We provide additional results for both synthetic and real results, including UHD (4K). We further
demonstrate our generative model with text and image prompts conditioning, alongside a Nearest
neighbor analysis, showing that our method doesn’t simply copy from the dataset.
We also evaluate our different steps in the diffusion procedure with high resolution ablation
examples and show results of making non tileable photographs into tileable materials. Finally, we
show the real photographs used and where they were sources from.