EAGE Student Webinar on Self-supervised blind-spot denoising by Claire Emma Birnie

19 September 2022 -  9:00AM - 10:00AM CEST

Student Webinar by Claire Emma Birnie

Self-supervised blind-spot denoising 

In this lecture I will introduce blind-spot networks and discuss how they can be trained in a self-supervised manner, overcoming the common limitation that deep learning for seismic denoising tasks typically require noisy-clean training pairs. The lecture will begin with the simplest denoising example, random noise suppression, progressing into how blind-spot networks can be extended for coherent noise suppression. Finally, the lecture will conclude by introducing how these approaches can be further improved via the inclusion of physics into the denoising task, for a deblending example.