Seismic data reconstruction uncertainty from the unsupervised deep learning models

What is it about?

Seismic data reconstruction provides complete data that contributes to improved imaging quality. We introduce the Monte Carlo estimation method into the unsupervised deep learning framework. This integration framework reconstructs the seismic data while also providing confidence (uncertainty) in the reconstructed part. The proposed framework not only significantly improves the reconstruction accuracy compared to the traditional methods, but also yields the uncertainty value on the reconstructed seismic traces. This ability to estimate uncertainty is not available in most of traditional methods and conventional deep learning methods

Featured Image

Photo by Zbynek Burival on Unsplash

Photo by Zbynek Burival on Unsplash

Why is it important?

Reconstruction uncertainty from the proposed framework provides the confidence in the use of reconstructed seismic traces by subsequent processors.

Perspectives