Skip to content

Dataset · wildintelproject

Wildlife camera trap images in YOLO format

wildintelproject/donadataset

Nearly 34 thousand camera trap frames from Doñana National Park (Spain) annotated with 15 animal categories.

What's inside

The dataset contains 33,914 camera trap images from Spain's Doñana National Park, divided into train, val, and test splits. The images contain 38,492 annotated objects (bounding boxes in YOLO format) across 15 animal categories — including wild mammals identified to species or genus (deer, wild boars, foxes, Iberian lynxes, rabbits, etc.) and domestic animals, as well as empty frames.

What you can do with it

  • Training object detectors (e.g., YOLO family) for automated camera trap frame analysis
  • Creating biodiversity monitoring systems and rare wildlife recognition
Tasks
Object detection
Size
10K–100K rows
Downloads this month
29K
Likes
0

How to load it

from datasets import load_dataset

ds = load_dataset("wildintelproject/donadataset")