A Study on the Collision of Artificial Intelligence and Art Based on Generative Adversarial Networks (GAN)

The cross-collision of artificial intelligence and art has attracted significant attention in related fields, such as the migration and integration of painting styles. Due to the inherent differences of different painting techniques, direct application of existing methods does not bring satisfactory results. This paper proposes a multi-layer Pyramid Generative Adversarial Network (MPGAN), an end-to-end Generative Adversarial Network(GAN) based architecture, whose advantages are in image quality and better-scaled fusion of painting styles compared with state-of-the-art algorithms demonstrate the effectiveness of the method. The method's effectiveness is demonstrated by its advantages in image quality compared to the state-of-the-art algorithms and its better scaling fusion of painting styles.