Chen Geng | Stanford CS

Chen Geng ? My first name is Chen, and my last name is Geng.
I prefer to be addressed by my first name Chen. I also go by "Ken" sometimes.
Possible pronunciation: Chen(ch-uhn) Geng(guh-ng).

Hi👋! I'm a CS Ph.D. student at Stanford, fortunate to be advised by Prof. Jiajun Wu. Previously, I got my bachelor's degree at Zhejiang University, working with Prof. Xiaowei Zhou and Prof. Sida Peng. My research is in part supported by an NVIDIA Graduate Fellowship.

My research lies at the intersection of 4D Computer Vision, Graphics, and Machine Learning. I'm broadly interested in data-driven modeling of the physical world and applications of such models in robotics and natural science. Currently, I'm obsessed with developing neural simulators for the (inverse) modeling of macroscopic mechanical systems.

Email: X × Y, where X = {gengchen}, Y = {@cs.stanford.edu}

Bio / Google Scholar / X / Bluesky / GitHub / LinkedIn

Recent News 📰

04/2026: Invited talk at the Simulation Intelligence Group at CMU on "Rethinking Inverse Simulation through the Lens of Lagrangian Mechanics".
02/2026: Four papers are accepted to CVPR 2026.
01/2026: One paper is accepted to ICLR 2026.
01/2026: We are organizing the 2nd workshop on 4D Vision at CVPR 2026. Submit your work on 4D vision to our workshop and share it with the community!
01/2026: This year I will co-organize the Stanford Vision Seminar. Subscribe to our mailing list here!
12/2025: Excited to announce that I received the NVIDIA Graduate Fellowship for 2026-2027!
04/2025: We will be organizing the workshop on "Generating Digital Twins from Images and Videos" at ICCV 2025.
04/2025: Our paper "Birth and Death of a Rose" is selected as an oral presentation at CVPR 2025.
03/2025: One paper is accepted to SIGGRAPH 2025.
02/2025: Two papers are accepted to CVPR 2025.
02/2024: One paper is accepted to CVPR 2024.
01/2024: One paper is accepted to ICLR 2024.

Invited Talks 🎤

Rethinking Inverse Simulation through the Lens of Lagrangian Mechanics [Abstract]

Sep 2026	Gradient Spaces Lab, Stanford University Host: Iro Armeni
Aug 2026	Physical Vision Group, Nanyang Technological University Host: Chuanxia Zheng
Jul 2026	Department of Computer Science, University of British Columbia & University of Toronto Hosts: Peter Chen (UBC), Eitan Grinspun and David Levin (UofT)
Apr 2026	Simulation Intelligence Group, Carnegie Mellon University Host: Minchen Li

Recent Research 🔬 (show selected / show all)

(* denotes equal contribution, ^ denotes student (co-)mentored, representative works are highlighted)
For the comprehensive list, check out my Google Scholar page.

	NeuROK: Generative 4D Neural Object Kinematics Chen Geng, Guangzhao He, Yue Gao, Yunzhi Zhang, Shangzhe Wu, Jiajun Wu CVPR 2026* [Paper] [Project Page] tl;dr: We turn any static 3D shape into an interactable 4D asset with a neural simulator built on the minimal inductive bias of Lagrangian mechanics.
	Choreographing a World of Dynamic Objects Yanzhe Lyu^, Chen Geng, Karthik Dharmarajan, Yunzhi Zhang, Hadi Alzayer, Shangzhe Wu, Jiajun Wu CVPR 2026 [Paper] [Project Page] tl;dr: We propose a universal pipeline for generating 4D scenes composed of dynamic objects by distilling from video generative models.
	ART: Articulated Reconstruction Transformer Zizhang Li, Cheng Zhang, Zhengqin Li, Henry Howard-Jenkins, Zhaoyang Lv, Chen Geng, Jiajun Wu, Richard Newcombe, Jakob Engel, Zhao Dong CVPR 2026 [Paper] [Project Page] tl;dr: A transformer-based model for reconstructing articulated dynamic objects from images.
	Coupled Diffusion Sampling for Training-Free Multi-View Image Editing Hadi Alzayer, Yunzhi Zhang, Chen Geng, Jia-Bin Huang, Jiajun Wu CVPR 2026 [Paper] [Project Page] tl;dr: Multi-view image editing by sampling from two diffusion models of different modalities concurrently.
	GenFusion: Feed-forward Human Performance Capture via Progressive Canonical Space Updates Youngjoong Kwon, Yao He, Heejung Choi, Chen Geng, Zhengmao Liu, Jiajun Wu, Ehsan Adeli ICLR 2026 [Paper] [Project Page] tl;dr: A feed-forward method for human performance capture that progressively updates a canonical space with incoming monocular RGB frames, using probabilistic regression to produce sharp novel-view renderings.
	Anymate: A Dataset and Baselines for Learning 3D Object Rigging Yufan Deng^, Yuhao Zhang^, Chen Geng, Shangzhe Wu†, Jiajun Wu† SIGGRAPH 2025 [Paper] [Project Page] [Demo] [Code] [arXiv] tl;dr: We propose a dataset and benchmark for supervised-learning-based 4D object rigging methods.
	Category-Agnostic Neural Object Rigging Guangzhao He^, Chen Geng, Shangzhe Wu, Jiajun Wu CVPR 2025 [Paper] [Project Page] [arXiv] tl;dr: We discover animatable motion subspaces for any 4D objects.
	Birth and Death of a Rose Chen Geng, Yunzhi Zhang, Shangzhe Wu, Jiajun Wu CVPR 2025 (Oral Presentation, 3.3% of the accepted papers) [Paper] [Project Page] [Code (Coming Soon)] [arXiv] tl;dr: We generate temporal 4D object intrinsics from 2D foundation models.
	Relightable and Animatable Neural Avatar from Sparse-View Video Zhen Xu, Sida Peng, Chen Geng, Linzhan Mou, Zihan Yan, Jiaming Sun, Hujun Bao, Xiaowei Zhou CVPR 2024 (Highlight, 11.9% of the accepted papers) [Paper] [Project Page] [Code] [arXiv] [Video] tl;dr: We estimate physically based intrinsics of dynamic characters from monocular videos.
	Neural Polynomial Gabor Fields for Macro Motion Analysis Chen Geng, Hong-Xing "Koven" Yu, Sida Peng, Xiaowei Zhou, Jiajun Wu ICLR 2024 [Paper] [Project Page] [Code (Coming Soon)] [OpenReview] tl;dr: We discover a low-dimensional interpretable motion representation for dynamic scenes with macro motion.
	Tree-Structured Shading Decomposition Chen Geng, Hong-Xing "Koven" Yu, Sharon Zhang, Maneesh Agrawala, Jiajun Wu ICCV 2023 [Paper] [Project Page] [Video] [Code] [MarkTechPost] tl;dr: We decompose the shading of objects into a tree-structured representation, which can be edited or interpreted by users easily. Abstract: We study the problem of obtaining a tree-structured representation for shading objects. Prior work typically uses the parametric or measured representation to model shading, which is neither interpretable nor easily editable. Our method uses the shade tree representation, which combines basic shading nodes and compositing methods, to model and decomposes the material shading. Such a representation enables users to edit previously rigid material appearances in an efficient and intuitive manner. In particular, novice users who are unfamiliar with the construction of such shade trees can quickly obtain such a representation. The extraction of such a representation enables the editing and understanding of object shading, even for novice users. The biggest challenge in this task is that the discrete structure of the shade tree is not differentiable. We propose a hybrid algorithm to address this issue. First, given an input image, a recursive amortized inference model is leveraged to initialize a guess of the tree structure and corresponding leaf node parameters. Then, we apply an optimization-based method to fine-tune the result. Experiments show that our method works well on synthetic images, realistic images, and non-realistic vector drawings, surpassing the baselines significantly.
	Learning Neural Volumetric Representations of Dynamic Humans in Minutes Chen Geng, Sida Peng, Zhen Xu, Hujun Bao, Xiaowei Zhou CVPR 2023* [Paper] [Project Page] [Code] tl;dr: We accelerate the learning of neural volumetric videos of dynamic humans by over 100 times. Abstract: This paper addresses the challenge of quickly reconstructing free-viewpoint videos of dynamic humans from sparse multi-view videos. Some recent works represent the dynamic human as a canonical neural radiance field (NeRF) and a motion field, which are learned from videos through differentiable rendering. They generally require a lengthy optimization process. Other generalization methods leverage learned prior from datasets and reduce the optimization time by only finetuning on new scenes, at the cost of visual fidelity. In this paper, we propose a novel method for creating viewpoint-free human performance synthesis from sparse view videos in minutes with competitive visual quality. Specifically, we leverage the human body prior to define a novel part-based voxelized NeRF representation, which distributes the representational power of the canonical human model efficiently. Furthermore, we propose a novel dimensionality reduction 2D motion parameterization scheme to increase the convergence rate of the human deformation field. Experiments demonstrate that our approach can be trained 100 times faster than prior per-scene optimiztion methods while being competitive in the rendering quality. We show that given a video capturing a human performer of 100 frames, our model typically takes about 5 minutes for training to produce photorealistic free-viewpoint videos on a single RTX 3090 GPU. The code will be released for reproducibility.
	Implicit Neural Representations with Structured Latent Codes for Human Body Modeling Sida Peng, Chen Geng, Yuanqing Zhang, Yinghao Xu, Qianqian Wang, Qing Shuai, Xiaowei Zhou, Hujun Bao TPAMI 2023 [Paper] [Code] [IEEE Xplore] tl;dr: Our approach reconstruct geometry and appearance of human performers with high accuracy from sparse observations.
	Novel View Synthesis of Human Interactions from Sparse Multi-view Videos Qing Shuai, Chen Geng, Qi Fang, Sida Peng, Wenhao Shen, Xiaowei Zhou, Hujun Bao SIGGRAPH 2022 (Featured in the technical paper trailer) [Paper] [Bibtex] [Code] [Project Page] tl;dr: Given sparse multi-view videos of crowded scenes with multiple human performers, our approach is able to generate high-fidelity novel views and accurate instance masks. @inproceedings{multinb, title = {Novel View Synthesis of Human Interactions from Sparse Multi-view Videos}, author = {Qing, Shuai and Chen, Geng and Qi, Fang and Sida, Peng and Wenhao, Shen and Xiaowei, Zhou and Hujun, Bao}, booktitle = {SIGGRAPH Conference Proceedings}, year = {2022}, }

Experience 🧑‍🎓

	NVIDIA June 2026 - Present, Santa Clara, California Research Scientist Intern Hosts: Nicholas Sharp and Donglai Xiang
	Stanford University Sept 2023 - Present, Stanford, California PhD Candidate in Computer Science Advisor: Jiajun Wu
	Zhejiang University Sept 2019 - June 2023, Hangzhou, China B.Eng.(Honours) in Computer Science Chu Kochen Honors College (Mixed Class and ACEE) Cumulative GPA: 94.38/100, 3.99/4.0 Major GPA: 96.67/100, 4.0/4.0 Advisor: Xiaowei Zhou and Sida Peng

Professional Activities 🏛️

Co-organizer

2026	Stanford Vision Seminar
2026	CVPR Workshop on 4D Vision
2025	International Conference on 3D Vision (3DV)
2025	ICCV Workshop on Generating Digital Twins from Images and Videos
2024	Stanford GCafé (Graphics Lunch Seminar)

Mentorship & Service

2026	Faculty-Student Liaison at Stanford CS
2023	Stanford PhD Application Support Program for Underrepresented Group (SASP)
2023	Stanford CS Undergraduate Mentoring Program

Conference Reviewer

CVPR (2024–), ECCV (2024–), ICCV (2025–), CoRL (2026–), SIGGRAPH (2025, 2026), SIGGRAPH Asia (2024, 2025), ICML (2024–, Position Papers for 2026), ICLR (2025–), NeurIPS (2024–), NeurIPS Workshop Proposals (2025–), NeurIPS D&B (2024), ICRA (2024), 3DV (2023), AAAI (2024), Pacific Graphics (2024)

Journal Reviewer

RA-L	IEEE Robotics and Automation Letters
ToG	ACM Transactions on Graphics
T-PAMI	IEEE Transactions on Pattern Analysis and Machine Intelligence
TVCG	IEEE Transactions on Visualization and Computer Graphics
TMLR	Transactions on Machine Learning Research

Misc.

Finalist, Qualcomm Innovation Fellowship, 2024

This website is adapted from this awesome template
Last updated: July, 2026