Hybrid ABBA-GMRES for Unmatched Backprojectors in Large Scale X-Ray Computerized Tomography
Ryan Bentley, Mirjeta Pasha, Malena Sabaté Landman, Luisa Yang, Jeffery Zhang · 2026
Read Hybrid ABBA-GMRES for Unmatched Backprojectors in Large Scale X-Ray Computerized Tomography by Ryan Bentley, Mirjeta Pasha, Malena Sabaté Landman, Luisa Yang, Jeffery Zhang (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-2602-17892.txt.
Description
Ryan Bentley and team develop hybrid ABBA-GMRES methods for unmatched backprojectors in large-scale X-ray computerized tomography. Matrix-free repeat methods matter in tomography given the cost of forming the system matrix. Forward and back projectors use varied grids or speedups, yielding unmatched pairs. Those pairs violate adjoint rules behind old least-squares solvers, so runs no longer match a least-squares problem.
Prior work explored Krylov solvers such as AB-GMRES and BA-GMRES, showing semi-convergence. Under matched conditions, they reduce to LSQR and LSMR. This paper builds hybrid AB- and BA-GMRES methods folding Tikhonov regularization into iterations. It examines links to hybrid LSQR and LSMR variants. L-curve and broad cross validation rules for auto parameters are assessed for settling and images. Tests on 2D CT problems show the methods curb semi-convergence, yield better images, and stop steadier than rivals.
Catalog record
| CREATOR | Ryan Bentley, Mirjeta Pasha, Malena Sabaté Landman, Luisa Yang, Jeffery Zhang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-17892