Exemplar-based synthesis of geology using kernel discrepancies and generative neural networks
This work addresses the need for efficient geological image synthesis for subsurface reservoir modeling, but it is incremental as it combines existing kernel methods with generative neural networks.
The authors tackled the problem of generating realistic geological images from a single exemplar by minimizing patch distribution discrepancies using maximum mean discrepancy, and they achieved synthesis that reproduces visual patterns and spatial statistics of the exemplar.
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.