---
# For reference on dataset card metadata, see the spec: https://github.com/huggingface/hub-docs/blob/main/datasetcard.md?plain=1
# Doc / guide: https://huggingface.co/docs/hub/datasets-cards
{}
---

# Dataset Card for Arabic Digits

<!-- Provide a quick summary of the dataset. -->

## Dataset Details

### Dataset Description

<!-- Provide a longer summary of what this dataset is. -->
This dataset contains 70,000 Arabic handwritten digits, written by 700 participants. It is intended for Arabic digit recognition tasks using machine learning. The dataset is split into a training set of 60,000 images and a test set of 10,000 images, covering 10 Arabic digits (labeled 0–9). Each digit was written ten times by each writer. The images are in grayscale, 28×28 pixels, and were collected from different institutions to ensure diversity in handwriting styles. The dataset is derived from the MADBase database.


- **License:** Open Database License (ODbL)

### Dataset Sources

<!-- Provide the basic links for the dataset. -->

- **Homepage:** https://github.com/mloey/Arabic-Handwritten-Digits-Dataset
- **Paper:** El-Sawy, A., El-Bakry, H., & Loey, M. (2017). CNN for handwritten arabic digits recognition based on LeNet-5. In Proceedings of the International Conference on Advanced Intelligent Systems and Informatics 2016 2 (pp. 566-575). Springer International Publishing.

## Dataset Structure

<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->

Total images: 70,000

Splits:

- **Train**: 60,000 images (85.7%)

- **Test**: 10,000 images (14.3%)

Classes (labels): 10 (Arabic digits), labeled 0–9

Image specs: PNG format, 28×28 pixels, grayscale

## Example Usage
Below is a quick example of how to load this dataset via the Hugging Face Datasets library.
```
from datasets import load_dataset

# Load the dataset
dataset = load_dataset("randall-lab/arabic-digits", split="train", trust_remote_code=True)
# dataset = load_dataset("randall-lab/arabic-digits", split="test", trust_remote_code=True)

# Access a sample from the training set
example = dataset["train"][0]
image = example["image"]
label = example["label"]

image.show()  # Display the image
print(f"Label: {label}")
```

## Citation

<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->

**BibTeX:**

@inproceedings{el2017cnn,
  title={CNN for handwritten arabic digits recognition based on LeNet-5},
  author={El-Sawy, Ahmed and El-Bakry, Hazem and Loey, Mohamed},
  booktitle={Proceedings of the International Conference on Advanced Intelligent Systems and Informatics 2016 2},
  pages={566--575},
  year={2017},
  organization={Springer}
}