# Visual WetlandBirds Dataset: Videos of 13 Wetland Bird Species with Behavior and Species Bounding-Box Labels, Alicante, Spain (University of Alicante)

## Facts

- Provider: Rodriguez-Juan et al. (University of Alicante)
- License: CC-BY-4.0 (https://zenodo.org/records/14639444)
- Size: 181 files
- Format: MP4, JSON Lines (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/academic-datasets/Visual-WetlandBirds-Dataset
- Access: public, readable without an account
- DOI: https://doi.org/10.5281/zenodo.14639444
- Category: Environmental Science
- Source: https://huggingface.co/datasets/academic-datasets/Visual-WetlandBirds-Dataset
- Page: https://scigantic.com/archive/e205886c-a450-565c-8eae-6987eca40f9b
- Keywords: Environmental Science, MP4, JSON Lines, Hugging Face

## About

Video dataset for fine-grained bird species identification and behavior recognition, recorded in the La Mata and El Hondo natural park wetlands of Alicante, southeastern Spain, for the CHAN-TWIN project by University of Alicante researchers (Rodriguez-Juan et al.). Each frame-level annotation gives a bird bounding box, an individual bird id, the species and the behavior (for example preening), so the data supports spatio-temporal action detection as well as classification. The 13 species include White Wagtail, Glossy Ibis, Squacco Heron, Black-winged Stilt, Yellow-legged Gull, Common Gallinule, Black-headed Gull, Eurasian Coot, Little Ringed Plover, Eurasian Magpie, Gadwall, Mallard and Northern Shoveler. This Hugging Face conversion holds 178 MP4 videos under videos/ and a JSONL annotation file (about 9.5 GB), loadable with datasets.load_dataset and torchvision or PyAV video decoding. Useful for automated wildlife monitoring, ecological camera footage analysis and behavior recognition models. Original dataset on Zenodo (DOI 10.5281/zenodo.14639444), licensed CC BY 4.0; cite Rodriguez-Juan et al., Visual WetlandBirds Dataset (arXiv 2501.08931).

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("academic-datasets/Visual-WetlandBirds-Dataset")
print(ds)
```

## Data structure

File types in sampled files: MP4, JSON Lines

Example file: `videos/metadata.jsonl`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/e205886c-a450-565c-8eae-6987eca40f9b
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/e205886c-a450-565c-8eae-6987eca40f9b
- Over MCP: `get_schema_card` with id `e205886c-a450-565c-8eae-6987eca40f9b` at https://api.scigantic.com/mcp

## Related topics

- [Environmental Science datasets](https://scigantic.com/topics/environmental-science)
- [JSON Lines datasets](https://scigantic.com/topics/json-lines)
- [MP4 datasets](https://scigantic.com/topics/mp4)
