{"title":"用于改进稀疏采集数据地震成像和不确定性量化的生成扩散模型","authors":"Xingchen Shi, Shijun Cheng, Weijian Mao, Wei Ouyang","doi":"arxiv-2407.21683","DOIUrl":null,"url":null,"abstract":"Seismic imaging from sparsely acquired data faces challenges such as low\nimage quality, discontinuities, and migration swing artifacts. Existing\nconvolutional neural network (CNN)-based methods struggle with complex feature\ndistributions and cannot effectively assess uncertainty, making it hard to\nevaluate the reliability of their processed results. To address these issues,\nwe propose a new method using a generative diffusion model (GDM). Here, in the\ntraining phase, we use the imaging results from sparse data as conditional\ninput, combined with noisy versions of dense data imaging results, for the\nnetwork to predict the added noise. After training, the network can predict the\nimaging results for test images from sparse data acquisition, using the\ngenerative process with conditional control. This GDM not only improves image\nquality and removes artifacts caused by sparse data, but also naturally\nevaluates uncertainty by leveraging the probabilistic nature of the GDM. To\novercome the decline in generation quality and the memory burden of large-scale\nimages, we develop a patch fusion strategy that effectively addresses these\nissues. Synthetic and field data examples demonstrate that our method\nsignificantly enhances imaging quality and provides effective uncertainty\nquantification.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"74 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Generative Diffusion Model for Seismic Imaging Improvement of Sparsely Acquired Data and Uncertainty Quantification\",\"authors\":\"Xingchen Shi, Shijun Cheng, Weijian Mao, Wei Ouyang\",\"doi\":\"arxiv-2407.21683\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Seismic imaging from sparsely acquired data faces challenges such as low\\nimage quality, discontinuities, and migration swing artifacts. Existing\\nconvolutional neural network (CNN)-based methods struggle with complex feature\\ndistributions and cannot effectively assess uncertainty, making it hard to\\nevaluate the reliability of their processed results. To address these issues,\\nwe propose a new method using a generative diffusion model (GDM). Here, in the\\ntraining phase, we use the imaging results from sparse data as conditional\\ninput, combined with noisy versions of dense data imaging results, for the\\nnetwork to predict the added noise. After training, the network can predict the\\nimaging results for test images from sparse data acquisition, using the\\ngenerative process with conditional control. This GDM not only improves image\\nquality and removes artifacts caused by sparse data, but also naturally\\nevaluates uncertainty by leveraging the probabilistic nature of the GDM. To\\novercome the decline in generation quality and the memory burden of large-scale\\nimages, we develop a patch fusion strategy that effectively addresses these\\nissues. Synthetic and field data examples demonstrate that our method\\nsignificantly enhances imaging quality and provides effective uncertainty\\nquantification.\",\"PeriodicalId\":501270,\"journal\":{\"name\":\"arXiv - PHYS - Geophysics\",\"volume\":\"74 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-07-31\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Geophysics\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2407.21683\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Geophysics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2407.21683","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Generative Diffusion Model for Seismic Imaging Improvement of Sparsely Acquired Data and Uncertainty Quantification
Seismic imaging from sparsely acquired data faces challenges such as low
image quality, discontinuities, and migration swing artifacts. Existing
convolutional neural network (CNN)-based methods struggle with complex feature
distributions and cannot effectively assess uncertainty, making it hard to
evaluate the reliability of their processed results. To address these issues,
we propose a new method using a generative diffusion model (GDM). Here, in the
training phase, we use the imaging results from sparse data as conditional
input, combined with noisy versions of dense data imaging results, for the
network to predict the added noise. After training, the network can predict the
imaging results for test images from sparse data acquisition, using the
generative process with conditional control. This GDM not only improves image
quality and removes artifacts caused by sparse data, but also naturally
evaluates uncertainty by leveraging the probabilistic nature of the GDM. To
overcome the decline in generation quality and the memory burden of large-scale
images, we develop a patch fusion strategy that effectively addresses these
issues. Synthetic and field data examples demonstrate that our method
significantly enhances imaging quality and provides effective uncertainty
quantification.