README.md

metadata

license: cc-by-4.0
task_categories:
  - object-detection
pretty_name: DonaDataset

DonaDataset

Version: REPLACE_WITH_VERSION

Camera-trap image dataset for object detection, from Doñana National Park, Spain.

Dataset format

This dataset is distributed in YOLO format for object detection.

After downloading and extracting all tar shards into the same directory, the expected structure is:

.
├── donana.yaml
├── images/
│   ├── train/
│   ├── val/
│   └── test/
└── labels/
    ├── train/
    ├── val/
    └── test/

The YOLO configuration file uses relative paths:

path: .
train: images/train
val: images/val
test: images/test

Splits

• train: 3 images
• val: 3 images
• test: 3 images

Classes

• 0: Ave
• 1: Felis catus
• 2: Bos taurus
• 3: Canis familiaris
• 4: Herpestes ichneumon
• 5: Meles meles
• 6: Oryctolagus cuniculus
• 7: Dama dama
• 8: Genetta genetta
• 9: Equus sp
• 10: Lepus granatensis
• 11: Lynx pardinus
• 12: Cervus elaphus
• 13: Vulpes vulpes
• 14: Sus scrofa
• 15: Empty

Data files

The image and label files are stored inside tar shards under:

data/train/
data/val/
data/test/

Each tar file preserves YOLO-compatible internal paths, for example:

images/train/example.jpg
labels/train/example.txt

Integrity verification

The export includes:

checksums-sha256.txt
manifest.csv
manifest-files-sha256.csv
verification_report_local.json

Use these files to verify downloaded shards and internal image/label hashes.

Example extraction

From the directory containing the downloaded files:

mkdir -p extracted
cp donana.yaml extracted/

for f in data/train/*.tar data/val/*.tar data/test/*.tar; do
  tar -xf "$f" -C extracted/
done

Then train with a YOLO-compatible framework using:

extracted/donana.yaml

Hugging Face repository

wildintelproject/donadataset

Contributing

Issues and pull requests are welcome in the source code repository: https://github.com/wildintelproject/donadataset

Funding

This work is part of the WildINTEL project, funded by the Biodiversa+ Joint Research Call 2022-2023 "Improved transnational monitoring of biodiversity and ecosystem change for science and society (BiodivMon)".

License

This dataset is released under: CC-BY-4.0.

Citation

Please check CITATION.cff for citation information.

Limitations and ethical considerations

Before using this dataset, users should review possible limitations related to camera-trap data, class imbalance, annotation quality, environmental bias, privacy, and sensitive ecological information.

Zenodo Sandbox DOI

This dataset version has an associated Zenodo linked dataset record.

• DOI: 10.5281/zenodo.565913
• DOI URL: https://doi.org/10.5281/zenodo.565913
• Zenodo record: https://sandbox.zenodo.org/records/565913
• Record type: zenodo_linked_dataset_record

Note: this is a Zenodo Sandbox DOI for workflow testing only. It must not be used for formal citation.