OliseNS/AerialBirdDetection_2ndPlace_NexusLADevDays

---
license: cc-by-4.0
task_categories:
- object-detection
tags:
- computer-vision
- aerial-imagery
- drone-imagery
- helicopter-imagery
- uav
- top-down-view
- nadir-view
- birds-on-ground
- ground-level
- wildlife
- avian
- bird-nests
- nesting-sites
- ornithology
- yolo
- yolo-format
- second-place-winner
- nexus-la-devdays
- nestscope
- hackathon
pretty_name: "Aerial Bird Detection — 2nd Place Winner (Nexus LA DevDays)"
---

# Aerial Bird Detection - 2nd Place Nexus LA DevDays

## Overview
This dataset was curated and modified to train computer vision models for detecting birds and nests in aerial and drone imagery. It was developed as part of the **NestScope** application by team **SONA**, which secured 2nd place at the Nexus LA DevDays competition.

This repository contains a hybrid dataset: it combines original aerial imagery and annotations collected by Olisemeka Nmarkwe with a specifically modified subset of the global "Big Bird" drone imagery dataset.

## Dataset Composition
* **Original Contributions:** Custom dataset additions, modifications, and formatting by [Olisemeka Nmarkwe](https://olisemeka.dev/).
* **Base Dataset:** Modified subset of *Big Bird: A global dataset of birds in drone imagery annotated to species level. Version 1.0*.

## Intended Use
This dataset is optimized for training lightweight object detection models (such as YOLO) to identify avian wildlife and nesting sites from top-down aerial perspectives.

---

## How to Use This Dataset
The dataset is packaged as a single `.zip` file containing standard YOLO format directories (`images/` and `labels/`). You can download and extract it directly into your training environment using Python or your terminal.

### Option 1: Python (Recommended for Runpod Jupyter Notebook / Google Colab VScode Extension)
Make sure you have the Hugging Face Hub library installed (`pip install huggingface_hub`).

```python
from huggingface_hub import hf_hub_download
import zipfile

# 1. Download the zip file
print("Downloading dataset...")
zip_path = hf_hub_download(
repo_id="OliseNS/AerialBirdDetection_2ndPlace_NexusLADevDays",
filename="Dataset.zip",
repo_type="dataset"
)

# 2. Extract it to a local folder
print("Extracting files...")
with zipfile.ZipFile(zip_path, 'r') as zip_ref:
zip_ref.extractall("./nexus_bird_data")

print("Dataset ready for YOLO training in ./nexus_bird_data!")
```

### Option 2: Terminal / CLI
If you prefer to download and unzip directly from the command line:

```bash
# Download the zip file
wget [https://huggingface.co/datasets/OliseNS/AerialBirdDetection_2ndPlace_NexusLADevDays/resolve/main/uq_modified_dataset_stage_2.zip](https://huggingface.co/datasets/OliseNS/AerialBirdDetection_2ndPlace_NexusLADevDays/resolve/main/uq_modified_dataset_stage_2.zip)

# Unzip into a new directory
unzip uq_modified_dataset_stage_2.zip -d ./nexus_bird_data
```

Once extracted, you can point your YOLO `data.yaml` file to the newly created `nexus_bird_data` directory.

---

## License
This combined dataset is released under the **Creative Commons Attribution 4.0 International (CC BY 4.0)** license.

You are free to share (copy and redistribute the material in any medium or format) and adapt (remix, transform, and build upon the material) for any purpose, even commercially, under the following terms:
* **Attribution:** You must give appropriate credit, provide a link to the license, and indicate if changes were made.

### Required Acknowledgements & Citations
Because this repository contains modified data from the Big Bird dataset, any re-use of this dataset in research outputs, conference papers, or published research must explicitly acknowledge the original creators by including the following two citations:

**1. Dataset Citation:**
> Wilson, Joshua, Fuller, Richard, Francis, Roxane, Marchowski, Dominik, Pfeifer, Christian, Corregidor Castro, Alejandro, Luque Fernández, César, McEvoy, John, Sarasola, José Hernán, Oosthuizen, Chris, Hodgson, Jarrod, Rawson, Sophie, Sockhill, Nicola, and Wong-Topp, Samantha (2026). Big Bird: A global dataset of birds in drone imagery annotated to species level. Version 1.0. The University of Queensland. Data Collection. https://doi.org/10.48610/27809f1

**2. Associated Publication Citation:**
> Wilson, J. P., Amano, T., Bregnballe, T., Corregidor-Castro, A., Francis, R., Gallego-García, D., Hodgson, J. C., Jones, L. R., Luque-Fernández, C. R., Marchowski, D., McEvoy, J., McKellar, A. E., Oosthuizen, W. C., Pfeifer, C., Renner, M., Sarasola, J. H., Sokač, M., Valle, R., Zbyryt, A., & Fuller, R. A. (2026). Big Bird: A global dataset of birds in drone imagery annotated to species level. *Remote Sensing in Ecology and Conservation*. https://doi.org/10.1002/rse2.70059