Postdoctoral Researcher @ vLAR Group

李金玺 Jinxi Li

The Hong Kong Polytechnic University

Physics Learning · 3D Vision · World Model

Email Google Scholar
Portrait of Jinxi Li

About Me

I am a Postdoctoral Researcher in the vLAR Group at The Hong Kong Polytechnic University, working with Prof. Bo Yang. My research focuses on physics learning, 3D vision, and world model. I am interested in building models that can reconstruct, understand, and predict the physical dynamics of 3D scenes from visual observations.

I received my Ph.D. from The Hong Kong Polytechnic University in January 2026. Before that, I obtained my M.S. in Electrical and Computer Engineering from Northeastern University, and my B.Econ. in Finance from Guanghua School of Management, Peking University.

Education

2021.09 - 2026.01

The Hong Kong Polytechnic University

Ph.D.

2019.09 - 2021.05

Northeastern University, College of Engineering

M.S., Electrical and Computer Engineering

2014.09 - 2018.07

Peking University, Guanghua School of Management

B.Econ., Finance

Publications

Accepted papers are listed in reverse chronological order. * indicates equal contribution.

FoundObj project thumbnail

FoundObj: Self-supervised Foundation Models as Rewards for Label-free 3D Object Segmentation

Zihui Zhang, Zhixuan Sun, Yafei Yang, Jinxi Li, Jiahao Chen, Bo Yang

International Conference on Machine Learning (ICML), 2026

EvObj project thumbnail

EvObj: Learning Evolving Object-centric Representations for 3D Instance Segmentation without Scene Supervision

Jiahao Chen, Zihui Zhang, Yafei Yang, Jinxi Li, Shenxing Wei, Zhixuan Sun, Bo Yang

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026

PhysInOne project thumbnail

PhysInOne: Visual Physics Learning and Reasoning in One Suite

Siyuan Zhou*, Hejun Wang*, Hu Cheng*, Jinxi Li*, Dongsheng Wang*, Junwei Jiang*, Yixiao Jin*, Jiayue Huang*, Shiwei Mao*, Shangjia Liu, Yafei Yang, Hongkang Song, Shenxing Wei, Zihui Zhang, Peng Huang, Shijie Liu, Zhengli Hao, Hao Li, Yitian Li, Wenqi Zhou, Zhihan Zhao, Zongqi He, Hongtao Wen, Shouwang Huang, Peng Yun, Bowen Cheng, Pok Kazaf Fu, Wai kit Lai, Jiahao Chen, Kaiyuan Wang, Zhixuan Sun, Ziqi Li, Haochen Hu, Di Zhang, Chun Ho Yuen, Bing Wang, Zhihua Wang, Chuhang Zou, Bo Yang

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026

arXiv GitHub

TRACE project thumbnail

TRACE: Learning 3D Gaussian Physical Dynamics from Multi-view Videos

Jinxi Li, Ziyang Song, Bo Yang

IEEE/CVF International Conference on Computer Vision (ICCV), 2025

arXiv GitHub

RayletDF project thumbnail

RayletDF: Raylet Distance Fields for Generalizable 3D Surface Reconstruction from Point Clouds or Gaussians Highlight

Shenxing Wei*, Jinxi Li*, Yafei Yang, Siyuan Zhou, Bo Yang

IEEE/CVF International Conference on Computer Vision (ICCV), 2025

arXiv GitHub

FreeGave project thumbnail

FreeGave: 3D Physics Learning from Dynamic Videos by Gaussian Velocity

Jinxi Li, Ziyang Song, Siyuan Zhou, Bo Yang

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025

arXiv GitHub

OSN project thumbnail

OSN: Infinite Representations of Dynamic 3D Scenes from Monocular Videos

Ziyang Song, Jinxi Li, Bo Yang

International Conference on Machine Learning (ICML), 2024

arXiv GitHub

NVFi project thumbnail

NVFi: Neural Velocity Fields for 3D Physics Learning from Dynamic Videos

Jinxi Li, Ziyang Song, Bo Yang

Conference on Neural Information Processing Systems (NeurIPS), 2023

arXiv GitHub

RayDF project thumbnail

RayDF: Neural Ray-surface Distance Fields with Multi-view Consistency

Zhuoman Liu, Bo Yang, Yan Luximon, Ajay Kumar, Jinxi Li

Conference on Neural Information Processing Systems (NeurIPS), 2023

arXiv GitHub

Probabilistic symmetry project thumbnail

Probabilistic Symmetry for Improved Trajectory Forecasting

Sophia Sun, Robin Walters, Jinxi Li, Rose Yu

Conference on Learning for Dynamics and Control (L4DC), 2023

arXiv GitHub

ECCO project thumbnail

Trajectory Prediction using Equivariant Continuous Convolution

Robin Walters*, Jinxi Li*, Rose Yu

International Conference on Learning Representations (ICLR), 2021

arXiv GitHub