Skip to content

Dataset · keremberke

Car photos for license plate object detection

keremberke/license-plate-object-detection

Image dataset with bounding box annotations for car license plates, split into train, validation, and test sets.

What's inside

The dataset contains 8,823 images (6,176 in train, 1,765 in valid, and 882 in test) with a single license_plate annotation class for the object detection task. Data imported from Roboflow and ready to load via the Hugging Face datasets library.

What you can do with it

  • Training or fine-tuning object detectors (e.g., YOLO or RT-DETR) for license plate localization
  • Building the first stage of an Automatic License Plate Recognition (ALPR) pipeline before running OCR
Tasks
Object detection
Size
1K–10K rows
Downloads this month
1.8K
Likes
42

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("keremberke/license-plate-object-detection")