๐ณ๐๐๐ โ ๐ฝ๐๐๐๐๐๐ \tt Deep\,\,Needlet : A CNN based full sky component separation method in Needlet space
Debabrata Adak ยท 2025
Read ๐ณ๐๐๐ โ ๐ฝ๐๐๐๐๐๐ \tt Deep\,\,Needlet : A CNN based full sky component separation method in Needlet space by Debabrata Adak (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-2501-07469.txt.
Description
Debabrata Adak, writing alone in 2025, used a neural network to clean the oldest light in the universe from foreground glare. Microwave background maps come mixed with emissions from our own galaxy and beyond. A classic method blends filtered maps by minimum variance; the new one learns the blend.
The network works in a wavelet-like needlet space that keeps both position and scale in view, and it respects the sky's rotational symmetry. Training used realistic simulations at the Planck satellite's frequencies. Recovered maps tracked the true signal closely, with slight leakage near the galactic plane, and the power spectrum held to fine scales. On real Planck data the map agreed with the mission's legacy products. Foreground leftovers shrank below the classic method's.
Catalog record
| CREATOR | Debabrata Adak |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-07469