Solving geophysical inverse problems with measurement-guided diffusion models
Matteo Ravasi · 2025
Read Solving geophysical inverse problems with measurement-guided diffusion models by Matteo Ravasi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-04881.txt.
Description
Matteo Ravasi tests two ways to steer diffusion models toward earth data that fits the measurements. Diffusion models can dream up many plausible underground maps from sparse seismic reads. The steering math is the hard part, since the guide term has no clean form. His 2025 paper compares two samplers on seismic interpolation and seismic inversion. One guides with an adjoint push on the leftover error. The other uses a pseudo-inverse built on a fuzzy one-step guess.
Two lessons emerge from wide trials. The inverse problem must be rescaled so the target sits in a fixed range. Sampler choice then shapes how honest the spread of answers is. The work serves teams who need not just one map, but a fair sense of doubt.
Catalog record
| CREATOR | Matteo Ravasi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 75 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-04881