{"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-learning-inversion-a-next-generation","title":"Deep-learning inversion: a next generation seismic velocity-model building method","arxiv_id":"1902.06267","date":"2019-02-17","proceeding":null,"authors":["Fangshu Yang","Jianwei Ma"],"abstract":"Seismic velocity is one of the most important parameters used in seismic\nexploration. Accurate velocity models are key prerequisites for reverse-time\nmigration and other high-resolution seismic imaging techniques. Such velocity\ninformation has traditionally been derived by tomography or full-waveform\ninversion (FWI), which are time consuming and computationally expensive, and\nthey rely heavily on human interaction and quality control. We investigate a\nnovel method based on the supervised deep fully convolutional neural network\n(FCN) for velocity-model building (VMB) directly from raw seismograms. Unlike\nthe conventional inversion method based on physical models, the supervised\ndeep-learning methods are based on big-data training rather than\nprior-knowledge assumptions. During the training stage, the network establishes\na nonlinear projection from the multi-shot seismic data to the corresponding\nvelocity models. During the prediction stage, the trained network can be used\nto estimate the velocity models from the new input seismic data. One key\ncharacteristic of the deep-learning method is that it can automatically extract\nmulti-layer useful features without the need for human-curated activities and\ninitial velocity setup. The data-driven method usually requires more time\nduring the training stage, and actual predictions take less time, with only\nseconds needed. Therefore, the computational time of geophysical inversions,\nincluding real-time inversions, can be dramatically reduced once a good\ngeneralized network is built. By using numerical experiments on synthetic\nmodels, the promising performances of our proposed method are shown in\ncomparison with conventional FWI even when the input data are in more realistic\nscenarios. Discussions on the deep-learning methods, training dataset, lack of\nlow frequencies, and advantages and disadvantages of the new method are also\nprovided.","url_abs":"http://arxiv.org/abs/1902.06267v1","url_pdf":"http://arxiv.org/pdf/1902.06267v1.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-learning-inversion-a-next-generation","repo_url":"https://github.com/YangFangShu/FCNVMB-Deep-learning-based-seismic-velocity-model-building","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"seismic-imaging","task_name":"Seismic Imaging"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1902.06267","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}