
بإشراف: Dr. Mohammed Alsuhaibani
Despite recent advances in generative AI, creating high-quality Arabic calligraphy remains a challenge. Most models treat it as decorative art rather than a structured writing system, resulting in outputs that lack the precision and aesthetic balance of traditional calligraphy. We developed Medad, an AI model that transforms plain Arabic text into authentic calligraphic art. Users input text through a web interface, which then is converted to an image and then processed by a Vision Transformer (ViT) model. The model generates a calligraphy-style image while preserving the text’s structure and artistic integrity. A final enhancement step improves the image resolution for better visual quality. Medad reimagines Arabic calligraphy in the digital age, offering designers, artists, and developers a tool that combines cultural depth with modern AI capabilities.
Mustafa Sallat
412117552