← Back to Sessions

Advancing Seismic Full Waveform Inversion: A Hybrid Approach of Machine Learning and Physical Models for Improved Generalizability and Efficiency

Description: 

Computational seismic full waveform inversion (FWI) is crucial for energy exploration, civil infrastructure, Earthquake detection and early warning, and so on. However, nearly all of the earth’s interior is inaccessible to direct observation. Inference of unknown subsurface properties, therefore, relies on indirect and limited seismic measurements taken at or near the surface. The relevant data analysis capability for solving computational seismic imaging problems is inadequate, mainly due to the ill-posed nature of the problems and the high computational costs of solving them. Recently, machine learning (ML) based computational methods have been pursued in the context of scientific computational imaging problems. Some success has been attained when an abundance of simulations and labels are available. Nevertheless, ML models, trained using physical simulations, usually suffer from weak generalizability when applied to a moderately different real-world dataset. Moreover, obtaining corresponding training labels is typically prohibitively expensive due to the high demand for subject-matter expertise. On the other hand, different from problems in a typical computer vision context, many scientific imaging problems are governed by underlying physical equations. For example, the wave equation, describing how a wave signal is propagated through a subsurface medium over time, is the governing physics for seismic imaging problems. To fully unleash the power and flexibility of ML for solving large-scale computational seismic FWI problems, I have developed new computational methods to bridge the technical gap by addressing the critical issues of weak generalizability, label scarcity, and high training cost. In this talk, I will discuss the details of my R&D effort in leveraging both the power of machine learning and underlying physics. A series of numerical experiments are conducted using datasets from synthetic simulations to field applications to evaluate the effectiveness of the new FWI methods.

Session: Machine Learning for Full Waveform Inversion: From Hybrid to End-to-End Approaches - I

Type: Oral

Date: 5/3/2024

Presentation Time: 04:30 PM (local time)

Presenting Author: Youzuo

Student Presenter: No

Invited Presentation: Yes

Authors