Learning Geometrically Consistent Mesh Corrections

Ștefan Săftescu, Paul Newman

Abstract

Building good 3D maps is a challenging and expensive task, which requires high-quality sensors and careful, time-consuming scanning. We seek to reduce the cost of building good reconstructions by correcting views of existing low-quality ones in a post-hoc fashion using learnt priors over surfaces and appearance. We train a convolutional neural network model to predict the difference in inverse-depth from varying viewpoints of two meshes – one of low quality that we wish to correct, and one of high-quality that we use as a reference. In contrast to previous work, we pay attention to the problem of excessive smoothing in corrected meshes. We address this with a suitable network architecture, and introduce a loss-weighting mechanism that emphasises edges in the prediction. Furthermore, smooth predictions result in geometrical inconsistencies. To deal with this issue, we present a loss function which penalises re-projection differences that are not due to occlusions. Our model reduces gross errors by 45.3%–77.5%, up to five times more than previous work.

Download


Paper Citation


in Harvard Style

Săftescu Ș. and Newman P. (2020). Learning Geometrically Consistent Mesh Corrections.In Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP, ISBN 978-989-758-402-2, pages 664-675. DOI: 10.5220/0008947806640675


in Bibtex Style

@conference{visapp20,
author={Ștefan Săftescu and Paul Newman},
title={Learning Geometrically Consistent Mesh Corrections},
booktitle={Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP,},
year={2020},
pages={664-675},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0008947806640675},
isbn={978-989-758-402-2},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP,
TI - Learning Geometrically Consistent Mesh Corrections
SN - 978-989-758-402-2
AU - Săftescu Ș.
AU - Newman P.
PY - 2020
SP - 664
EP - 675
DO - 10.5220/0008947806640675