{"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/deep-fluids-a-generative-network-for","title":"Deep Fluids: A Generative Network for Parameterized Fluid Simulations","arxiv_id":"1806.02071","date":"2018-06-06","proceeding":null,"authors":["Byung-soo Kim","Vinicius C. Azevedo","Nils Thuerey","Theodore Kim","Markus Gross","Barbara Solenthaler"],"abstract":"This paper presents a novel generative model to synthesize fluid simulations\nfrom a set of reduced parameters. A convolutional neural network is trained on\na collection of discrete, parameterizable fluid simulation velocity fields. Due\nto the capability of deep learning architectures to learn representative\nfeatures of the data, our generative model is able to accurately approximate\nthe training data set, while providing plausible interpolated in-betweens. The\nproposed generative model is optimized for fluids by a novel loss function that\nguarantees divergence-free velocity fields at all times. In addition, we\ndemonstrate that we can handle complex parameterizations in reduced spaces, and\nadvance simulations in time by integrating in the latent space with a second\nnetwork. Our method models a wide variety of fluid behaviors, thus enabling\napplications such as fast construction of simulations, interpolation of fluids\nwith different parameters, time re-sampling, latent space simulations, and\ncompression of fluid simulation data. Reconstructed velocity fields are\ngenerated up to 700x faster than re-simulating the data with the underlying CPU\nsolver, while achieving compression rates of up to 1300x.","url_abs":"http://arxiv.org/abs/1806.02071v2","url_pdf":"http://arxiv.org/pdf/1806.02071v2.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":"deep-fluids-a-generative-network-for","repo_url":"https://github.com/ccsi-toolset/deeperfluids","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":null,"task_name":"CPU"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1806.02071","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}