Attention-guided generative adversarial networks for semantic extraction and synthesis of cultural heritage imagery

Image segmentation, as an important research direction in the field of digital image processing, is the foundation of many computer vision tasks. Its core goal is to classify pixels in images and extract valuable information. Against the backdrop of accelerating modernization, the inheritance and protection of intangible cultural heritage (ICH) graphics in Gansu face many challenges, and traditional protection methods are no longer able to meet the needs of the times. In recent years, the rise of AIGC (AI Generated Content) technology has provided a new solution for the digital protection and innovative inheritance of intangible cultural heritage graphics. This article proposes a graphic semantic extraction and generation algorithm for Gansu intangible cultural heritage based on AIGC technology. This algorithm combines deep learning (DL) technology to first extract semantic features of Gansu intangible cultural heritage graphics, capturing their unique cultural connotations and artistic styles; Subsequently, using Generative Adversarial Networks (GANs) for image generation, high-quality reproduction and innovative expression of intangible cultural heritage graphics can be achieved. The results indicate that the proposed algorithm performs well in both image generation accuracy and visual effects.