Research on Art and Design Idea Generation and Design Inspiration Inspiration Relying on Intelligent Recommendation Algorithm

Abstract

Intelligent generation of art design ideas is a necessary way to promote the diversification of artistic expression, but there are problems such as imprecise intelligent recommendation and unsatisfactory generation effect. This study combines the generative adversarial network with TimeSVD++ algorithm for generating recommendation model of art design.TS-GAN model is to use TimeSVD++ algorithm to obtain the temporal dynamic features of user preference in the process of art design, and use them to construct the generative model. The generated features are then input into a discriminative model based on multi-layer fully connected neural networks, which improves the discriminative effect on art design styles. Performance verification and investigation are conducted to analyze the effectiveness of the TS-GAN model. The research found that compared with the current mainstream CFGAN model, the TS-GAN model outperformed 6.14%, 8.94%, and 9.60% respectively in the three indicators of P@10, N@10, and M@10. The mean value of subjects' satisfaction with the art design idea generation recommendation results was 87.02%. Integrating intelligent algorithms with art design can promote the development of art design to be more diversified, which makes art designers obtain more creativity and inspirational inspiration.