|
Autonomous Character-Scene Interaction Synthesis from Text Instruction
Nan Jiang*,
Zimo He*,
Zi Wang,
Hongjie Li,
Yixin Chen,
Siyuan Huang,
Yixin Zhu
SIGGRAPH Asia, 2024
[PDF]
 
[arXiv]
 
[Project]
We propose a framework synthesizing multi-stage scene-aware human motions autonomously and directly from the text instruction and goal location. We also introduce LINGO, a comprehensive language-annotated MoCap dataset, featuring Human-Scene Interaction (HSI) motions.
|
|
Scaling Up Dynamic Human-Scene Interaction Modeling
Nan Jiang*,
Zhiyuan Zhang*,
Hongjie Li,
Xiaoxuan Ma,
Zan Wang,
Yixin Chen,
Tengyu Liu,
Yixin Zhu,
Siyuan Huang
CVPR, 2024
(Highlight)
[PDF]
 
[arXiv]
 
[Code]
 
[Data]
 
[Project]
 
[Hugging Face]
We introduce TRUMANS, an extensive MoCap Human-Scene Interaction (HSI) dataset capturing a wide array of human behaviors across 100 indoor scenes, noted for its diversity, quality, and scalability. We further develop a diffusion-based autoregressive method for the real-time generation of HSI, adaptable to any length and conditioned on 3D scenes and action labels.
|
|
Stanford Vision and Learning Lab (SVL), Stanford University, USA
Jun 2024 - Present
Research intern
Advisor: Prof. Jiajun Wu
|
|
Beijing Institute for General Artificial Intelligence (BIGAI), China
Sept 2023 - Present
Research volunteer
Advisor: Dr. Siyuan Huang
|
|
Cognitive Reasoning (CoRe) Lab, Institute for Artificial Intelligence, Peking University, China
Jan 2023 - Present
Student research assistant
Advisor: Prof. Yixin Zhu
|
|
Peking University, China
Aug 2021 - Present
Junior Student
|
|