Planar Surface Reconstruction from Sparse Views

TL;DR

This paper introduces a method for reconstructing planar surfaces in indoor scenes from only two views with unknown camera poses, improving the state of the art in sparse-view scene reconstruction.

Contribution

It jointly estimates camera pose and reconstructs planar surfaces, leveraging the dominance of planes in indoor environments, which was not effectively utilized in prior methods.

Findings

Outperforms previous methods on Matterport3D scenes

Successfully reconstructs dominant planar structures from sparse views

Jointly estimates camera pose and scene geometry

Abstract

The paper studies planar surface reconstruction of indoor scenes from two views with unknown camera poses. While prior approaches have successfully created object-centric reconstructions of many scenes, they fail to exploit other structures, such as planes, which are typically the dominant components of indoor scenes. In this paper, we reconstruct planar surfaces from multiple views, while jointly estimating camera pose. Our experiments demonstrate that our method is able to advance the state of the art of reconstruction from sparse views, on challenging scenes from Matterport3D. Project site: https://jinlinyi.github.io/SparsePlanes/