Annotated object detection dataset with bounding boxes for training YOLO, R-CNN, SSD, and other object detection models. Perfect for computer vision research and deep learning projects.
Everything you need for object detection model training
Precise bounding box coordinates for every object in each image with pixel-perfect accuracy.
10 different object categories including persons, vehicles, animals, and common objects.
Pre-split dataset with 80% training and 20% validation data for immediate use.
Ready-to-use annotations in YOLO format (txt files) for YOLOv5, YOLOv7, and YOLOv8.
Complete COCO JSON annotations for compatibility with Detectron2 and other frameworks.
High resolution PNG and JPG images with diverse lighting and background conditions.
Interactive visualization of dataset distribution
Draw bounding boxes and explore annotation formats
Click and drag to draw bounding boxes
Select a class from the dropdown above# Draw boxes to see annotations here
# Format: class_id x_center y_center width heightSample images with annotations from the dataset
Organized directory structure for easy integration
# YOLO Annotation Format
# class_id x_center y_center width height
# All values are normalized (0-1)
0 0.453125 0.546875 0.234375 0.687500
1 0.765625 0.421875 0.312500 0.234375
2 0.156250 0.718750 0.187500 0.25000010 annotated object categories in the dataset
Follow these steps to train your object detection model
Download and extract the dataset zip file to your project directory.
unzip object-detection.zip -d ./data/
Install required Python packages for your chosen framework.
pip install ultralytics opencv-python
Start training using YOLOv8 or your preferred framework.
yolo train model=yolov8n.pt data=data.yaml epochs=100
Use your trained model for object detection on new images.
yolo predict model=best.pt source=image.jpg
Get the complete object detection dataset with all annotations and start training your models today!