---
license: mit
language:
- ar
tags:
- OCR
- VLM
- HTR
- Arabic
pretty_name: Muharaf-public
size_categories:
- 1K<n<10K
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
dataset_info:
  features:
  - name: id
    dtype: string
  - name: image
    dtype: image
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 3435611946
    num_examples: 1216
  download_size: 3063317838
  dataset_size: 3435611946
---
# Muharaf-public images
This dataset contains 1,218 full page images of Arabic handwriting and the corresponding text. The images are Manuscripts from the 19th to 21st century.
See the official code, paper, and zenodo archive below. Their work has been accepted to NeurIPS 2024.
[![Code](https://img.shields.io/badge/Code-black?style=flat-square&logo=github)](https://github.com/MehreenMehreen/muharaf) | [![Paper](https://img.shields.io/badge/Paper-black?style=flat-square&logo=arxiv)](https://arxiv.org/abs/2406.09630) | [![Dataset](https://img.shields.io/badge/Dataset-black?style=flat-square&logo=zenodo&logoColor=white&logoSize=auto)](https://zenodo.org/records/11492215)
:-------------------------:|:-------------------------:|:----

# How to use

```python
from datasets import load_dataset
import matplotlib.pyplot as plt

# Load your dataset in streaming mode to be loaded faster
# https://huggingface.co/docs/datasets/v1.10.1/dataset_streaming.html

ds = load_dataset("Omarkhaledok/muharaf-public-pages", streaming=True)

# Take one sample from the training set

sample = next(iter(ds['train']))

# Number of samples to visualize

SAMPLE_SIZE=1

# Display the image using matplotlib

plt.imshow(sample['image'])
plt.axis('off')  # Hide axes for better visualization
plt.show()

# Print the corresponding text

print("Corresponding text:")
print(sample['text'])

# # Example: iterate through 5 samples and visualize them
# # Uncode the following lines to visualize `SAMPLE_SIZE` of images and the corresponding text

# for i, sample in enumerate(ds['train']):
#   if i >= SAMPLE_SIZE:
#     break
#   plt.imshow(sample['image'])
#   plt.axis('off')
#   plt.title(f"Sample {i+1}")
#   plt.show()
#   print("Text:", sample['text'])
#   print("-"*50)

```
# Citation

If you are using this dataset in your work, please cite the official paper below.

```
@inproceedings{
  saeed2024muharaf,
  title={Muharaf: Manuscripts of Handwritten Arabic Dataset for Cursive Text Recognition},
  author={Mehreen Saeed and Adrian Chan and Anupam Mijar and Joseph Moukarzel and Gerges Habchi and Carlos Younes and Amin Elias and Chau-Wai Wong and Akram Khater},
  booktitle={The Thirty-eight Conference on Neural Information Processing Systems Datasets and Benchmarks Track},
  year={2024},
  url={https://openreview.net/forum?id=1s8l1tnTXW}
}
```