Year: 2025
Type: conference-paper
Abstract: Deep generative learning has emerged as a significant area of research in seismic data processing. Among the various generative neural network architectures, Generative Adversarial Networks (GANs) and Denoising Diffusion Probabilistic Models (DDPMs) have seen widespread application in tasks such as denoising, interpolation, imaging, and inversion. In this study, we compare the performance of GANs and DDPMs in predicting particle velocity components from the pressure component in Ocean Bottom Node (OBN) acquisitions. High-quality predictions could enable the use of single-component receivers in certain locations, potentially reducing the overall cost of the OBN surveys by reducing the need for more expensive multi-component nodes.
Institution: Universidade Federal do Rio Grande do Norte
Language: English
Subfield: Geophysics
Field: Earth and Planetary Sciences
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7125712349
