Papers › DKM: Dense Kernelized Feature Matching for Geometry Estimation

DKM: Dense Kernelized Feature Matching for Geometry Estimation

1 Feb 2022CVPR 2023 1arXiv:2202.00667archive 2025-07-28

Johan Edstedt, Ioannis Athanasiadis, Mårten Wadenbäck, Michael Felsberg

Feature matching is a challenging computer vision task that involves finding correspondences between two images of a 3D scene. In this paper we consider the dense approach instead of the more common sparse paradigm, thus striving to find all correspondences. Perhaps counter-intuitively, dense methods have previously shown inferior performance to their sparse and semi-sparse counterparts for estimation of two-view geometry. This changes with our novel dense method, which outperforms both dense and sparse methods on geometry estimation. The novelty is threefold: First, we propose a kernel regression global matcher. Secondly, we propose warp refinement through stacked feature maps and depthwise convolution kernels. Thirdly, we propose learning dense confidence through consistent depth and a balanced sampling approach for dense confidence maps. Through extensive experiments we confirm that our proposed dense method, \textbf{D}ense \textbf{K}ernelized Feature \textbf{M}atching, sets a new state-of-the-art on multiple geometry estimation benchmarks. In particular, we achieve an improvement on MegaDepth-1500 of +4.9 and +8.9 AUC$@5^{\circ}$ compared to the best previous sparse method and dense method respectively. Our code is provided at https://github.com/Parskatt/dkm

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

parskatt/dkm officialmentioned in papermentioned on GitHubpytorchNOASSERTION report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Camera Pose EstimationGeometric MatchingImage MatchingPose EstimationVisual Localization

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Matching ZEB DKM Mean AUC@5° 46.2 #4 of 10 Archive leaderboard report
Pose Estimation InLoc DKM DUC1-Acc@0.25m,10° 51.5 #4 of 6 Archive leaderboard report
Pose Estimation InLoc DKM DUC1-Acc@0.5m,10° 75.3 #4 of 6 Archive leaderboard report
Pose Estimation InLoc DKM DUC1-Acc@1.0m,10° 86.9 #4 of 6 Archive leaderboard report
Pose Estimation InLoc DKM DUC2-Acc@0.25m,10° 63.4 #4 of 6 Archive leaderboard report
Pose Estimation InLoc DKM DUC2-Acc@0.5m,10° 82.4 #4 of 6 Archive leaderboard report
Pose Estimation InLoc DKM DUC2-Acc@1.0m,10° 87.8 #4 of 6 Archive leaderboard report
Visual Localization Aachen Day-Night v1.1 Benchmark DKM Acc@0.25m, 2° 70.2 #7 of 7 Archive leaderboard report
Visual Localization Aachen Day-Night v1.1 Benchmark DKM Acc@0.5m, 5° 90.1 #7 of 7 Archive leaderboard report
Visual Localization Aachen Day-Night v1.1 Benchmark DKM Acc@5m, 10° 97.4 #7 of 7 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

ConvolutionDepthwise Convolution

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections