Request Access to VolHuMe
VolHuMe is available for approved non-commercial research. Please review the license below before submitting your request. Access requests are manually reviewed.
By requesting access, you agree to the VolHuMe Dataset License, including its restrictions on redistribution, participant identification, and publication of identifiable qualitative results.
Log in or Sign Up to review the conditions and access this dataset content.
VolHuMe
VolHuMe: A High-Resolution Large Scale Dataset of Volumetric Human Meshes
VolHuMe is a volumetric human dataset designed for research in computer vision and computer graphics. It provides high-resolution human capture data and associated 3D representations for studying human reconstruction, modeling, and related tasks.
Project Website | Paper (ICIP 2026) | arXiv
Dataset Structure
The dataset is organized by subject, with separate archives for each available data modality.
VolHuMe/
├── README.md
├── LICENSE
├── AUTHORIZED_SUBJECTS.md
└── subjects/
├── <subject_id>/
│ ├── rgb.zip
│ ├── depth.zip
│ ├── pointclouds.zip
│ ├── scans.zip
│ └── rigged_avatar.zip
└── ...
Each subject directory may contain:
| Archive | Description |
|---|---|
rgb.zip |
Multi-view RGB images |
depth.zip |
Depth maps |
pointclouds.zip |
3D point clouds |
scans.zip |
Volumetric 3D meshes and associated data |
rigged_avatar.zip |
Rigged 3D avatar |
Release status: VolHuMe is being released incrementally. Additional subjects and data modalities, including SMPL-X fittings and animations, may be added as they become available. Not all archives are necessarily available for every subject.
For further details about the dataset and its acquisition methodology, please refer to the VolHuMe paper.
Access and Usage
VolHuMe is available for non-commercial scientific research and education under the terms of the license below.
Access is granted individually and requires manual approval. To request access, sign in to Hugging Face, review the license and Authorized Qualitative Subset, and complete the access request form.
Approved researchers may use the dataset for training, evaluation, benchmarking, and publication of scientific results, subject to the license conditions.
In particular:
- Quantitative results may be published provided they do not disclose participants' identities.
- Qualitative results may be published if participants are not reasonably recognizable.
- Identifiable qualitative results may only be published for subjects explicitly listed in AUTHORIZED_SUBJECTS.md.
- Redistribution of the dataset and attempts to identify or re-identify participants are prohibited.
License
VolHuMe Dataset License
For Non-Commercial Scientific Research
Copyright © 2026 University of Trento (MMLab) and Tampere University.
By accessing or using the VolHuMe dataset, you agree to the following terms.
1. Permitted Use
VolHuMe is made available free of charge for non-commercial scientific research and education.
Users may use the dataset for training, evaluation, benchmarking, and development of research methods, including the publication of scientific results.
Commercial use requires prior written authorization from the dataset providers.
2. Redistribution
Redistribution, resale, sublicensing, or public sharing of the dataset, in whole or in part, is prohibited without prior written permission.
Access to the dataset must not be shared with unauthorized users.
3. Privacy and Publication
Users must not attempt to identify or re-identify dataset participants or link their data to external identity information.
Quantitative results may be published provided they do not disclose participants' identities.
Qualitative results, including images, videos, meshes, renderings, and reconstructions, may be published under the following conditions:
- Authorized subjects: Identifiable results may be published only for the subject IDs explicitly listed in AUTHORIZED_SUBJECTS.md, subject to the conditions stated there.
- All other subjects: Qualitative results may be published only when the individuals are not reasonably recognizable.
Users must not publicly release derived data or trained models that expose identifiable participant information or reproduce substantial portions of the dataset.
4. Attribution
Publications using VolHuMe must cite:
@INPROCEEDINGS{11630100,
author={Martinelli, Giulia and Bisagno, Niccolò and Garau, Nicola and Rahtu, Esa and Conci, Nicola},
booktitle={2026 IEEE International Conference on Image Processing (ICIP)},
title={VolHuMe: A High-Resolution Large Scale Dataset of Volumetric Human Meshes},
year={2026},
pages={1-6},
doi={10.1109/ICIP61757.2026.11630100}
}
5. Disclaimer
The dataset is provided "AS IS", without warranties of any kind, to the extent permitted by applicable law. The providers are not responsible for misuse of the dataset.
The providers reserve the right to revoke access in the event of a license violation. No ownership rights are transferred to users.
6. Contact
For questions or commercial licensing requests:
By downloading or using VolHuMe, you acknowledge that you have read and agree to these terms.
Dataset Providers
VolHuMe is provided by:
- University of Trento — MMLab (Multimedia Signal Processing and Understanding Lab)
- Tampere University
For dataset inquiries, access issues, or privacy-related concerns, please contact:
giulia.martinelli-2@unitn.it | giulia.martinelli@tuebingen.mpg.de
- Downloads last month
- 398