Affine steerers for structured keypoint description

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Bökman, Georg, Edstedt, Johan, Felsberg, Michael, Kahl, Fredrik
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929474552987648
author Bökman, Georg
Edstedt, Johan
Felsberg, Michael
Kahl, Fredrik
author_facet Bökman, Georg
Edstedt, Johan
Felsberg, Michael
Kahl, Fredrik
contents We propose a way to train deep learning based keypoint descriptors that makes them approximately equivariant for locally affine transformations of the image plane. The main idea is to use the representation theory of GL(2) to generalize the recently introduced concept of steerers from rotations to affine transformations. Affine steerers give high control over how keypoint descriptions transform under image transformations. We demonstrate the potential of using this control for image matching. Finally, we propose a way to finetune keypoint descriptors with a set of steerers on upright images and obtain state-of-the-art results on several standard benchmarks. Code will be published at github.com/georg-bn/affine-steerers.
format Preprint
id arxiv_https___arxiv_org_abs_2408_14186
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Affine steerers for structured keypoint description
Bökman, Georg
Edstedt, Johan
Felsberg, Michael
Kahl, Fredrik
Computer Vision and Pattern Recognition
We propose a way to train deep learning based keypoint descriptors that makes them approximately equivariant for locally affine transformations of the image plane. The main idea is to use the representation theory of GL(2) to generalize the recently introduced concept of steerers from rotations to affine transformations. Affine steerers give high control over how keypoint descriptions transform under image transformations. We demonstrate the potential of using this control for image matching. Finally, we propose a way to finetune keypoint descriptors with a set of steerers on upright images and obtain state-of-the-art results on several standard benchmarks. Code will be published at github.com/georg-bn/affine-steerers.
title Affine steerers for structured keypoint description
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2408.14186