# On Demand Solid Texture Synthesis Using Deep 3D Networks

arXiv: https://arxiv.org/abs/2001.04528 · HTML: https://mlsanity.com/paper/2001.04528/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 13 Jan 2020.

## Built on by (most cited first, 7 of 7)

- [Mesh Neural Cellular Automata](https://mlsanity.com/paper/2311.02820/) (2023)
- [G-Style: Stylized Gaussian Splatting](https://mlsanity.com/paper/2408.15695/) (2024)
- [STS-GAN: Can We Synthesize Solid Texture with High Fidelity from Arbitrary 2D Exemplar?](https://mlsanity.com/paper/2102.03973/) (2021)
- [Implicit Neural Representation of Tileable Material Textures](https://mlsanity.com/paper/2402.02208/) (2024)
- [ARTEMIS: Using GANs with Multiple Discriminators to Generate Art](https://mlsanity.com/paper/2311.08278/) (2023)
- [3D microstructural generation from 2D images of cement paste using generative adversarial networks](https://mlsanity.com/paper/2204.01645/) (2022)
- [PolyCrysDiff: Controllable Generation of Three-Dimensional Computable Polycrystalline Material Structures](https://mlsanity.com/paper/2603.11695/) (2026)

## Abstract (arXiv)

This paper describes a novel approach for on demand volumetric texture synthesis based on a deep learning framework that allows for the generation of high quality 3D data at interactive rates. Based on a few example images of textures, a generative network is trained to synthesize coherent portions of solid textures of arbitrary sizes that reproduce the visual characteristics of the examples along some directions. To cope with memory limitations and computation complexity that are inherent to both high resolution and 3D processing on the GPU, only 2D textures referred to as "slices" are generated during the training stage. These synthetic textures are compared to exemplar images via a perceptual loss function based on a pre-trained deep network. The proposed network is very light (less than 100k parameters), therefore it only requires sustainable training (i.e. few hours) and is capable of very fast generation (around a second for 2563 voxels) on a single GPU. Integrated with a spatially seeded PRNG the proposed generator network directly returns an RGB value given a set of 3D coordinates. The synthesized volumes have good visual results that are at least equivalent to the state-of-the-art patch based approaches. They are naturally seamlessly tileable and can be fully generated in parallel.