Attribute-Conditioned Layout GAN for Automatic Graphic Design

Modeling layout is an important first step for graphic design. Recently, methods generating layouts have progressed, particularly with Generative Adversarial Networks (GANs). However, the problem of specifying locations and sizes design elements usually involves constraints respect to element attributes, such as area, aspect ratio reading-order. Automating attribute conditional remains a complex unsolved problem. In this article, we introduce Attribute-conditioned Layout GAN incorporate attributes generation by forcing both generator discriminator meet conditions. Due complexity designs, further propose dropout method make look at partial lists learn their local patterns. addition, various loss designs following different principles optimization. We demonstrate that proposed can synthesize conditioned on attributes. It also adjust well-designed new while retaining elements' original reading-orders. The effectiveness our validated through user study.