Koshur Pixel: a large-scale synthetic ocr dataset for kashmiri

Koshur Pixel is a newly released large-scale synthetic OCR dataset specifically designed for the Kashmiri language, consisting of 613,078 image-text pairs generated from the KS-PRET-5M corpus using the SynthOCR-Gen framework. The dataset features diverse fonts and text granularities, along with over 25 augmentation strategies to simulate real-world document conditions. This resource is crucial for developing OCR systems for low-resource languages, facilitating the digitization of Kashmiri texts and enhancing language technology applications in under-resourced linguistic contexts.