurity, and sustainability.

  • 12/25       I'll be serving on Senior Program Committee at IJCAI

  • 11/25       Received an Adobe Research Gift. Thank you, Adobe! 

  • 10/25      Two papers accepted to 3DV 2026: OVMono3D and Point-MAE-Zero. Congrats Jin and Xuweiyi! 

  • 10/25      Received MathWorks Research Award. Thank you, MathWorks!

  • 09/25     Received NVIDIA Academic Grant Program Award. Thank you, NVIDIA!

  • 09/25     Gave a talk at NSF-Simons AI Institute for Cosmic Origins on Computer Vision for Scientific Discovery [Record]

  • 09/25      Two papers accepted to NeurIPS 2025: LabelAny3D and Frame-In-N-Out. Congrats Jin and Boyang!

  • 09/25      Teaching 3D Computer Vision this Fall.

  • 08/25     Received AMD’s High Performance Compute Fund. Thank you, AMD!

  • 06/25       I'll be serving on Senior Program Committee at AAAI for Social Impact Track.

  • 05/25      Our NSF NAIRR Pilot project has been awarded! 

  • 03/25      Gave a talk at MathWorks: Recognize Anything in 3D with Minimal Human Supervision        

  • 02/25      One papers accepted to CVPR 2025: Self-supervised Learning for Mid-Level Vision. Congrats Xuweiyi! 

  • 01/25      Teaching Computer Vision this Spring.

  • 12/24       Received an Adobe Research Gift. Thank you, Adobe! 

  • 12/24        Gave a talk at UVA AIML Seminar: Recognize Anything in 3D with Minimal Human Supervision

  • 09/24      Teaching 3D Computer Vision this Fall.

Publications  [Google Scholar]

WildRayZer: Self-supervised Large View Synthesis in Dynamic Environments 

Xuweiyi Chen, Wentao Zhou, Zezhou Cheng

Computer Vision and Pattern Recognition (CVPR), 2026.

[PDF | Project Website | Dataset | Source code (Coming soon)] 

Keywords:  self-supervised learning, 3D understanding

Empowering Dynamic Urban Navigation with Stereo and Mid-Level Vision 

Wentao Zhou, Xuweiyi Chen, Vignesh Rajagopal, Jeffrey Chen,  Rohan Chandra, Zezhou Cheng

arXiv: 2512.10956

[PDF | Project Website | Source code] 

Keywords:  mid-level vision, robot navigation

Point-MoE: Towards Cross-Domain Generalization in 3D Semantic Segmentation via Mixture-of-Experts 

Xuweiyi Chen, Wentao Zhou, Aruni RoyChowdhury, Zezhou Cheng

The International Conference on Learning Representations (ICLR) 2026

[PDF | Project Website] 

Keywords:  3D understanding 

Semantic-Free Procedural 3D Shapes Are Surprisingly Good Teachers 

Xuweiyi Chen, Zezhou Cheng

International Conference on 3D Vision (3DV) 2026

[PDF | Project Website | Dataset | Source code] 

Keywords:  Self-supervised learning, 3D representation learning 

Open Vocabulary Monocular 3D Object Detection 

Jin Yao, Hao Gu, Xuweiyi Chen, Jiayun Wang, Zezhou Cheng

International Conference on 3D Vision (3DV) 2026

[PDF | Project Website | Source code] 

Keywords: 3D recognition 

2025

Frame In-N-Out: Unbounded Controllable Image-to-Video Generation 

Boyang Wang, Xuweiyi Chen, Matheus Gadelha, Zezhou Cheng

Conference on Neural Information Processing Systems (NeurIPS) 2025

[PDF | Project Website | Source code] 

Keywords:  Controllable video generation 

LabelAny3D: Label Any Object 3D in the Wild 

Jin Yao, Radowan Mahmud Redoy, Sebastian Elbaum, Matthew B. Dwyer, Zezhou Cheng

Conference on Neural Information Processing Systems (NeurIPS) 2025

[PDF | Project Website | Source code | 3D Annotation Tool] 

Keywords:  3D understanding 

Probing the Mid-level Vision Capabilities of Self-Supervised Learning 

Xuweiyi Chen, Markus Marks, Zezhou Cheng

Computer Vision and Pattern Recognition (CVPR), 2025.

[PDF | Project Website | Source code] 

Keywords:  Self-supervised learning, 3D representation learning 

SAB3R: Semantic-Augmented Backbone in 3D Reconstruction 

Xuweiyi Chen, Tian Xia, Sihan Xu, Jed Jianing Yang, Joyce Chai, Zezhou Cheng

3D-LLM/VLA Workshop @ CVPR 2025

[PDF | Project Website] 

Keywords:  3D understanding 

E-BARF: Bundle Adjusting Neural Radiance Fields from a Moving Event Camera 

Zhipeng Tang, Shifan Zhu, Zezhou Cheng, Donghyun Kim, Erik Learned-Miller

Event-based Vision @ CVPR 2025

[PDF] 

Keywords:  3D understanding 

Before 2024 Fall 

Using spatio-temporal information in weather radar data to detect and track communal bird roosts 

Gustavo Perez, Wenlong Zhao, Zezhou Cheng, Maria Carolina Belotti, Yuting Deng, Victoria Simons, Elske Tielens, Jeffrey Kelly, Kyle Horton, Subhransu Maji, Daniel Sheldon

Remote Sensing in Ecology and Conservation. 2024; 

[PDF | Source code] 

Keywords: AI for ecology

LU-NeRF: Scene and Pose Estimation by Synchronizing Local Unposed NeRFs 

Zezhou Cheng, Carlos Esteves, Varun Jampani, Abhishek Kar, Subhransu Maji, Ameesh Makadia

IEEE International Conference on Computer Vision (ICCV), 2023.

[PDF | Project Website] 

Keywords: 3D pose estimation, Neural rendering

Accidental Turntables: Learning 3D Pose by Watching Objects Turn 

Zezhou Cheng,  Matheus Gadelha, Subhransu Maji

ICCV workshop on Recovering 6D Object Pose (ICCV workshop), 2023.

[PDF | Project Website | Poster] 

Keywords: unsupervised learning; 3D pose estimation.

ZeoNet: 3D convolutional neural networks for predicting adsorption in nanoporous zeolites 

Yachan Liu*, Gustavo Perez*, Zezhou Cheng, Aaron Sun, Samuel Hoover, Wei Fan, Subhransu Maji, Peng Bai (*equal contribution)

Journal of Materials Chemistry A, 2023.

[PDF | Project Website] 

Keywords: 3D shape understanding; AI for Chemistry 

A Semi-Automated System to Annotate Communal Roosts in Large-Scale Weather Radar Data 

Wenlong Zhao, Gustavo Perez, Zezhou Cheng, Maria Carolina Belotti, Yuting Deng, Victoria Simons, Elske Tielens, Jeffrey Kelly, Kyle Horton, Subhransu Maji, Daniel Sheldon

NeurIPS 2023 Computational Sustainability: Pitfalls and Promises from Theory to Deployment

[PDF | Project Website] 

Keywords: AI for ecology 

Cross-Modal 3D Shape Generation and Manipulation 

Zezhou Cheng, Menglei Chai, Jian Ren, Hsin-Ying Lee, Kyle Olszewski, Zeng Huang, Subhransu Maji, Sergey Tulyakov

European Conference on Computer Vision (ECCV) 2022

[PDF | Source code | Project Website | Poster | Video] 

Keywords: 3D shape generation and manipulation

Long-term analysis of persistence and size of swallow and martin roosts in the US Great Lakes 

Maria Belotti, Yuting Deng, Wenlong Zhao, Victoria Simons, Zezhou Cheng, Gustavo Perez, Elske Tielens, Subhransu Maji, Daniel Sheldon, Jeffrey Kelly, Kyle Horton

Remote Sensing in Ecology and Conservation, 2022

[PDF] 

Keywords: AI for ecology

Quantifying Long-term Phenological Patterns of Aerial Insectivores Roosting in the Great Lakes Region using Weather Surveillance Radar 

Yuting Deng, Maria Belotti, Wenlong Zhao, Zezhou Cheng, Gustavo Perez, Elske Tielens, Victoria Simons, Daniel Sheldon, Subhransu Maji, Jeff Kelly, Kyle Horton

Global Change Biology, 2022 (impact factor: 13.21)

[PDF] 

Keywords: AI for ecology

Improving Few-Shot Part Segmentation using Coarse Supervision 

Oindrila Saha, Zezhou Cheng, Subhransu Maji

European Conference on Computer Vision (ECCV) 2022

[PDF | Source code | Project Website] 

Keywords: Self-/Semi-/Weakly-supervised learning

GANORCON: Are Generative Models Useful for Few-shot Segmentation? 

Oindrila Saha, Zezhou Cheng, Subhransu Maji

Computer Vision and Pattern Recognition (CVPR), 2022.

[PDF | Source code | Project Website]

Keywords: Self-/Semi-/Weakly-supervised learning

On Equivariant and Invariant Learning of Object Landmark Representations 

Zezhou Cheng,  Jong-Chyi Su, Subhransu Maji

IEEE International Conference on Computer Vision (ICCV), 2021.

[PDF | Source code | Project Website | arXiv  | Poster  | Supplementary]

Keywords: Self-/Semi-/Weakly-supervised learning

AI for conservation: learning to track birds with radar

Zezhou Cheng, Subhransu Maji, Daniel Sheldon

 ACM XRDS 2021

[PDF] 

Keywords: AI for ecology

A Realistic Evaluation of Semi-Supervised Learning for Fine-Grained Classification

 Jong-Chyi Su, Zezhou Cheng, Subhransu Maji

Computer Vision and Pattern Recognition (CVPR), 2021, (Oral).

[PDF | Source code | arXiv | Poster]

Keywords: Self-/Semi-/Weakly-supervised learning

Detecting and Tracking Communal Bird Roosts in Weather Radar Data

Zezhou Cheng, Saadia Gabriel, Pankaj Bhambhani, Daniel Sheldon, Subhransu Maji, Andrew Laughlin, David Winkler

 Association for the Advancement of Artificial Intelligence (AAAI), 2020, (Oral)

[PDF | Source code | Project Website | arXiv | Supplementary] 

Keywords: AI for ecology

A Bayesian Perspective on the Deep Image Prior

Zezhou Cheng, Matheus Gadelha, Subhransu Maji, Daniel Sheldon 

Computer Vision and Pattern Recognition (CVPR), 2019.

[PDF | Source code | Project Website | arXiv | Poster | Supplementary ]

Keywords: Image restoration; Self-/Semi-/Weakly-supervised learning

Colorization Using Neural Network Ensemble 

Zezhou Cheng, Qingxiong Yang, and Bin Sheng 

IEEE Transactions on Image Processing (TIP), 2017.

[PDF | Souce code | Project Website]

Keywords: Image restoration

Deep Colorization

Zezhou Cheng, Qingxiong Yang, and Bin Sheng 

IEEE International Conference on Computer Vision (ICCV), 2015. 

[PDF | Souce code | Colorized Images (view online)] (My undergraduate thesis)

Keywords: Image restoration

Services

Reviewer:

Conferences: NeurIPS/ ICCV / CVPR / AAAI / WACV / ICLR / ICML etc.

Journals: IJCV / TIP / TMM, etc.

Area Chair / Senior PC:

AAAI / IJCAI

Awards

Travel Award from Doctoral Consortium, CVPR 2023 

Outstanding Reviewer, CVPR 2021

Outstanding Synthesis Award, UMass Amherst CICS, 2020 (top 2 synthesis) 

AAAI-2020 Student Scholarship  2020

Best Poster Award at the New England Computer Vision Workshop held at Brown University  2019

UMass Amherst CICS Edward Riseman and Allen Hanson Scholarship  2018

National Scholarship, Ministry of Education, China  2016 (at Shanghai Jiao Tong University)

National Scholarship, Ministry of Education, China  2014 (at Sichuan University)

Media Coverage

Climate Change AI: Deep learning of nanoporous materials for chemical separations

UMass CICS Outstanding Synthesis: CICS Graduate Students Honored for Outstanding Synthesis Projects


Google Sites
Report abuse
Google Sites
Report abuse