Jin Cao

曹晋

My goal is to digitalize the physical world — building virtual worlds that power entertainment and robotics.

Portrait of Jin Cao

relighted by IC-Light

About· 01 ·

I'm a fourth-year undergraduate student at Xi'an Jiaotong University, majoring in CS. I am now a Research Intern at the Stanford Vision and Learning Lab (SVL), working closely with Koven Yu and Prof. Jiajun Wu on 3D visual generation. Meanwhile, I'm a Research Intern at ZJU3DV, Zhejiang University, working with Prof. Sida Peng and Prof. Xiaowei Zhou on robust 3D reconstruction. Previously, I worked with Prof. Xiangyong Cao and Prof. Deyu Meng on low-level image processing at Xi'an Jiaotong University.

My research interest is to build virtual worlds that digitalize the physical world and enable downstream applications such as entertainment and robotics. Toward this goal, I focus on learning world representations from multi-modal data and building generative models on top of them. Recently, I have been interested in long-context generative world modeling, physically-grounded generation, and controllable generation in support of this goal.

News· 02 ·

Selected Publications· 03 ·

Under Preparation

ShadowDancer: Teaching Video World Models Any Action by Learning Unified Dynamics Representations from a Video and Its Shadow

Jin Cao, Zian Meng, Kaipeng Zhang

* indicates equal contribution. See the full list on Google Scholar.

Education· 04 ·

The University of Texas at Austin

Incoming · Austin, USA

Incoming Ph.D. in Computer Science, advised by Prof. Qixing Huang and Prof. Georgios Pavlakos

Xi'an Jiaotong University

2022.09 – 2026.06 · Shaanxi, China

Undergraduate in Computer Science

Special Class for the Gifted Young 少年班

2021.09 – 2022.06 · Xi'an Jiaotong University

Admitted through China's elite program for gifted youth — ranked top 13 / 235, with direct admission to undergraduate study.

Experience· 05 ·

Alaya Lab, Shanda AI

2026.04 – Present · Shanghai, China

Research Intern at Alaya Lab with Dr. Kaipeng Zhang · Dynamics Representation Learning & Fully-Controllable Video Generation

Stanford Vision and Learning Lab (SVL)

2025.02 – 2026.06 · California, USA

Research Intern with Koven Yu and Prof. Jiajun Wu · Iterative 3D Scene Generation

ZJU3DV, Zhejiang University

2024.12 – 2025.10 · Zhejiang, China

Research Intern with Prof. Sida Peng and Prof. Xiaowei Zhou · Robust Radiance Field Reconstruction

Xi'an Jiaotong University

2024.03 – 2025.02 · Shaanxi, China

Research with Prof. Xiangyong Cao and Prof. Deyu Meng · Low-Level Image Processing (Super Resolution, Dehazing)