Back to the Feature: Learning Robust Camera Localization from Pixels to Pose

Sarlin, Paul-EdouardEidgenössische Technische Hochschule Zürich (ETH),Swiss Federal Institute of Technology in Zürich (ETH) (author)

Camera pose estimation in known scenes is a 3D geometry task recently tackled by multiple learning algorithms. Many regress precise geometric quantities, like poses or 3D points, from an input image. This either fails to generalize to new viewpoints or ties the model parameters to a specific scene. In this paper, we go Back to the Feature: we argue that deep networks should focus on learning robust and invariant visual features, while the geometric estimation should be left to principled algorithms. We introduce PixLoc, a scene-agnostic neural network that estimates an accurate 6-DoF pose from an image and a 3D model. Our approach is based on the direct alignment of multiscale deep features, casting camera localization as metric learning. PixLoc learns strong data priors by end-to-end training from pixels to pose and exhibits exceptional generalization to new scenes by separating model parameters and scene geometry. The system can localize in large environments given coarse pose priors but also improve the accuracy of sparse feature matching by jointly refining keypoints and poses with little overhead. The code will be publicly available at github.com/cvg/pixloc.

Unagar, AjaykumarEidgenössische Technische Hochschule Zürich (ETH),Swiss Federal Institute of Technology in Zürich (ETH) (author)
Larsson, Mans (author)
Germain, HugoÉcole des Ponts ParisTech (author)
Toft, Carl,1990Chalmers tekniska högskola,Chalmers University of Technology(Swepub:cth)tcarl (author)
Larsson, ViktorEidgenössische Technische Hochschule Zürich (ETH),Swiss Federal Institute of Technology in Zürich (ETH) (author)
Pollefeys, MarcEidgenössische Technische Hochschule Zürich (ETH),Swiss Federal Institute of Technology in Zürich (ETH) (author)
Lepetit, VincentÉcole des Ponts ParisTech (author)
Hammarstrand, Lars,1979Chalmers tekniska högskola,Chalmers University of Technology(Swepub:cth)e8rull (author)
Kahl, Fredrik,1972Chalmers tekniska högskola,Chalmers University of Technology(Swepub:cth)kahlf (author)
Sattler, Torsten,1983Chalmers tekniska högskola,Chalmers University of Technology(Swepub:cth)torsat (author)
Eidgenössische Technische Hochschule Zürich (ETH)École des Ponts ParisTech (creator_code:org_t)

In:Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, s. 3246-32561063-6919

By the author/editor: Sarlin, Paul-Edo ...; Unagar, Ajaykuma ...; Larsson, Mans; Germain, Hugo; Toft, Carl, 1990; Larsson, Viktor; show more...; Pollefeys, Marc; Lepetit, Vincent; Hammarstrand, La ...; Kahl, Fredrik, 1 ...; Sattler, Torsten ...; show less...

About the subject

NATURAL SCIENCES: NATURAL SCIENCES; and Computer and Inf ...; and Bioinformatics

ENGINEERING AND TECHNOLOGY: ENGINEERING AND ...; and Other Engineerin ...; and Media Engineerin ...

NATURAL SCIENCES: NATURAL SCIENCES; and Computer and Inf ...; and Computer Vision ...

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.