IRIS: A Bayesian Approach for Image Reconstruction in Radio Interferometry with expressive Score-Based priors.
Noé Dia, M. J. Yantovski-Barth, Alexandre Adam, Micah Bowles, Laurence Perreault-Levasseur, Yashar Hezaveh, Anna Scaife · 2025
Read IRIS: A Bayesian Approach for Image Reconstruction in Radio Interferometry with expressive Score-Based priors. by Noé Dia, M. J. Yantovski-Barth, Alexandre Adam, Micah Bowles, Laurence Perreault-Levasseur, Yashar Hezaveh, Anna Scaife (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-2501-02473.txt.
Description
Noé Dia and co-authors present, in a 2025 paper, a statistical method for turning noisy radio telescope data into pictures of planet-forming disks. Radio interferometers sample a sky image only indirectly, so rebuilding the picture is a guessing game with many plausible answers. Their method, IRIS, works in a Bayesian frame that reports not one image but a spread of likely ones. A score-based model trained on galaxy photos supplies the prior sense of what skies look like.
Paired with a noise model in the measurement domain, the prior yields samples of possible disk images from a real survey data set. Even though galaxy photos are the wrong training subject, the samples stayed plausible. Coverage tests on simulations checked that the uncertainty bands were honest. The authors offer the framework as a principled upgrade on classic radio imaging.
Catalog record
| CREATOR | Noé Dia, M. J. Yantovski-Barth, Alexandre Adam, Micah Bowles, Laurence Perreault-Levasseur, Yashar Hezaveh, Anna Scaife |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-02473