DeformMaster: An Interactive Physics-Neural World Model for Deformable Objects from Videos
DeformMaster: An Interactive Physics-Neural World Model for Deformable Objects from Videos
This research presents DeformMaster, a video-derived interactive physics-neural world model designed to handle deformable objects.
The paper introduces DeformMaster, a video-derived interactive physics-neural world model for deformable objects that infers physical states from visual observations and supports future dynamics rollout, material-parameter variation, and novel-view synthesis, with experiments demonstrating performance gains over state-of-the-art baselines on real-world sequences.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- code: Reported pending Open source S5
- weights: not_reported
- dataset: not_reported
- paper: not_reported
- code: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
DeformMaster's ability to roll out future dynamics and render dynamic appearance, outperforming state-of-the-art baselines while supporting novel action rollout, material-parameter variation, and dynamic novel-view synthesis. Code and video demos are available at the project page: https://can-lee.github.io/deformmaster-web/
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org