Downloads · 30 days
0
huajian-zeng/stablehand
stablehand is a machine learning model from huajian-zeng. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for pytorch. The card lists the license as other.
Pretrained checkpoints for StableHand on HOT3D and ARCTIC.
Downloads · 30 days
0
Access
Public
Updated Sep 7, 2026
Repo size
584 MB
Likes
1
Public
Click a slice to open those files.
.pt584 MB · 100%
From the Hugging Face model README
Pretrained checkpoints for StableHand on HOT3D and ARCTIC.
| Paper | arXiv 2605.18553 |
| Project page | https://huajian-zeng.github.io/projects/stablehand/ |
| Code | https://github.com/huajian-zeng/stablehand |
| Video | YouTube |
| Directory | Description |
|---|---|
dit_hot3d/ | HOT3D motion prior checkpoint and configuration |
dit_arctic/ | ARCTIC motion prior checkpoint and configuration |
qn/ | Quality Network checkpoint and configuration |
Follow the installation and MANO setup instructions, then run from the code repository:
bash scripts/download_pretrained.sh
bash scripts/download_data.sh
bash demo.sh
bash scripts/evaluate.sh all
The benchmark caches and example videos are available at huajian-zeng/stablehand-data.
@article{zeng2026stablehand,
title = {StableHand: Quality-Aware Flow Matching for World-Space Dual-Hand Motion Estimation from Egocentric Video},
author = {Zeng, Huajian and Yao, Chaohua and Zhang, Yuantai and Yang, Jiaqi and Potamias, Rolandos Alexandros and Zuo, Xingxing},
journal = {arXiv preprint arXiv:2605.18553},
year = {2026},
}
Research use only. Source datasets and upstream components retain their respective terms.