Creator(s)
djw12@nyu.edu
Language(s)
Arabic
Centuries
17th, 18th, 19th, 20th
CER on Validation Set
15.4%
Size (Nr. of Words)
23,337
Model ID
57467
Rasam Dataset from the BULAC library (selected pages)
The Svoboda Diaries Project Data (complete)
Biographies of Noteworthy Persons from the QDL (selected pages)
Ground truth was provided openly by the first two projects. In the case of the third, it was transcribed by the HTR working group at NYU Abu Dhabi, especially Ibrahim Ali, Saqer Almarri, Fadi John, Duoaa Magdi Khalifa, Suphan Kirmizialtin, and David Wrisley.
You can use this model to automatically transcribe Handwritten documents with Handwritten Text Recgnition in Transkribus.