_README_ #1

This dataset is released in conjunction with the paper "VoDaSuRe: A Large-Scale Dataset Revealing Domain Shift in Volumetric Super-Resolution"

The VoDaSuRe dataset consists of 32 volumetric scans of 16 samples acquired using laboratory CT (Lab-CT). It includes:
- High-resolution (HR) volumes
- Low-resolution (LR) volumes (unregistered)
- Registered and intensity-matched low-resolution (REG) volumes

For details on the features of VoDaSuRe, see the links below:
1. Project page - https://augusthoeg.github.io/VoDaSuRe/
2. ArXiv paper - https://arxiv.org/abs/2603.23153
3. GitHub repository (frameworks, pipelines, data loaders) - https://github.com/AugustHoeg/VoxelSR

--------

The dataset contains training and test splits stored under the "ome/train" and "ome/test" subfolders. 

All data is stored in the OME-Zarr format, a hierarchical, chunked storage format that supports efficient and lazy loading.

Each .zarr file has the following structure:

ome.zarr
├── HR  (High-resolution volume)
│   ├── 0  (full resolution)
│   ├── 1  (2× downsampled)
│   ├── 2  (4× downsampled)
│   └── 3  (8× downsampled)
│
├── LR  (Unregistered low-resolution volume)
│   ├── 0  (full resolution)
│   ├── 1  (2× downsampled)
│   ├── 2  (4× downsampled)
│   └── 3  (8× downsampled)
│
└── REG (Registered + intensity-matched low-resolution volume)
    ├── 0  (full resolution)
    └── 1  (2× downsampled)

---------

To download the data, click on a chosen file/folder, then click the download button in the top right corner.

Alternatively, the data can be easily fetched through the command line, for example with:
> wget https://archive.compute.dtu.dk/downloads/public/projects/VoDaSuRe

The data should be decompressed before use:
> cd VoDaSuRe && bash extract_files.sh

Note: The full dataset requires approximately 500 GB of disk space.

---------

If you use our dataset, please cite our work:

@article{hoeg2026vodasure,
  title={VoDaSuRe: A Large-Scale Dataset Revealing Domain Shift in Volumetric Super-Resolution},
  author={August Leander Høeg and Sophia Wiinberg Bardenfleth and Hans Martin Kjer and Tim Bjørn Dyrby and Vedrana Andersen Dahl and Anders Dahl},
  journal={Proceedings of the Computer Vision and Pattern Recognition Conference},
  year={2026},
  url={https://augusthoeg.github.io/VoDaSuRe/}
}