External Division of Two Bregman Proximity Operators for Poisson Inverse Problems
Kazuki Haishima, Kyohei Suzuki, and Konstantinos Slavakis · 2026
Read External Division of Two Bregman Proximity Operators for Poisson Inverse Problems by Kazuki Haishima, Kyohei Suzuki, and Konstantinos Slavakis (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-2602-11482.txt.
Description
Recovering sparse vectors from Poisson-noisy straight models suffers bias under standard control. Kazuki Haishima and two co-authors offer an tool built by externally dividing two Bregman proximity tools, promoting sparsity while mitigating bias from classic absolute-value control. The tool plugs into the established NoLips recipe in place of its standard proximity step.
Two complementary reformulations clarify the tool's shape in primal and dual spaces for the Poisson problem. Numerical tests show more stable convergence than usual divergence-based approaches and greatly superior results on built data and an image restoration task. The work offers a principled plug-and-play upgrade for Poisson inverse problems.
Catalog record
| CREATOR | Kazuki Haishima, Kyohei Suzuki, and Konstantinos Slavakis |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-11482