license: cc-by-4.0
pipeline_tag: robotics
SoMA: A Real-to-Sim Neural Simulator for Robotic Soft-body Manipulation
SoMA is a 3D Gaussian Splat simulator for soft-body manipulation. It couples deformable dynamics, environmental forces, and robot joint actions in a unified latent neural space for end-to-end real-to-sim simulation.
Modeling interactions over learned Gaussian splats enables controllable, stable long-horizon manipulation and generalization beyond observed trajectories without predefined physical models.
- Paper: https://arxiv.org/abs/2602.02402
- Project Page: https://city-super.github.io/SoMA/
Description
SoMA is a Gaussian splat neural simulator that models deformable object dynamics from real-world robot manipulation, enabling action-conditioned, stable long-horizon simulation with high-fidelity, multi-view–consistent rendering. By coupling deformable dynamics, environmental forces, and robot joint actions in a unified latent neural space, SoMA improves resimulation accuracy and generalization on real-world robot manipulation by 20%, enabling stable simulation of complex tasks such as long-horizon cloth folding.
Citation
If you find this work useful, please cite:
@article{huang2026soma,
title={SoMA: A Real-to-Sim Neural Simulator for Robotic Soft-body Manipulation},
author={Huang, Mu and Wang, Hui and Kerui Ren and Linning Xu and Yunsong Zhou and Mulin Yu and Bo Dai and Jiangmiao Pang},
journal={arXiv preprint arXiv:2602.02402},
year={2026}
}