1 code implementation • CVPR 2023 • Daniel Barath, Denys Rozumny, Ivan Eichhardt, Levente Hajder, Jiri Matas
Dominant instances are found via a RANSAC-like sampling and a consolidation process driven by a model quality function considering previously proposed instances.
1 code implementation • CVPR 2021 • Daniel Barath, Dmytro Mishkin, Ivan Eichhardt, Ilia Shipachev, Jiri Matas
We propose ways to speed up the initial pose-graph generation for global Structure-from-Motion algorithms.
1 code implementation • ECCV 2020 • Ivan Eichhardt, Daniel Barath
We propose a new approach for combining deep-learned non-metric monocular depth with affine correspondences (ACs) to estimate the relative pose of two calibrated cameras from a single correspondence.
1 code implementation • 1 May 2019 • Ivan Eichhardt, Daniel Barath
The technique requires the epipolar geometry to be pre-estimated between each image pair.
no code implementations • ECCV 2018 • Ivan Eichhardt, Dmitry Chetverikov
This paper presents a novel algorithm to estimate the relative pose, i. e. the 3D rotation and translation of two cameras, from two affine correspondences (ACs) considering any central camera model.