A Review of the Research and Development of Adversarial Generative Networks in Interior Graphic Design

Beijing University of Posts and Telecommunications, International School, 100876 Beijing, China

^* Corresponding author: 2023213681@bupt.cn

Abstract

This study provides a comprehensive overview of the research and development of adversarial generative networks in interior graphic design. With the continuous development of adversarial generative networks, the level of Generative Adversarial Networks (GAN) has reached a very outstanding level, and it has also developed in interior graphic design. This article will be divided into four categories: early research, optimization methods for refining workflows, optimization methods for introducing graph networks, and other optimization methods. The framework, characteristics, advantages, and disadvantages of these methods will be introduced. Early research mainly focused on two relatively simple frameworks based on CGAN. The optimization methods for refining workflow are based on the degree of refinement, with a focus on introducing two representative articles and also mentioning some outstanding research results. The graph network section focuses on two studies, House GAN and House GAN++, while also mentioning FlooPlan GAN. The other optimization methods section introduces the introduction perspective of ActFloor GAN technology and energy-saving strategies. The final section summarizes the work of this article and provides prospects for future development.