Research map / Image generation / GANs
Learning Texture Manifolds with the Periodic Spatial GAN
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add periodic noise dimensions (ablation) | PSGAN · Text and honeycomb textures | periodicity reproduction | Helped |
| Add periodic noise dimensions (ablation) | PSGAN · Honeycomb texture | local hexagonal structure | Helped |
| Add global noise dimensions (ablation) | PSGAN · Merrigum House image | texture diversity | Helped |
| Use independent noise sampling without global dimensions (ablation) | SGAN · Merrigum House image | texture diversity | Hurt |
| Set periodic dimensions to zero (ablation) | PSGAN · Text and honeycomb textures | periodicity reproduction | Hurt |
| Set global dimensions to zero (ablation) | PSGAN · Multiple textures | texture selection | Hurt |
| Use more than 2 periodic dimensions (ablation) | PSGAN · Periodic texture learning | training convergence speed | Helped |
| Use more than 2 periodic dimensions (ablation) | PSGAN · Periodic texture learning | sinusoid beating artifacts | Hurt |
| Remove batch normalization from the discriminator (ablation) | PSGAN · Single-image training | result quality | Helped |
| Fix periodic wave numbers while varying global dimensions (ablation) | PSGAN · Facade texture morphing | periodicity consistency | Helped |
Closest papers by meaning
- Pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis (CVPR 2021 · similarity 0.89)
- Exemplar-Based Pattern Synthesis With Implicit Periodic Field Network (CVPR 2022 · similarity 0.89)
- TextureGAN: Controlling Deep Image Synthesis With Texture Patches (CVPR 2018 · similarity 0.89)
- GramGAN: Deep 3D Texture Synthesis From 2D Exemplars (NeurIPS 2020 · similarity 0.88)
- Convolutional Generation of Textured 3D Meshes (NeurIPS 2020 · similarity 0.87)
- What You See is What You GAN: Rendering Every Pixel for High-Fidelity Geometry in 3D GANs (CVPR 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.