{"id":17408945,"url":"https://github.com/lukasmosser/stochastic_seismic_waveform_inversion","last_synced_at":"2025-04-15T17:33:02.577Z","repository":{"id":76534579,"uuid":"189036905","full_name":"LukasMosser/stochastic_seismic_waveform_inversion","owner":"LukasMosser","description":"Official Implementation of \"Stochastic seismic waveform inversion using generative adversarial networks as a geological prior\"","archived":false,"fork":false,"pushed_at":"2019-12-09T16:54:41.000Z","size":35067,"stargazers_count":49,"open_issues_count":2,"forks_count":16,"subscribers_count":4,"default_branch":"master","last_synced_at":"2025-03-29T00:23:10.574Z","etag":null,"topics":["bayesian-inference","generative-adversarial-network","geophysics","machine-learning"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/LukasMosser.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2019-05-28T13:45:58.000Z","updated_at":"2025-03-05T07:22:58.000Z","dependencies_parsed_at":"2023-07-17T02:00:23.233Z","dependency_job_id":null,"html_url":"https://github.com/LukasMosser/stochastic_seismic_waveform_inversion","commit_stats":null,"previous_names":[],"tags_count":1,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LukasMosser%2Fstochastic_seismic_waveform_inversion","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LukasMosser%2Fstochastic_seismic_waveform_inversion/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LukasMosser%2Fstochastic_seismic_waveform_inversion/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LukasMosser%2Fstochastic_seismic_waveform_inversion/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/LukasMosser","download_url":"https://codeload.github.com/LukasMosser/stochastic_seismic_waveform_inversion/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":249118953,"owners_count":21215647,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["bayesian-inference","generative-adversarial-network","geophysics","machine-learning"],"created_at":"2024-10-16T23:06:56.569Z","updated_at":"2025-04-15T17:33:00.419Z","avatar_url":"https://github.com/LukasMosser.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Stochastic seismic waveform inversion using generative adversarial networks as a geological prior\n\nAuthors: [Lukas Mosser](https://twitter.com/porestar), [Olivier Dubrule](https://www.imperial.ac.uk/people/o.dubrule), [Martin J. Blunt](https://www.imperial.ac.uk/people/m.blunt) \n\n[Pytorch](https://pytorch.org) implementation of [Stochastic seismic waveform inversion using generative adversarial networks as a geological prior](https://arxiv.org/abs/1806.03720)\n\n## Model Architecture\n\nThe model architecture consists of two parts:  \n - the generative adversarial network (implemented in [Pytorch](https://pytorch.org))   \n - the acoustic wave equation forward solver implemented in  ([Devito](https://www.opesci.org/devito)).  \nThe coupling between the two defines a fully differentiable computational graph.\n\n## Movie representation of samples from the prior\n\n\u003cimg src=\"https://github.com/LukasMosser/stochastic_seismic_waveform_inversion/raw/master/results/animations/movie_prior.gif\" width=\"400\"\u003e\n\n## Movie representation of samples from the posterior (27 sources)\n\n\u003cimg src=\"https://github.com/LukasMosser/stochastic_seismic_waveform_inversion/raw/master/results/animations/movie_posterior.gif\" width=\"400\"\u003e\n\n## Usage\n\nTo perform the inversion using the available pre-trained generator network use ``` apps/main_paper_version_revised.py ```  \n(Sorry for the long name, but older versions of the code were kept for reference purposes)\n  \n## Trained Models\nPre-trained models are available in the  [checkpoints](checkpoints/) directory.\n\n## Results and Data\n\nThe resulting datasets are available in this [Google Drive](https://drive.google.com/drive/folders/1xLkLwDxAGVmfz-o2DzImgr8fP0fQNHW4?usp=sharing)  \n \nEach run was made reproducible by setting the run-number = seed command-line argument.  \nComputations were performed on Imperial College CX1 supercomputing facilities.  \nTotal duration: 12 hours wall-time on 32-core nodes ~ 50 nodes simultaneously.  \n\n## Figures from paper\n\nThe figures from the paper can be reproduced using ```notebooks/Paper_Figures.ipynb```.  \nAll figures are located in ```results/figures```\n\n## Devito Optimizations\n\nThe library used to represent the forward solver has a number of optimizations that allow it to parallelize across\ncores and nodes using MPI.  \nWe suggest the following environment variables be set to maximize for performance:\n\n```\nDEVITO_OPENMP=\"1\";\nDEVITO_DLE=\"advanced\"\nDEVITO_LOGGING=\"INFO\"\nDEVITO_ARCH=\"gcc\"\n```\n\nAn example bash script used to perform the numerical computations on Imperial's CX1 cluster can be found in ```scripts/cluster_run.sh```\n\n## Citing\n\n```\n@article{mosser2018stochastic,\n  title={Stochastic seismic waveform inversion using generative adversarial networks as a geological prior},\n  author={Mosser, Lukas and Dubrule, Olivier and Blunt, Martin J},\n  journal={arXiv preprint arXiv:1806.03720},\n  year={2018}\n}\n```\n\n## Acknowledgements\n\nThe author would like to acknolwedge the developers of the [Devito](https://www.opesci.org/devito/).  \nIf you use their software, please acknowledge them in your references.  \nO. Dubrule would like to thank Total for seconding him as a visiting professor at Imperial College London.\n\n## License\n\n[MIT](LICENSE)","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flukasmosser%2Fstochastic_seismic_waveform_inversion","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Flukasmosser%2Fstochastic_seismic_waveform_inversion","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flukasmosser%2Fstochastic_seismic_waveform_inversion/lists"}