Abstract: This disclosure describes a system and method for effectively training a machine learning model to identify features in DAS and/or seismic imaging data with limited or no human labels. This is accomplished using a masked autoencoder (MAE) network that is trained in multiple stages. The first stage is a self-supervised learning (SSL) stage where the model is generically trained to predict data that has been removed (masked) from an original dataset. The second stage involves performing additional predictive training on a second dataset that is specific to a particular geographic region, or specific to a certain set of desired features. The model is fine-tuned using labeled data in order to develop feature extraction capabilities.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating realistic synthetic seismic data items. One of the methods includes obtaining a plurality of synthetic seismic data items; obtaining a plurality of real seismic data items; processing each of the plurality of synthetic seismic data items using a machine learning model; processing each of the plurality of real seismic data items using the same machine learning model; determining a range for values for one or more parameters of a synthetic seismic data generator by comparing the synthetic seismic data items and the real seismic data items in an embedding space of the machine learning model; and selecting, as realistic synthetic seismic data items, a plurality of synthetic seismic data items that have been generated with a respective combination of values for the one or more parameters that is within the determined range.