English

Pano Popups: Indoor 3D Reconstruction with a Plane-Aware Network

Computer Vision and Pattern Recognition 2020-02-25 v2

Abstract

In this work we present a method to train a plane-aware convolutional neural network for dense depth and surface normal estimation as well as plane boundaries from a single indoor 360360^\circ image. Using our proposed loss function, our network outperforms existing methods for single-view, indoor, omnidirectional depth estimation and provides an initial benchmark for surface normal prediction from 360360^\circ images. Our improvements are due to the use of a novel plane-aware loss that leverages principal curvature as an indicator of planar boundaries. We also show that including geodesic coordinate maps as network priors provides a significant boost in surface normal prediction accuracy. Finally, we demonstrate how we can combine our network's outputs to generate high quality 3D "pop-up" models of indoor scenes.

Keywords

Cite

@article{arxiv.1907.00939,
  title  = {Pano Popups: Indoor 3D Reconstruction with a Plane-Aware Network},
  author = {Marc Eder and Pierre Moulon and Li Guan},
  journal= {arXiv preprint arXiv:1907.00939},
  year   = {2020}
}

Comments

2019 International Conference on 3D Vision (3DV). IEEE, 2019

R2 v1 2026-06-23T10:09:04.882Z