“New progress has been made in the image description method of Dongba painting, and experts have proposed a method that combines prompt learning and visual semantic generative fusion, effectively improving the accuracy of description and cultural expression ability.”