---
dataset_info:
  features:
  - name: text
    dtype: string
  - name: image
    dtype: image
  - name: difficulty
    dtype: string
  splits:
  - name: train
    num_bytes: 4410348228.488
    num_examples: 403744
  - name: valid
    num_bytes: 561605435.133
    num_examples: 52091
  - name: test
    num_bytes: 626949323.1
    num_examples: 59565
  download_size: 5528589124
  dataset_size: 5598902986.721001
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
---
# Original

## Brno Mobile OCR Dataset

Brno Mobile OCR Dataset (B-MOD) is a collection of 2 113 templates (pages of scientific papers). Those templates were captured using 23 various mobile devices under unrestricted conditions ensuring that the obtained photographs contain various amount of blurriness, illumination etc. In total, the dataset contains 19 725 photographs and more than 500k text lines with precise transcriptions. The template pages are divided into three subsets (training, validation and testing).

This dataset may be used for non-commercial research purpose only. If you publish material based on this dataset, we request you to include a reference to the paper:

[Kišš, M., Hradiš, M. and Kodym, O. Brno Mobile OCR Dataset. In: 2019 International Conference on Document Analysis and Recognition (ICDAR). September 2019, p. 1352–1357. ISSN: 1520-5363.](https://pero.fit.vutbr.cz/publications)

You can download the dataset and evaluate your OCR system below. Our OCR system is available on the github. If you have any question, please contact ikiss@fit.vutbr.cz or ihradis@fit.vutbr.cz.

# Note
This dataset is without test set as it has no labels. If you consider submitting to the leaderboard download the data from the original source: https://zenodo.org/records/15310982