A study of the intersection of art and design in the advertising industry and computer science and technology

In this paper, computer science and technology based on Generative Adversarial Network (GAN) is applied to advertisement art design to realize the automatic generation of advertisement layout and image style migration, so as to improve the efficiency of advertisement design and art design effect. Based on the layout generation principle of LayoutGAN model, we propose an automatic advertisement layout generation model FL-GAN to realize the layout generation of apparel print advertisement. On the basis of the automatically generated advertisement layout, contour constraints are introduced, and a GAN-based advertisement image style migration model is further proposed to provide methods and assistance for advertisement art design. In the simulation experiments to test the effect of style migration, except for the water painting dataset, the FID scores of this paper’s advertisement image style migration model are better than those of other comparative methods, and the decoupling indexes PPL in the datasets of Tangka Flower Painting, Ink and Water Landscape Painting, Face Painting, Bonsai Painting, and Water Painting are 65, 45, 27.5, 79.1, and 48.6 respectively, which are the best among all the methods. In application practice, the cosmetic advertisement A1 designed with the style migration model of this paper has the longest gaze duration and gaze count data, both in the spokesperson and brand name interest area, and also in the advertising slogan and product interest area.