While recent text-to-image models excel at generating realistic content, they struggle to capture the nuanced visual characteristics …