Physics-Guided Deep Learning For High Resolution X-ray Imaging
Shao Xian Lee, Aashwin Ananda Mishra, Ariel Arnott, Meriame Berboucha, Nina Boiadjieva, Gourab Chatterjee, Eric Cunningham, Nick Czapla, Gilliss Dyer, Jonathan Ehni, Robert Ettelbrick, Anna Grassi, Mickael Grech, Philip Hart, Dimitri Khaghani, Hae Ja Lee, Peregrine McGehee, Bob Nagler, Paul Neumayer, Caterina Riconda, Marc Welch, Andrea Zabala, Eric Galtier, Quynh L. Nguyen · 2026
Read Physics-Guided Deep Learning For High Resolution X-ray Imaging by Shao Xian Lee, Aashwin Ananda Mishra, Ariel Arnott, Meriame Berboucha, Nina Boiadjieva, Gourab Chatterjee, Eric Cunningham, Nick Czapla, Gilliss Dyer, Jonathan Ehni, Robert Ettelbrick, Anna Grassi, Mickael Grech, Philip Hart, Dimitri Khaghani, Hae Ja Lee, Peregrine McGehee, Bob Nagler, Paul Neumayer, Caterina Riconda, Marc Welch, Andrea Zabala, Eric Galtier, Quynh L. Nguyen (2026) 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-2605-01543.txt.
Description
Shao Xian Lee and 23 co-authors clean structured artifacts out of X-ray images from high-energy-density experiments, where each shot may be the only one of its kind. Drifting patterns in the imaging system ride on top of the real signal. Standard flat-field correction leaves their ghosts in the final transmission maps, biasing measures of density, speed, and feature size.
The team model the artifacts as a separable layer and train a U-Net to estimate that layer from each shot. Against Fourier filtering and a dynamic flat-field method, the physics-guided network lifts image similarity scores from poor to near unity in synthetic tests, while keeping thin filament profiles intact. Deep ensembles add uncertainty estimates for each rebuild, a guard flagging shots unlike the training data. The method targets fusion and high-energy-density labs that cannot afford a spoiled single shot.
Catalog record
| CREATOR | Shao Xian Lee, Aashwin Ananda Mishra, Ariel Arnott, Meriame Berboucha, Nina Boiadjieva, Gourab Chatterjee, Eric Cunningham, Nick Czapla, Gilliss Dyer, Jonathan Ehni, Robert Ettelbrick, Anna Grassi, Mickael Grech, Philip Hart, Dimitri Khaghani, Hae Ja Lee, Peregrine McGehee, Bob Nagler, Paul Neumayer, Caterina Riconda, Marc Welch, Andrea Zabala, Eric Galtier, Quynh L. Nguyen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.01543