Solid Texture Synthesis using Generative Adversarial Networks

Solid texture synthesis (STS), as an effective way to extend 2D exemplar to a 3D solid volume, exhibits advantages in numerous application domains. However, existing methods generally synthesize solid texture with specific features, which may result in the failure of capturing diversified textural information. In this paper, we propose a novel generative adversarial nets-based approach (STS-GAN) to hierarchically learn solid texture with a feature-free nature. Our multi-scale discriminators evaluate the similarity between patch from exemplar and slice from the generated volume, promoting the generator to synthesize realistic solid textures. Experimental results demonstrate that the proposed method can generate high-quality solid textures with similar visual characteristics to the exemplar.