The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
CryoFastAR Example Data and Pretrained Checkpoints
This repository hosts the example data and pretrained model checkpoints distributed for CryoFastAR inference experiments and research reproducibility.
CryoFastAR estimates particle poses for feed-forward cryo-EM ab initio reconstruction. Its accompanying paper appeared at ICCV 2025. Implementation details, inference scripts, and upstream licensing information are available in the project repository.
Files
| Archive | Size in bytes | Contents |
|---|---|---|
data.zip |
4,606,744,358 | Experimental example in real/10028/10028/ and synthetic example in synthetic/spliceosome/5nrl/ |
cryofastar-ckpts.zip |
3,971,274,829 | checkpoints/real.pth and checkpoints/synthetic.pth |
The experimental example includes NumPy arrays for particle images, poses, translations, and CTF parameters. The synthetic example includes particle images and poses.
Download and use
Download both archives with the Hugging Face CLI.
hf download cellverse/cryofastar-assets data.zip cryofastar-ckpts.zip --repo-type dataset --local-dir .
From the CryoFastAR project directory, extract the example data into data/. The checkpoint archive already includes the checkpoints/ directory.
unzip data.zip -d data
unzip cryofastar-ckpts.zip -d .
Follow the experimental or synthetic inference instructions in the project repository.
Direct downloads are available for data.zip and cryofastar-ckpts.zip.
License and use
The data and model weights are licensed under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0), following the upstream project's licensing for these assets.
Noncommercial academic research, teaching, evaluation, and reproducibility studies are permitted subject to the license. Users must provide appropriate attribution, link to the license, and identify modifications. Commercial use requires separate written permission from the relevant rights holders. The NonCommercial condition covers uses primarily intended for commercial advantage or monetary compensation.
For commercial licensing, the upstream project lists contact@cellverse.tech and jiakai0309@gmail.com.
The complete license is included in LICENSE. The Creative Commons license summary and legal code provide the authoritative terms. The license text in this repository was obtained from the SPDX license text collection. Applicable notices and rights associated with underlying third-party material must also be respected.
本数据集及预训练权重采用 CC BY-NC 4.0 许可。允许遵守许可条件的非商业学术研究、教学、评估和结果复现。使用时应注明作者及来源、提供许可证链接,并标明修改。商业用途须另行取得相关权利人的书面授权。完整条款见 LICENSE。
Paper and citation
Jiakai Zhang, Shouchen Zhou, Haizhao Dai, Xinhang Liu, Peihao Wang, Zhiwen Fan, Yuan Pei, and Jingyi Yu. CryoFastAR: Fast Cryo-EM Ab initio Reconstruction Made Easy. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2025, pages 8462–8471.
Read the paper at CVF Open Access or arXiv, 2506.05864.
Please cite this paper when using the resources in academic work. The entry is also available in citation.bib.
@inproceedings{Zhang_2025_ICCV,
author = {Zhang, Jiakai and Zhou, Shouchen and Dai, Haizhao and Liu, Xinhang and Wang, Peihao and Fan, Zhiwen and Pei, Yuan and Yu, Jingyi},
title = {{CryoFastAR}: Fast Cryo-EM Ab initio Reconstruction Made Easy},
booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
month = {October},
year = {2025},
pages = {8462--8471},
url = {https://openaccess.thecvf.com/content/ICCV2025/html/Zhang_CryoFastAR_Fast_Cryo-EM_Ab_initio_Reconstruction_Made_Easy_ICCV_2025_paper.html}
}
Integrity
SHA-256 digests of the uploaded archives are provided below.
fba9b18b041977641d07dbf3cba68b1a6b693b4c2bc2cefe7519be3d014b10fb data.zip
8f8f49ef081e6abb8183c7bb4be5e0bcb3df74e0e584e955d20e42d6fbbbdaed cryofastar-ckpts.zip
- Downloads last month
- 60