nielsr HF Staff commited on
Commit
fed2b96
·
verified ·
1 Parent(s): c1d20fb

Improve model card: Add robotics pipeline tag, links, and description

Browse files

This PR enhances the model card by:
- Adding the `robotics` pipeline tag for better discoverability on the Hugging Face Hub.
- Confirming the existing `mit` license.
- Linking to the official paper: [Real-to-Sim Robot Policy Evaluation with Gaussian Splatting Simulation of Soft-Body Interactions](https://huggingface.co/papers/2511.04665).
- Providing a direct link to the project page: https://real2sim-eval.github.io/.
- Including a link to the GitHub repository for the code and detailed instructions: https://github.com/kywind/real2sim-eval.
- Adding a concise description of the framework's purpose and key features.
- Including the BibTeX citation for proper attribution.

These additions will make the artifact more informative and discoverable for users.

Files changed (1) hide show
  1. README.md +30 -3
README.md CHANGED
@@ -1,3 +1,30 @@
1
- ---
2
- license: mit
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ pipeline_tag: robotics
4
+ ---
5
+
6
+ # Real-to-Sim Robot Policy Evaluation with Gaussian Splatting Simulation of Soft-Body Interactions
7
+
8
+ This repository contains the official implementation of the **Real-to-Sim-Policy-Eval** framework, presented in the paper [Real-to-Sim Robot Policy Evaluation with Gaussian Splatting Simulation of Soft-Body Interactions](https://huggingface.co/papers/2511.04665).
9
+
10
+ This framework introduces a real-to-sim policy evaluation approach that constructs soft-body digital twins from real-world videos. It renders robots, objects, and environments with photorealistic fidelity using 3D Gaussian Splatting. The goal is to provide a reproducible, scalable, and accurate method for evaluating robotic manipulation policies, particularly for tasks involving deformable objects, which are typically challenging and costly to evaluate in the real world.
11
+
12
+ Key components of the framework include constructing simulation assets, deploying trained PhysTwin for soft-body simulation, policy training via a submodule, and evaluating policies within the constructed simulation environment.
13
+
14
+ For more detailed information, code, installation, and usage instructions, please refer to the following resources:
15
+ - **Paper**: [Real-to-Sim Robot Policy Evaluation with Gaussian Splatting Simulation of Soft-Body Interactions](https://huggingface.co/papers/2511.04665)
16
+ - **Project Website**: https://real2sim-eval.github.io/
17
+ - **GitHub Repository**: https://github.com/kywind/real2sim-eval
18
+
19
+ ## Citation
20
+
21
+ If you find this work useful, please consider citing the paper:
22
+
23
+ ```bibtex
24
+ @article{zhang2025real,
25
+ title={Real-to-Sim Robot Policy Evaluation with Gaussian Splatting Simulation of Soft-Body Interactions},
26
+ author={Zhang, Kaifeng and Sha, Shuo and Jiang, Hanxiao and Loper, Matthew and Song, Hyunjong and Cai, Guangyan and Xu, Zhuo and Hu, Xiaochen and Zheng, Changxi and Li, Yunzhu},
27
+ journal={arXiv preprint arXiv:2511.04665},
28
+ year={2025}
29
+ }
30
+ ```