253 citations · 265 across the 17 of their papers we have counts for
17 papers · 1 filter
A Unified Framework for Vision Transformers Equivariant to Discrete Subgroups of
Tīkun Ông, Georg Bökman
Vision transformers have become a dominant architecture for visual recognition. However, standard models do not explicitly encode the planar symmetries that arise in many vision do…
Who Handles Orientation? Investigating Invariance in Feature Matching
David Nordström, Johan Edstedt, Fredrik Kahl +1
Finding matching keypoints between images is a core problem in 3D computer vision. However, modern matchers struggle with large in-plane rotations. A straightforward mitigation is…
LoMa: Local Feature Matching Revisited
David Nordström, Johan Edstedt, Georg Bökman +6
Local feature matching has long been a fundamental component of 3D vision systems such as Structure-from-Motion (SfM), yet progress has lagged behind the rapid advances of modern d…
RoMa v2: Harder Better Faster Denser Feature Matching
Johan Edstedt, David Nordström, Yushan Zhang +7
Dense feature matching aims to estimate all correspondences between two images of a 3D scene and has recently been established as the gold standard due to its high accuracy and rob…
Platonic Transformers: A Solid Choice For Equivariance
Mohammad Mohaiminul Islam, Rishabh Anand, David R. Wessels +7
While widespread, Transformers lack inductive biases for geometric symmetries common in science and computer vision. Existing equivariant methods often sacrifice the efficiency and…
A Framework for Reducing the Complexity of Geometric Vision Problems and its Application to Two-View Triangulation with Approximation Bounds
Felix Rydell, Georg Bökman, Fredrik Kahl +1
In this paper, we present a new framework for reducing the computational complexity of geometric vision problems through targeted reweighting of the cost functions used to minimize…
Flopping for FLOPs: Leveraging equivariance for computational efficiency
Georg Bökman, David Nordström, Fredrik Kahl
Incorporating geometric invariance into neural networks enhances parameter efficiency but typically increases computational costs. This paper introduces new equivariant neural netw…
Affine steerers for structured keypoint description
Georg Bökman, Johan Edstedt, Michael Felsberg +1
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…
From 2D to 3D: AISG-SLA Visual Localization Challenge
Jialin Gao, Bill Ong, Darld Lwi +13
Research in 3D mapping is crucial for smart city applications, yet the cost of acquiring 3D data often hinders progress. Visual localization, particularly monocular camera position…
DeDoDe v2: Analyzing and Improving the DeDoDe Keypoint Detector
Johan Edstedt, Georg Bökman, Zhenjun Zhao
In this paper, we analyze and improve into the recently proposed DeDoDe keypoint detector. We focus our analysis on some key issues. First, we find that DeDoDe keypoints tend to cl…
Steerers: A framework for rotation equivariant keypoint descriptors
Georg Bökman, Johan Edstedt, Michael Felsberg +1
Image keypoint descriptions that are discriminative and matchable over large changes in viewpoint are vital for 3D reconstruction. However, descriptions output by learned descripto…
Leveraging Cutting Edge Deep Learning Based Image Matching for Reconstructing a Large Scene from Sparse Images
Georg Bökman, Johan Edstedt
We present the top ranked solution for the AISG-SLA Visual Localisation Challenge benchmark (IJCAI 2023), where the task is to estimate relative motion between images taken in sequ…
DeDoDe: Detect, Don't Describe -- Describe, Don't Detect for Local Feature Matching
Johan Edstedt, Georg Bökman, Mårten Wadenbäck +1
Keypoint detection is a pivotal step in 3D reconstruction, whereby sets of (up to) K points are detected in each view of a scene. Crucially, the detected points need to be consiste…
RoMa: Robust Dense Feature Matching
Johan Edstedt, Qiyu Sun, Georg Bökman +2
Feature matching is an important computer vision task that involves estimating correspondences between two images of a 3D scene, and dense methods estimate all such correspondences…
In Search of Projectively Equivariant Networks
Georg Bökman, Axel Flinth, Fredrik Kahl
Equivariance of linear neural network layers is well studied. In this work, we relax the equivariance condition to only be true in a projective sense. We propose a way to construct…
A case for using rotation invariant features in state of the art feature matchers
Georg Bökman, Fredrik Kahl
The aim of this paper is to demonstrate that a state of the art feature matcher (LoFTR) can be made more robust to rotations by simply replacing the backbone CNN with a steerable C…
ZZ-Net: A Universal Rotation Equivariant Architecture for 2D Point Clouds
Georg Bökman, Fredrik Kahl, Axel Flinth
In this paper, we are concerned with rotation equivariance on 2D point cloud data. We describe a particular set of functions able to approximate any continuous rotation equivariant…