Datasets:
metadata
license: agpl-3.0
language:
- en
tags:
- poultry
- disease-detection
- object-detection
- computer-vision
- veterinary
- agriculture
- yolo
task_categories:
- object-detection
size_categories:
- n<1K
Poultry Disease Detection Dataset
A custom-annotated dataset of poultry images for detecting 26 disease lesions and clinical signs across 3 major poultry diseases: Newcastle Disease, Coccidiosis, and Fowlpox.
Dataset Details
- ~406 labeled images in YOLO format
- 26 classes (lesions and clinical signs)
- Annotations created with CVAT
- Images from farm and veterinary postmortem contexts
Format
Labels are in YOLO format (.txt per image):
<class_id> <x_center> <y_center> <width> <height>
All coordinates normalized 0–1.
Classes
See dataset.yaml for the full class list.
Associated Model
Fine-tuned YOLOv11n trained on this dataset: Evet-Africa/poultry-disease-detector
License
AGPL-3.0 — must be open-sourced if used in derivative works.
Citation
@misc{poultry-disease-dataset-2025,
title = {Poultry Disease Detection Dataset},
year = {2025},
publisher = {Hugging Face},
howpublished = {\url{https://huggingface.co/datasets/Evet-Africa/poultry-disease-dataset}}
}