Learning a Depth Covariance Function

CVPR 2023 pp. 13122-13131

doi:10.1109/CVPR52729.2023.01261 /cvpr/2023/dexheimer2023cvpr-learning/

Abstract

We propose learning a depth covariance function with applications to geometric vision tasks. Given RGB images as input, the covariance function can be flexibly used to define priors over depth functions, predictive distributions given observations, and methods for active point selection. We leverage these techniques for a selection of downstream tasks: depth completion, bundle adjustment, and monocular dense visual odometry.

PDF CVPR Semantic Scholar

Cite

Text

Dexheimer and Davison. "Learning a Depth Covariance Function." Conference on Computer Vision and Pattern Recognition, 2023. doi:10.1109/CVPR52729.2023.01261

Markdown

[Dexheimer and Davison. "Learning a Depth Covariance Function." Conference on Computer Vision and Pattern Recognition, 2023.](https://mlanthology.org/cvpr/2023/dexheimer2023cvpr-learning/) doi:10.1109/CVPR52729.2023.01261

BibTeX

@inproceedings{dexheimer2023cvpr-learning,
  title     = {{Learning a Depth Covariance Function}},
  author    = {Dexheimer, Eric and Davison, Andrew J.},
  booktitle = {Conference on Computer Vision and Pattern Recognition},
  year      = {2023},
  pages     = {13122-13131},
  doi       = {10.1109/CVPR52729.2023.01261},
  url       = {https://mlanthology.org/cvpr/2023/dexheimer2023cvpr-learning/}
}