SatBird
Species distribution model
Habitat suitability
Benchmarked Dataset
SatBird is a dataset and benchmark models for the task of predicting bird species encounter rates jointly at a specific location using remote sensing data.
| Model metadata | Value |
|---|---|
| Tool Type | Benchmarked Dataset |
| Broad Task | Species distribution model |
| Specific Task | Habitat suitability |
| Model type | Convolutional Neural Network, ResNet, SATLAS, SatMAE |
| Description | SatBird is a dataset and benchmark models for the task of predicting bird species encounter rates jointly at a specific location using remote sensing data. |
| Task Specific | Yes |
| Ecology Specific | Yes |
| Language(s) | Python |
| Last Edited Time | 5/23/24 7:16 |
| Related Publication(s) | https://doi.org/10.48550/arXiv.2311.00936 |
| Dependencies | pytorch, torchaudio, torchvision, etc. |
| Tool URL (Github etc.) | https://github.com/RolnickLab/SatBird/ |
| Last Update (time ago) | Last updated within 6 months |
| License | GPL-3.0 |
| Contact Name | Melisande Teng |
| Contact Email | mailto:tengmeli@mila.quebec |
| Contact Responsiveness | Very responsive |
| HuggingFace URL | nan |
| Reproducibility Method | nan |