Summary of Automatic Differentiation-based Full Waveform Inversion with Flexible Workflows, by Feng Liu et al.
Automatic Differentiation-based Full Waveform Inversion with Flexible Workflows
by Feng Liu, Haipeng Li, Guangyuan Zou, Junlun Li
First submitted to arxiv on: 30 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Signal Processing (eess.SP); Geophysics (physics.geo-ph)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents an open-source framework called Automatic Differentiation-based Full Waveform Inversion (ADFWI) that simplifies the design, development, and evaluation of novel approaches in full waveform inversion (FWI). ADFWI leverages automatic differentiation to compute gradients for wave equations in various types of media, including isotropic acoustic and vertically or horizontally transverse isotropic elastic. The framework also includes a suite of objective functions, regularization techniques, and optimization algorithms. Additionally, ADFWI is integrated with deep learning for implicit model reparameterization via neural networks, which introduces learned regularization and rapid estimation of uncertainty through dropout. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper makes it easier to use full waveform inversion (FWI) by creating a special tool called ADFWI. It uses something called automatic differentiation to help solve problems with wave equations in different types of materials. The tool also includes many different ways to measure how good the results are, and how to make sure they’re not too good or bad. It even works with deep learning to make the results better and more predictable. |
Keywords
» Artificial intelligence » Deep learning » Dropout » Optimization » Regularization
