Haoran Li

About Me

I received my Ph.D. and Bachelor’s degree from the University of Science and Technology of China (USTC). During my Ph.D., I was supervised by Prof. Yong Liao, and previously worked with Prof. Pengyuan Zhou and Prof. Haiyong Xie. My research interests lie in world models, 3D scene generation and editing, 3D understanding, and video generation. I recently visited Nanyang Technological University (NTU), where I worked with Prof. Lin Wang on AI for Science and agent reasoning. I have now joined JD Explore Academy, working on the TGT project.

Research Interests

• World Models & Video Generation: 4D scene representations, real-time frame models, spatio-temporal consistency, long-horizon video prediction
• 3D Scene Generation & Editing: multi-object 3D scene synthesis, controllable layout generation, structure-aware 3D content editing
• 3D Perception & Understanding: multi-view geometry reasoning, 3D structure understanding, consistency across viewpoints and modalities
• AI for Science & Agent Reasoning: agent reasoning for scientific problems, model-based decision making, integrating generative models with scientific priors

Publications

* indicates equal contribution

1. 3D Generation& World Models

3D Perception & Understanding

AI for Science & Agent Reasoning

Experiences

JD Explore Academy
Apr. 2026 – Present
TGT Project

University of Science and Technology of China
Sep. 2020 – Mar. 2026
Ph.D.

University of Science and Technology of China
Sep. 2016 – Jul. 2020
Bachelor Degree

Tencent Interactive Entertainment Group (IEG)
Apr. 2024 – Jan. 2025
Research Intern (3D Scene Layout & 3D Generation)

Nanyang Technological University
Apr. 2025 – Oct. 2025
Visiting Researcher (AI for Science & Agent Reasoning)

Services

Reviewer for IEEE CVPR, ICCV, NeurIPS, ICME, and ICASSP

Reviewer for IEEE Transactions on Image Processing (TIP)

Powered by Jekyll and Minimal Light theme.