HyDeFuse: Provably Convergent Denoiser-Driven Hyperspectral Fusion

Sagar Kumar, Unni V. S, Kunal N. Chaudhury · 2025

Read HyDeFuse: Provably Convergent Denoiser-Driven Hyperspectral Fusion by Sagar Kumar, Unni V. S, Kunal N. Chaudhury (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-2509-02477.txt.

Description

Sagar Kumar, Unni V S, and Kunal Narayan Chaudhury present a 2025 image processing paper on fusing hyperspectral and multispectral satellite images. The first type records many colour bands but coarse detail; the second gives fine detail in fewer bands.

Their HyDeFuse method treats fusion as an inverse problem guided by a ready-made denoiser, which cleans images and acts as an implicit prior. Strong denoisers can destabilise iterations, so the design uses pseudo-linear denoisers with controlled behaviour. Applying the contraction mapping theorem, a classic fixed-point result, proves global linear convergence. Tests on public datasets confirm the theory and show competitive fusion quality. The work gives denoiser-based fusion a firmer mathematical footing for remote sensing use.

Catalog record

CREATORSagar Kumar, Unni V. S, Kunal N. Chaudhury
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE76 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-02477

Browse the full library · About the Tower of Babel