{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/texture-synthesis-with-spatial-generative","title":"Texture Synthesis with Spatial Generative Adversarial Networks","arxiv_id":"1611.08207","date":"2016-11-24","proceeding":null,"authors":["Nikolay Jetchev","Urs Bergmann","Roland Vollgraf"],"abstract":"Generative adversarial networks (GANs) are a recent approach to train\ngenerative models of data, which have been shown to work particularly well on\nimage data. In the current paper we introduce a new model for texture synthesis\nbased on GAN learning. By extending the input noise distribution space from a\nsingle vector to a whole spatial tensor, we create an architecture with\nproperties well suited to the task of texture synthesis, which we call spatial\nGAN (SGAN). To our knowledge, this is the first successful completely\ndata-driven texture synthesis method based on GANs.\n  Our method has the following features which make it a state of the art\nalgorithm for texture synthesis: high image quality of the generated textures,\nvery high scalability w.r.t. the output texture size, fast real-time forward\ngeneration, the ability to fuse multiple diverse source images in complex\ntextures. To illustrate these capabilities we present multiple experiments with\ndifferent classes of texture images and use cases. We also discuss some\nlimitations of our method with respect to the types of texture images it can\nsynthesize, and compare it to other neural techniques for texture generation.","url_abs":"http://arxiv.org/abs/1611.08207v4","url_pdf":"http://arxiv.org/pdf/1611.08207v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"texture-synthesis-with-spatial-generative","repo_url":"https://github.com/zalandoresearch/spatial_gan","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"texture-synthesis-with-spatial-generative","repo_url":"https://github.com/Bhargav4488/cvmlass4","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"texture-synthesis-with-spatial-generative","repo_url":"https://github.com/ubergmann/spatial_gan","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"texture-synthesis","task_name":"Texture Synthesis"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1611.08207","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}