Baseer: A Vision-Language Model for Arabic Document-to-Markdown OCR

Khalil Hennara, Muhammad Hreden, Mohamed Motasim Hamed +4

Arabic document OCR remains a challenging task due to the language's cursive script, diverse fonts, diacritics, and right-to-left orientation. While modern Multimodal Large Languag…