cs.CVNov 13, 2026

ALoMa: Affine Correspondences Beyond Patches

Authors: David Nordström, Georg Bökman, Viktor Larsson, Fredrik Kahl, Jonathan Ventura

Organizations: Chalmers University of Technology · University of Amsterdam · Lund University · California Polytechnic State University

Abstract

Affine correspondences (ACs) enable robust and efficient solvers for multiple-view geometry by constraining the local geometry of point correspondences (PCs). Despite their utility in minimal solvers, AC estimation remains understudied. We revisit AC estimation beyond local image patches, extending it to the image and image-pair levels. Rather than estimating the components of the affine transformation separately, we learn the full 4-DoF matrix jointly. Building on this formulation, we introduce ALoMa, an AC estimator using LoMa keypoint descriptions, demonstrating consistent improvements across a wide range of downstream tasks. Code at https://github.com/davnords/loma.

Figures & tables

Appendix figures & tables11 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. MVLoMa: Multi-View Local Feature Matching with Any Geometry Foundation Model

    Oct 22, 2026Xinyang Lu, Shengzhe Wei, Sheng Tang +3Image MatchingCross-View Image Matching

  2. SAMatcher: Co-Visibility Modeling with Segment Anything for Robust Feature Matching

    Jun 2, 2026Xu Pan, Qiyuan Ma, Mingyue Dong +3Multi-View GeometryImage Matching

  3. LoMa: Local Feature Matching Revisited

    Apr 6, 2026David Nordström, Johan Edstedt, Georg Bökman +6Multi-View GeometryImage Matching