Gen4D: Synthesizing Humans and Scenes in the Wild
Jerrin Bright, Zhibo Wang, Yuhao Chen, Sirisha Rambhatla, John Zelek, David Clausi Vision and Image Processing Lab Critical ML Lab University of Waterloo, Canada {jerrin.bright, zhibo.wang, yuhao.chen1, sirisha.rambhatla, jzelek}@uwaterloo.ca · 2025
Read Gen4D: Synthesizing Humans and Scenes in the Wild by Jerrin Bright, Zhibo Wang, Yuhao Chen, Sirisha Rambhatla, John Zelek, David Clausi Vision and Image Processing Lab Critical ML Lab University of Waterloo, Canada {jerrin.bright, zhibo.wang, yuhao.chen1, sirisha.rambhatla, jzelek}@uwaterloo.ca (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-05397.txt.
Description
Vision systems need huge amounts of training footage, and rare activities like sports are hard to film at scale. Jerrin Bright and co-authors build Gen4D in a 2025 paper, a pipeline that fabricates moving human scenes automatically. Older synthetic sets lean on fixed asset libraries, which limits variety in bodies, motion, and settings.
Gen4D encodes motion from experts, grows avatars from text prompts with diffusion methods, and synthesizes backgrounds that fit the people in them. From that pipeline the team built SportPAL, a large set covering baseball, ice hockey, and soccer. No hand modeling or scene design was needed. The pair offers a scalable base for training vision tools aimed at real-world human activity.
Catalog record
| CREATOR | Jerrin Bright, Zhibo Wang, Yuhao Chen, Sirisha Rambhatla, John Zelek, David Clausi Vision and Image Processing Lab Critical ML Lab University of Waterloo, Canada {jerrin.bright, zhibo.wang, yuhao.chen1, sirisha.rambhatla, jzelek}@uwaterloo.ca |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-05397