Synthesizing Designs With Inter-Part Dependencies Using Hierarchical Generative Adversarial Networks

Abstract

Real-world designs usually consist of parts with hierarchical dependencies, i.e., the geometry of one component (a child shape) is dependent on another (a parent shape). We propose a method for synthesizing this type of design. It decomposes the problem of synthesizing the whole design into synthesizing each component separately but keeping the inter-component dependencies satisfied. This method constructs a two-level generative adversarial network to train two generative models for parent and child shapes, respectively. We then use the trained generative models to synthesize or explore parent and child shapes separately via a parent latent representation and infinite child latent representations, each conditioned on a parent shape. We evaluate and discuss the disentanglement and consistency of latent representations obtained by this method. We show that shapes change consistently along any direction in the latent space. This property is desirable for design exploration over the latent space.

BézierGAN: Automatic Generation of Smooth Curves from Interpretable Low-Dimensional Parameters

“…As a generalization of the ellipse, superformula shapes are formed by periodic curves [19]. We generate two families of superformula shapes by using the following equations [3,20]:…”

“…Latent Regularity. Latent Space Consistency (LSC) measures the regularity of the latent space [20]. A high LSC indicates shapes change consistently along any direction in the latent space (e.g., a shape's roundness is monotonically increasing along one direction).…”