Ground-truth material released for printed Sorani Kurdish OCR research by Saman Idrees and Hossein Hassani. The associated study reports 522 line-image/transcription pairs. The repository distributes the material in three archive parts.
Line images paired with text transcriptions and distributed in archive parts. The first inspected archive contains 178 PNG images and 178 text files.
Saman Idrees , Hossein Hassani (2021)
This study adapts an Arabic Tesseract LSTM model to printed Sorani Kurdish using limited Kurdish ground truth and synthetic font-based training material. The authors report 95.45% average character …
Idrees, S., & Hassani, H. (2021). Exploiting Script Similarities to Compensate for the Large Amount of Data in Training Tesseract LSTM: Towards Kurdish OCR.
Licensed under GPL stated by the paper; exact repository license version not independently verified