We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Exemplar-based synthesis of geology using kernel discrepancies and generative neural networks

Abstract: We propose a framework for synthesis of geological images based on an exemplar image. We synthesize new realizations such that the discrepancy in the patch distribution between the realizations and the exemplar image is minimized. Such discrepancy is quantified using a kernel method for two-sample test called maximum mean discrepancy. To enable fast synthesis, we train a generative neural network in an offline phase to sample realizations efficiently during deployment, while also providing a parametrization of the synthesis process. We assess the framework on a classical binary image representing channelized subsurface reservoirs, finding that the method reproduces the visual patterns and spatial statistics (image histogram and two-point probability functions) of the exemplar image.
Subjects: Machine Learning (stat.ML); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Applications (stat.AP)
Cite as: arXiv:1809.07748 [stat.ML]
  (or arXiv:1809.07748v2 [stat.ML] for this version)

Submission history

From: Shing Chan [view email]
[v1] Thu, 20 Sep 2018 17:33:20 GMT (3066kb,D)
[v2] Fri, 21 Sep 2018 09:31:45 GMT (3066kb,D)

Link back to: arXiv, form interface, contact.