Does shallow geological knowledge help neural-networks to predict deep units?

Abstract

Geological interpretation of seismic images is a visual task that can be automated by training neural networks. While neural networks have shown to be effective at various interpretation tasks, a fundamental challenge is the lack of labeled data points in the subsurface. For example, the interpolation and extrapolation of well-based lithology using seismic images relies on a small number of known labels. Besides well-known data augmentation techniques, as well as regularization of the network output, we propose and test another approach to deal with the lack of labels. Non learning-based horizon trackers work very well in the shallow subsurface where seismic images are of higher quality and the geological units are roughly layered. We test if these segmented and shallow units can help train neural networks to predict deeper geological units that are not layered and flat. We show that knowledge of shallow geological units helps to predict deeper units when there are only a few labels for training using a dataset from the Sea of Ireland. We employ U-net based multi-resolution networks, and we show that these networks can be described using matrix-vector product notation in a similar fashion as standard geophysical inverse problems.

Deep connections between learning from limited labels & physical parameter estimation -- inspiration for regularization

“…Rather than standard tensor notation, we employ matrix-vector product descriptions to stay close to PDEconstrained optimization literature. We block-vectorize states and Lagrangian multipliers, while weight/parameter tensors are flattened into block-matrices, see [21,18,4]. To keep notation compact, we focus the ResNet [10] with timestep h. The network state y j at layer j is given by y j = y j−1 − hf (K j y j−1 ).…”