Computer Vision Based Multimedia Advertising Effect Enhancement Strategy in E-Commerce Practice

In the context of today’s Internet-developed era, novel advertisement design is part of the enterprise marketing means. To improve the effectiveness of multimedia advertising on e-commerce practices, the study developed a multimedia advertising design system based on a generative adversarial network. The study proposes a generative adversarial network mechanism for computer vision technology, constructs an intelligent generation model of advertisement images using the generative adversarial network, designs a multimedia advertisement image interaction system, and explores the visual communication effect of multimedia advertisements. In order to test the significance of the advertisement image generation model for the display of advertisement images, a performance comparison test is conducted with other image generation models, and then the multimedia advertisement proposed in this paper is put into e-commerce practice. The study shows that the advertisement image generation model can accurately generate multimedia advertisement images, and the generation of multimedia advertisements based on the antagonistic network has a significant effect on the pleasant mood of e-commerce customers (p>0.005). Therefore, the multimedia advertising enhancement strategy proposed in this paper is conducive to mobilizing customers’ positive mood in e-commerce, and its practical effect is significant.