GLOCR: GeezLab OCR Dataset

A Text Recognition (TR) and Optical Character Recognition (OCR) dataset for the Tigrinya language.

Content

The dataset contains a total of 710k image-label pairs from multiple data sources. In addition to the characters-only data, the major part of the dataset is a collection of multi-word text images with labels from three categories: News (from Haddas Ertra newspaper), the Bible, and random-trigrams of the 150k most common words in Tigrinya.

Examples

Components:

• Tigrinya News text-lines dataset.
  • Samples: train (200k), dev (15k), and test (15k)
  • Download
• Tigrinya Bible text-lines dataset.
  • Samples: train (80k), dev (10k), and test (10k)
  • Download
• Tigrinya text-lines dataset from top 150k words.
  • Samples: train (150k), dev (15k), and test (15k)
  • Download
• Tigrinya Characters dataset.
  • Samples: train (120k), dev (15k), and test (15k)
  • Download
• Tigrinya unsegmented full-page dataset.
  • Samples: 506 scanned pages with the corresponding text
  • Download

Download

The entire dataset (>1.3GB) is published via Harvard Dataverse and can be downloaded from there: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/RQTSD2

Cite

If you use this dataset in your product or research, please cite as follows:

@data{DVN/RQTSD2_2021,
    author = {Gaim, Fitsum},
    publisher = {Harvard Dataverse},
    title = {{GLOCR: GeezLab OCR Dataset}},
    year = {2021},
    version = {V1},
    doi = {10.7910/DVN/RQTSD2},
    url = {https://doi.org/10.7910/DVN/RQTSD2}
}